The short answerYes, mostly. Middlesbrough's opponent-adjusted xG difference is +1.59 a game, the best in the Championship by 0.64, against the 4th-toughest schedule. They have scored fewer goals than their chances deserve. The risk is defensive: chances against are up from 0.92 to 1.36 a game.

Middlesbrough are top of the Championship, and the underlying numbers say they deserve to be. Once every game is judged against the quality of the opponent, Kim Hellberg's side have created +1.59 more non-penalty expected goals than they have conceded per game. That is the best figure in the division by 0.64, and nobody else is within two-thirds of it. West Ham are second on +0.95, Wolves third on +0.74.

That's the case in one paragraph. Boro have 25 points from 12 games and lead Swansea by 2. More striking is how they have got there. In each of the 11 league games in our chance data, they created more than their opponents. Their results have actually lagged behind their football: they have scored 22 goals from chances worth 31.0. A team that is top while underperforming its own chances is a different proposition from a team that is top on a hot streak.

So is it built to last? Mostly, yes. The attack is the best in the league, and it is stronger now than at any point in Hellberg's 11 months. The question is the other end of the pitch, where he has knowingly traded some control for more chances.

Better than the table, not just top of it

Raw expected goals can flatter a side that has played weak opponents. So we compared each Championship game with what that opponent creates and concedes in its other matches. Beat a side that normally gives away very little, and the chances you created count for more. Concede a lot to a team that rarely creates anything, and the adjustment counts against you.

For Middlesbrough, the adjustment helps. Their raw margin is +1.46 per game (2.82 for, 1.36 against). Their opponents so far have been the 4th-toughest set in the league, and allowing for that adds 0.13 a game. They have not been fattening up on the bottom half.

One game skews the average. Against Cardiff on 17 October, Boro had 36 shots and chances worth 5.35 expected goals, and drew 2-2 after coming from two goals down. Take that game out and their adjusted margin falls to +1.22. That's still the best in the division.

The Index agrees, more cautiously. On our ratings, Middlesbrough are 1506, 89th in the world and 3rd in the Championship, level on rating with Wolves and behind West Ham (1608). The gap to West Ham is mostly history. At this point in a season, the current campaign carries only about 26% of a club's within-league score, so West Ham's recent Premier League seasons still count for a lot. On this season's football alone, Boro's expected points rate of 2.29 a game is the best in the division. Their 25 points are about 2.5 fewer than that football would normally earn. The Index rates them as rising.

Their raw margin is +1.46 per game (2.82 for, 1.36 against).

Eleven months of Hellberg

Hellberg was appointed on 24 November 2025, from Hammarby, after Rob Edwards left for Wolves. The seven league games in our data from just before the change were ordinary: a margin of +0.34 a game in raw terms, and +0.12 once opponents were counted.

Under Hellberg, the margin jumped. Across his 29 league games last season, Boro created 1.85 non-penalty expected goals a game and allowed 0.92, an opponent-adjusted margin of +0.89. They finished fifth on 80 points. They lost the play-off semi-final to Southampton, were reinstated when Southampton were expelled over the spying case, and then lost the final 1-0 to Hull City at Wembley.

This season, the attack has taken another step. Boro's 2.82 expected goals a game is up from 1.85 last season, and it's the most in the league. The defence has moved the other way: 1.36 a game against, up from 0.92. That is still the 4th-lowest figure in the Championship. Boro have kept only two clean sheets, though, and the 25 goals scored and conceded in their six games before the Bolton trip showed how open some of their matches have become. Hellberg has been clear that this is a choice. "You have to pick in football what you want to be and we want to be an offensive team," he told BBC Radio Tees last week.

How they play

The shape on the team sheet changes. This season Boro have started most often in a 3-4-2-1 or a 4-2-3-1, three times each, and have also used a 4-3-3, a 4-4-2 and a 4-1-4-1. The principles don't change. On our style measures, Boro have the most box presence, the most deep completions and close to the most possession in the division, each more than two standard deviations above the league norm.

In practice, that means a patient possession side that pins teams back and then floods the box. They lead the league in positional attacks (33.9 a game against an average of 26.9), shots (18.8) and crosses (20.7). They are near the bottom for counter-attacks, 18th of 24, because opponents rarely leave them space to break into. Their chances are good ones, too: 0.15 expected goals per shot, against a league average of 0.13. Volume and quality together are rare.

The defensive side of the model is control rather than a deep block. They face 10.5 shots a game, the 4th-fewest, because they have the ball so much. The cost is what happens when they lose it. As the volume of attacks has gone up this season, so have the chances against, which is what you would expect from a side committing more players forward. Hellberg's own description after the 5-1 win at Bolton was "We used the space behind the last line… then we played fluid football". That's a side willing to play through and behind a block, not just around it.

Is it sustainable?

Three things say yes. First, the margin is large enough to absorb a lot of regression. Even if Boro's chance creation fell back to last season's level, they would still be among the best sides in the division. Second, it isn't built on finishing. Boro have scored about nine fewer goals than their chances were worth, so the results have more room to improve than to fall. Will Lankshear has 11 league goals in 12 games and is the division's top scorer, but the team was creating chances before he was taking them. Third, it has lasted. Hellberg's Boro have been the better side on chances for most of a calendar year, against good opponents as well as bad ones.

Two things say watch carefully. The goals against are up, and a team that concedes 1.36 expected goals a game will keep drawing matches it dominates, as it did against Millwall and Cardiff. And the squad is being tested. David Strelec and Alfie Jones have both been out, Rokas Pukstas has a longer-term hamstring injury, and Hellberg drew criticism for resting Lankshear and Sebastian Berhalter against Cardiff. With "game, game, game", as he put it, the attack has to keep its level through rotation.

The rivals' numbers

The two sides closest to Boro on chances are worth a look, because they show what kind of lead this is. West Ham have scored the most goals in the division, but their adjusted margin of +0.95 has come against the 12th-toughest schedule, close to the middle. Wolves are third on +0.74. Both are good teams. Neither has matched Boro for volume. Middlesbrough have taken 18.8 shots a game, West Ham 15.8 and Wolves 15.0. Further down, Swansea are second in the table but only 10th on adjusted chances. That gap is the subject of our Championship real table.

The next test is Charlton at the Riverside on Saturday. Charlton have the 22nd-worst adjusted margin in the league, so it's a game Boro's numbers say they should control. The more useful evidence will come from the run after it, against the sides around them.

Our verdict

Top on merit, and built to last. Middlesbrough have the best football in the Championship this season, by a margin no other side comes close to, and against a tougher schedule than most. The attack is more than one striker's hot streak, and it is still improving under Hellberg rather than levelling off. The one thing to watch is how much they give away to get it. Two numbers matter over the next ten games. Does their expected goals against stay near 1.3 a game, or creep towards the league average of 1.63? And do the goals start to match the chances? If the first holds, the second should follow.

The short answerMost football data records events on the ball: passes, shots, tackles, carries. A run that drags a defender away, or a press that forces a hopeful clearance, often creates no event at all, so it earns nobody any credit. Our numbers have the same blind spot. We can see its effects in team totals, but we can't yet price the run itself.

Jude Bellingham did not score in the Clásico and did not register the assist for either goal. Our tactical analysis still argued that his night was won without the ball: a big tackle in the 33rd minute started the move that ended with Kylian Mbappé's overhead kick over the bar, and after half-time he was part of the block that held Barcelona to one goal.

Look at the numbers for that night and you will find the tackle, because a tackle is an event. What you won't find is most of the rest: where he stood so that passes never went his way, the ground he covered to get into that block, the space he closed before anyone tried to use it. When he wasn't touching the ball, the data barely knew he was there.

That is the problem with football data in one moment. It is very good at describing the ball. It is much worse at describing the 21 other people on the pitch.

What the numbers actually record

Almost every football statistic you have seen, ours included, is built from event data. Someone, or something, logs each time a player does something to the ball: a pass, a shot, a tackle, an interception, a dribble, a clearance. Each event has a time, a place, a player and an outcome.

That is a remarkable amount of information. It is how we know that Middlesbrough lead the Championship in shots (18.8 a game) and in positional attacks (33.9 a game, against an average of 26.9). It is how we can say that Ture Spendler is first of 63 Superettan full-backs for progressive passes.

But an event needs a ball. A centre-forward who runs across the near post to pull a centre-back with him, so that the winger has a free header at the far post, has done his job perfectly. In the event log he does not exist. The winger gets the shot. The crosser gets the chance created. The decoy gets nothing.

Even the busiest player has the ball for only a small part of a match. Event data describes those moments in great detail and the rest barely at all.

It is how we can say that Ture Spendler is first of 63 Superettan full-backs for progressive passes.

Bellingham, again

Our tactical analysis of the Clásico was headlined around one idea: Bellingham's night was won without the ball.

He didn't score. He didn't register the assist for either goal. On the ball his night was mixed: twice he had a chance to release Mbappé on a break and didn't find the pass. A box-score reading of his game would be a shrug.

Much of what made his night happened off the ball. A big tackle in the 33rd minute started a move that ended with Mbappé's overhead kick over the bar. After half-time, when Barcelona pushed, Bellingham was part of a block defending the Madrid box, not a spectator to it. That block held Barcelona, who had been scoring heavily all season, to one goal.

Even his best moment on the ball barely registers. Three minutes from time, his turn under pressure started the move for Mbappé's winner. The assist went to Eduardo Camavinga, who made the cutback. In the box score, Bellingham's part in the goal that decided the Clásico does not appear.

Some of that does show up. A tackle is an event. But "was in the right place so that the pass never went there" is not. The best defensive work often looks, in the data, like nothing happening.

Pressing: we can count the result, not the effort

Pressing is the clearest case of off-ball work that the numbers half-see.

We measure a team's pressing as the number of passes it allows its opponent before it makes a defensive action: a tackle, an interception, a foul, a challenge. Lower means more aggressive. When Barcelona went to Betis, they allowed 7.4 passes per defensive action. Betis allowed Barcelona 38.3. The LaLiga median this season is about 11.6.

That number is useful. It told us, accurately, that Barcelona squeezed and Betis retreated. But look at what it is made of. Every term in it is an on-ball event. A forward who sprints 30 yards to close down a centre-back, forcing a long ball that goes out of play, has done exactly what his coach wanted. Unless he touches the ball, his sprint isn't in the number.

We said as much in that piece: these counts tell you who made a tackle, not where on the pitch it happened. We couldn't split the numbers by half either, so we couldn't say how much of the Betis retreat was choice and how much was the red card.

The same limit hits players. When we wrote about Bruno Fernandes, the warning sign was that his successful defensive actions had fallen to the 6th percentile of his position (2.72 per 90). That is a fair number to worry about. But it measures the tackles and interceptions he completes. It cannot see whether he is still running to close passing lanes, or doing it a second later than he used to. Two players with the same count can be pressing very differently.

"Progressive runs" are not runs

A small confession about our own vocabulary.

When our scout reports talk about a player's progressive runs, as they did for Fernandes (down to the 10th percentile, 0.39 per 90) or Marcus Linday (91st percentile), they mean running with the ball: carries that move it meaningfully towards goal. They don't mean a run without it.

The name is standard, and we have kept it. But it is a neat example of the problem. Even the word "run", in football data, has quietly been taken over by the ball.

Where off-ball work leaves fingerprints

None of this means off-ball work is invisible. It leaves marks, mostly in team numbers.

Middlesbrough are the best example we have published. On our style measures they have the most box presence in the Championship, close to three standard deviations above the league norm. They create the most expected goals per game (2.82), and their chances are better than average: 0.15 expected goals per shot, against 0.13 for the league.

None of that happens without players making runs into the box. The attacking numbers are, in a sense, the receipt for all those runs. What we can't do is say whose runs they were. The box presence belongs to the team. The expected goals are assigned to whoever shot. The forward who made the run that freed the shooter adds to the total and gets none of the credit.

Defence works the same way. We have adjusted interceptions for how much of the ball a team has, because a side that is always defending gets more chances to intercept. That adjustment is why we could say that Lee Dae-gwang, a K League 2 striker, was in the top 5% for interceptions and the top 3% for successful defensive actions in early October. That suggests a forward who works hard without the ball. It is still an inference from the moments he wins it. A striker whose positioning means the pass is never attempted makes no interception at all.

When David Raum was in the 94th percentile for interceptions among Bundesliga full-backs last season, we wrote that he reads the game well. That is a reasonable reading, but it is a reading. Interceptions are the visible tip of good positioning. Plenty of good positioning never produces one.

The step-up blind spot

This matters most for the thing we care about most: saying which players can step up a level.

When we wrote about Spendler, the key evidence was that full-backs moving from the Superettan to the Allsvenskan kept a median 94% of their progressive passing. We can test that because passing is an event. We have the before and the after.

We can't run the same test on off-ball work, because we don't have it in the first place. If a full-back's real strength is recovery runs, or holding the line at the right moment, our step-up evidence will undersell him. If his weakness is drifting out of position when the ball is on the other side, our numbers may oversell him. Spendler's defensive actions rose from the 24th percentile to the 95th over three seasons. That is a real improvement in what he does when he engages. It is not, on its own, proof that his positioning has improved.

That is one reason every scout report we publish leans on watching the player as well as the numbers, and why we say when a verdict rests mostly on the ball.

How you would actually measure it

The honest answer is that you need to know where everyone is, all the time, not just where the ball is.

With that kind of positional data, you can start to ask the right questions. Which runs pulled a defender out of shape? Which run opened the lane that the pass went through? Which presser cut off the easy pass and forced the long one? You can then compare the chances a team had with the run against the chances it would probably have had without it. That gives the run a value even when the runner never touches the ball.

Work like that exists. It isn't yet in the numbers most of the public sees, and it isn't in ours. Our ratings, our chance models and our scout-report percentiles are built from what happens to the ball, and adjusted, carefully, for opponents and league strength.

What we promise

Three things.

First, when a verdict rests on on-ball numbers, we will say so. If a player's case depends on what he does without the ball, we will tell you that our numbers can't fully see it.

Second, we will keep separating the team fingerprint from the individual credit. Middlesbrough's box presence is a team fact. It is not evidence that any one Boro forward is good.

Third, we will keep watching the games. Bellingham's Clásico showed up in a box score as no goal and no assist. Most of what made it a good night would not appear there at all.

Somewhere in that match, runs and positions helped win a Clásico. Nobody counted it. We'd like to.

The short answerTure Spendler, Norrby's 23-year-old left-back, ranks first of 63 Superettan full-backs for progressive and long passes, and has improved in each of the last three seasons. Full-backs who stepped up from the Superettan to the Allsvenskan kept a median 94% of their progressive passing. His level: an Allsvenskan starter.

Of the 63 full-backs who have played 270 or more minutes in Sweden's Superettan this season, the one who moves the ball forward most is Norrby's left-back. Ture Spendler makes 11.5 progressive passes per 90 minutes, against a full-back median of 6.7. He also makes 10.4 long passes per 90, three times the median. He's first in the league's full-back pool for both. That isn't new. And when Superettan full-backs have stepped up to the Allsvenskan, this is the part of the game that has mostly come with them.

That's the case in one paragraph. Spendler is 23, and his numbers have got better in each of the last three seasons, the last of those a division higher. The players who have already made the jump he would make next suggest he would arrive in the Allsvenskan with his main strength largely intact. Our view is that he is a top-flight starter in waiting.

This is a Step Up piece. Each week we take one player from a lower league and ask a single question: what level is he really playing at, and what's the evidence he could do it one level higher? We first picked out Spendler in our list of eight players nobody is watching, as the safest bet on that list that his season was real. Here's the full case.

Three seasons, three steps up

Spendler arrived at Norrby from fourth-tier Räppe before the 2024 season, after coming through the academy at Öster. His first year was in the third tier, Ettan, and it was useful rather than special: 2,267 league minutes, good at long passing (the 85th percentile among Ettan full-backs), mid-table at progressive passing (the 67th), and well below average without the ball. He was in the 24th percentile for successful defensive actions and the 15th for interceptions.

In 2025, still in Ettan, he jumped. Over 2,502 minutes his progressive passing rose to the 95th percentile, his defensive actions to the 66th and his interceptions to the 74th. Norrby went up.

A step up a division is normally where a lower-league season gets found out. Spendler's numbers went up again. In the 2026 Superettan, against a pool of 63 full-backs, he is first for progressive passes and long passes, third for passes into the final third and third for successful defensive actions, which puts him in the 95th percentile. His interceptions are in the 85th. He has played 2,605 league minutes across 27 games, and public match logs show him starting every league game this season.

That's the answer to the spike-or-steady question, and it isn't close. The passing has been strong for three years and elite for two. The defending started poor and has climbed every season, in a better league each time it mattered: from the 24th percentile to the 66th to the 95th. A spike looks like one good season. This looks like a player who keeps getting better.

Over 2,502 minutes his progressive passing rose to the 95th percentile, his defensive actions to the 66th and his interceptions to the 74th.

Is it him, or is it Norrby?

There's an obvious objection. On our style measures, Norrby are one of the most direct teams in the Superettan, with a long-ball share two standard deviations above the league norm. A left-back in a team that goes long will make a lot of long passes. So how much of Spendler's profile is just the system?

Look at the other side of the pitch. Norrby's two main right-backs this season, in the same team and the same system, average about 6.2 progressive passes per 90 each. That's roughly the league median, and about half of Spendler's 11.5. The team's style explains some of the long balls. It doesn't explain why the left-back makes nearly twice as many forward passes as the right-backs. Long passes make up 23% of everything he plays, the highest share of any Superettan full-back, so he is the team's main outlet from the back.

It also tells you what he is and isn't. Spendler moves the ball by passing it, not by running with it. He's in the 15th percentile for progressive runs and the 18th for dribbles. His accuracy on progressive passes is mid-table, in the 47th percentile, so the volume comes with risk. This isn't a winger who has been moved back. He's a deep-lying passer who has been given the left touchline.

The step-up test

The next level for a Superettan full-back is the Allsvenskan, and the gap is real. On our ratings, the average Allsvenskan club is about 210 points stronger than the average Superettan club. At roughly 180 points a goal, that's more than a goal a game. Norrby, rated 988, are 12th in the Superettan.

So we went looking for players who have already made the move. Since 2023 there have been 15 cases of a Superettan full-back playing 900 or more league minutes and then playing 900 or more as a full-back in the Allsvenskan the next season. Some went up with their clubs and some moved to bigger ones. We compared their progressive passing either side of the move.

The pattern is clearer than you might expect. The median full-back kept 94% of his progressive passing after stepping up. Eight of the 15 kept 90% or more, and a quarter kept less than 79%. Defending didn't fall at all: the median mover made 6% more successful defensive actions per 90 in the Allsvenskan, which fits a full-back spending more of his time without the ball. The cases most like Spendler's are even more encouraging. Five movers arrived with 10 or more progressive passes per 90, and three of the five kept at least 98% of it. Spendler's 11.5 is higher than any of the 15 managed before they moved.

Now apply that to him. If he keeps the median share, he'd make about 10.8 progressive passes per 90 in the Allsvenskan. That would rank 4th of the 51 Allsvenskan full-backs with 270 or more minutes this season. If he has a harder landing and keeps only the lower-quartile share, it's about 9.1, which would still rank 15th, in the top third of the league.

The blunter version of the test is to drop his raw 2026 numbers straight into this season's Allsvenskan pool. They would rank in the 94th percentile for progressive passes, the 100th for long passes and the 94th for defensive actions. Those raw numbers won't survive the move unchanged. Better opponents press harder and give a left-back less time to pick out a long pass. But the movers tell you how much usually survives. For this skill, it's most of it.

The one number that doesn't travel as well is his duels. He wins 60% of his defensive duels, the 83rd percentile in the Superettan. Placed raw in the Allsvenskan pool, that falls to the 57th, about average. Wingers in the top flight are quicker and better, and his one-on-one defending is the part of his game an Allsvenskan coach would test first.

The level, and the fit

Put it together and the level is clear: a starting left-back in the Allsvenskan, at a mid-table club, with his passing as the reason to sign him. That's one division up. It isn't a leap to a top-five league, and nothing here argues for one yet.

The type of club matters as much as the league. A patient, short-passing side would ask him to do the thing he does least, carrying the ball. A side that wants its full-backs to switch play and find runners early would get the most out of him. Two Allsvenskan clubs fit that on our style measures. These are our suggestions based on fit and level, not clubs reported to be interested.

Brommapojkarna (1205, 11th in the Allsvenskan on our ratings). They're one of the league's most direct sides, A left-back who already plays a quarter of his passes long would fit how they want to attack, though his long-pass accuracy, mid-table in the Superettan, is something they would want to see improve.

Mjällby (1198, 12th). One of the most active defensive teams in the league on possession-adjusted interceptions. A full-back whose defensive actions and interceptions have climbed every season would suit a side that wants to win the ball back early.

Swedish press reported interest in him from an Allsvenskan club and from Norway's Sandefjord before this season. In the summer, a supposed approach from abroad turned out to be a scam, which Norrby reported to the police. He's still at Norrby, on a contract that runs to the end of 2027.

Our verdict

Step-up level: Allsvenskan starter. Spendler is the most productive forward passer among Superettan full-backs. He has improved in each of three seasons, including the one in which he went up a division. The full-backs who made this exact jump before him mostly kept their forward passing, so the evidence says his would travel too. His one-on-one defending is the part to watch: he wins his duels at this level, but on raw numbers he'd be only average at the next. Watch two things over his next ten games. Does his progressive passing stay above ten per 90 against the Superettan's better sides? And do his defensive duels hold up against the league's best wingers? If both do, he's ready.

The short answerBruno Fernandes had his most creative Premier League season in 2025-26, with 0.37 expected assists per 90, the 99th percentile among midfielders. This season his passing is still elite, but his progressive runs and defensive actions have fallen sharply. The best fit is a deeper creative role, ideally at Manchester United.

Bruno Fernandes is 32 and his contract runs out in June 2027. Manchester United have an option to extend it by a year, and talks over a longer deal are reported to be stuck on one question: how many more years he has at the top. Reports say Al Hilal remain interested, a Galatasaray link has cooled, and pundits are already asking whether January is the moment to cash in. Nothing has been agreed and United are reported to want to keep him. We can't tell you what will happen. We can test the assumption behind the whole debate: that he is about to decline.

Here's the non-obvious part. On our numbers, the decline everyone is pricing in hasn't reached his passing. His last full season was the most creative of the six we can rank. What is going is something else: the running and the defending. That changes which club, and which role, makes sense for him.

Six seasons in the top 6%

We ranked Fernandes against every Premier League midfielder with 900 league minutes in each season from 2020-21 to 2025-26, and against those with 270 minutes so far this season. His percentiles for expected assists, the quality of the chances he sets up regardless of the finish, read 96, 95, 96, 99, 94, 99. He has been in the top 6% of the league's midfielders in every one of those seasons.

Then look at 2025-26. He played 3,403 league minutes, scored 9 and set up 18. His expected assists came to 14.1, or 0.37 per 90, his highest rate in the six full seasons in our data. His 2.88 shot assists per 90 were the most of any Premier League midfielder with 900 minutes. A player whose creative engine was failing would not produce his best creative season at 31 going on 32.

This season's sample is small, eight games and 696 minutes, so we hold it loosely. But the passing signals are the same or stronger. He is in the 100th percentile among Premier League midfielders for progressive passes and for passes into the box, and the 99th for shot assists. He has 4 league goals from chances worth 3.4. His expected assists percentile has dipped to 82, still well inside the top fifth.

What is actually changing

The real signal is in the bottom rows of the chart. Fernandes has never been a big tackler. His percentile for successful defensive actions has sat between the 19th and 53rd over six seasons. But he has always carried the ball, in the 57th to 67th percentile for progressive runs every year. This season both have dropped: progressive runs to the 10th percentile (0.39 per 90) and defensive actions to the 6th (2.72 per 90).

Two cautions. Eight games is a small sample, and part of the change could be role rather than legs: a player asked to sit deeper and pass will carry less. Even so, it is the pattern you'd expect at 32. Passing range and vision age slowly. The repeated sprints of a pressing number 10 do not. Any club signing him, or extending him, is buying the first and should stop paying for the second.

This season's sample is small, eight games and 696 minutes, so we hold it loosely.

What that means for the fit

That is the useful answer to "should he stay or go". The question isn't whether Fernandes is still good. On creation he is still elite by any measure we have. The question is whether the team around him asks him to do the things that are fading. A side that needs its number 10 to lead the press and carry the ball 40 yards will see the decline first. A side that gives him the ball deeper, with runners ahead and a ball-winner beside him, will get the best of what he still has.

United, on our ratings, are 1750, 5th in the Premier League and 10th in the world. Fernandes's numbers say his best version for the next three years is probably as one of the deeper midfielders, receiving the ball facing play, rather than as the attacking midfielder behind the striker who is expected to press first. That's a coaching decision, not a transfer.

The destinations, on our ratings

Stay at Manchester United. Our model says this is the highest level available to him by a distance. United, at 1750, are the strongest side in the picture, and his passing numbers are still among the best in the league. If the contract can be structured around a deeper role and fewer minutes, staying is the best football fit.

Al Hilal. Reported to remain interested. They are rated 1678 and 27th in the world on our ratings, about 72 points below United. That is roughly the level of a mid-table Premier League side. The style question matters less there: a dominant team in its league lets a creator play deeper and receive the ball. It is the most comfortable football fit for the player he is becoming, and the biggest step down in competition.

Galatasaray. Linked in the summer, and that interest is now reported to have cooled. Rated 1600, 45th in the world, 150 points below United. They play in Europe and dominate the ball in their own league, which suits a deeper creator. It's a real step down, but it keeps him in the Champions League conversation.

We also show two clubs nobody has linked with him, Benfica (1739) and Sporting CP (1694), for context only. He made his name in Lisbon with Sporting before moving to Manchester. On our ratings the Portuguese giants sit in the same band as the clubs that have been linked.

Our verdict

Creation: elite. Legs: fading. Best fit: United, in a deeper role. The case for selling Fernandes rests on the idea that he's about to fall off. The numbers don't show that in his passing. His last full season was his most creative, and this season's passing is still at the very top. What they do show is a player doing less running and less defending, which is exactly how a creative midfielder should be managed into his mid-30s. United's sensible move is a contract that pays for the passing and a role that stops asking for the running. A buyer's sensible move is the same. Pay for three good years of passing, not for a pressing number 10.

There's a wider lesson here for anyone valuing older creative players. The public debate tends to treat age as one number that drags every part of a player's game down together. The data rarely works like that. Different skills age at different speeds, and the smart contract is the one that pays for what is still at the top.

What would change our mind? If his expected assists stay below the 85th percentile over a full half-season, the decline would have reached the thing that matters. Until then, the argument that he's finished is about his legs, not his passing.

The short answerDortmund’s rating falls −21 points once opponents are counted; Bremen’s moves −4. Bremen’s schedule ranks 3rd of 18 in the Bundesliga for toughness, Dortmund 13th. Ratings still favour Dortmund by about a goal a game, but the season gap is narrower than the brands suggest.

Borussia Dortmund host Werder Bremen in the DFB-Pokal second round tonight at Signal Iduna Park, 19:45 UK. On the league table and on our ratings this looks straightforward: Dortmund are third in the Bundesliga on our numbers, Bremen eighth, and the rating gap is about 1.1 goals a game. The same two sides drew 2-2 in the league earlier this month. Once you count who each club has actually played, the cup tie looks closer than the brands suggest.

That is the Index question for tonight. Not “who is the better club”, because Dortmund still are. The question is whether Dortmund’s season so far has been padded by a soft run of fixtures, and whether Bremen’s has been stress-tested by a hard one.

Signal Iduna Park on a Pokal night is a different animal from a Sunday league fixture, but the data question travels. Dortmund’s rating falls by 21 points once every game is judged against the opponent. Bremen’s barely moves. Bremen have also faced one of the tougher schedules in the division. That is still an edge for the hosts. It is not a procession.

The schedule gap

From this week, every performance in our ranking is weighed against the opponent it came against. Dortmund’s opponent adjustment is −21 rating points: the soft run has been costing them on the scale, not padding them. Bremen’s adjustment is −4. On schedule toughness within the Bundesliga, Bremen rank 3rd of 18; Dortmund rank 13th. Same division, different tests.

What the league games have looked like

Dortmund’s football still creates. In the league games in our detailed export they averaged more than two expected goals a game at home against Hamburg and Paderborn, and they took a point at Union Berlin in a game that was basically even on chances. The 2-2 with Bremen on 9 October is the useful reference for tonight: Dortmund created the better chances and still dropped points. Cup football, at home, with a crowd that expects a win, is a different night. It is not a different opponent.

Bremen’s season is messier and more interesting. They were hammered 4-1 at Freiburg, then beat Leipzig 3-1 at home despite conceding chances worth more than two expected goals. They have drawn with Köln, Dortmund and Paderborn. The pattern is a side that can hurt anyone on its day and can also leak. Our luck measure has them 3.9 points ahead of what their performances deserve, one of the bigger surpluses in the league. Results have been kinder to Bremen than the chances.

Ratings still favour Dortmund

None of this turns Bremen into favourites. Our ratings put Dortmund at 1660, 29th in the world and third in the Bundesliga. Bremen are 1469, 117th and eighth. Around 180 points is worth a goal a game, so the structural gap is still about a goal. Opponent adjustment answers a narrower question: how much of this season’s rating is about the fixtures rather than the teams.

The 2-2 with Bremen on 9 October is the useful reference for tonight: Dortmund created the better chances and still dropped points.

Dortmund are rated steady. Bremen are rated rising, which matches the eye test of a side that has competed with better opponents than the table admits. In a one-off cup tie, rising plus a hard schedule is exactly the profile that produces an away upset if the favourite is slightly flat.

The cup context

This is the second meeting in three weeks. The league draw already showed Bremen can score twice at this ground. The Pokal adds a different pressure: Dortmund are expected to progress, Bremen are expected to spoil. Cup ties compress variance. A side that has been taking more points than its chances warrant, facing a side whose rating has been trimmed by a soft schedule, is exactly the recipe for a noisy night even when the rating gap is real.

Dortmund’s recent league run includes a goalless draw at Union Berlin in which neither side created much, and a home draw with Bremen that will sit in both dressing rooms tonight. They have also beaten Stuttgart away and put three past Hoffenheim. The floor is high. The ceiling on any given midweek depends on whether the soft-schedule version of Dortmund or the rating version turns up.

Bremen’s away form is the other half of the story. Losses and draws on the road have mixed with the home win over Leipzig. Our rising trend flag on Bremen is not a prediction that they will win in Dortmund. It is a note that the underlying direction has been up while the fixture list has been hard. Strip out the schedule and they look less like an eighth-placed side hanging on and more like a mid-table side that has been stress-tested.

Three things to watch tonight. First, whether Dortmund’s chance volume holds against a Bremen side that has already sat deep at this ground and scored twice. Second, whether Bremen’s finishing runs hot again: they have taken more points than their chances warrant, and cup ties reward that kind of variance. Third, game state. If Dortmund score early, the schedule story probably does not matter. If it is still 0-0 after an hour, the adjusted numbers say Bremen are entitled to believe they belong.

None of this asks you to pick against Dortmund on the longer-run rating. It asks you to stop reading the brands as a formality. A side rated a goal a game better, with a soft-schedule adjustment of about twenty points against them, hosting a rising side with a hard schedule and a luck surplus, is a cup tie with a proper shape. That is the Index preview.

Cup football compresses the season into one night. The Index still prefers Dortmund. The schedule and luck columns are why Bremen get a proper preview rather than a formality line.

The verdict

Dortmund are better. The tie is closer than the brands. Ratings still favour Dortmund by about a goal a game. Bremen’s harder schedule and rising trend mean this is a cup night, not a formality. Respect the gap. Do not invent a different match.

We will update the Index after tonight. For now the preview stands on results to Sunday morning 25 October: Dortmund still ahead on rating, Bremen still closer than the brands suggest once schedules are counted, and the schedule gap still the main reason this cup night should not be waved through as a formality for the hosts.

What would change our mind? If Dortmund’s opponent adjustment stays deeply negative over a month of tougher fixtures, the soft-schedule caveat fades. If Bremen keep outperforming their chances by four points or more, the luck will eventually regress. Tonight is one match. The schedule story is why it is worth watching as a contest rather than a formality.

One more note on method. Opponent adjustment reweights each performance against the opponent faced. League and continental matches count fully; cup matches count half. That is why a Pokal preview can lean on the league-adjusted picture without pretending tonight is just another Bundesliga Sunday.

The short answerAlex Scott’s 2025-26 Premier League season ranks in the top fifth of midfielders for defensive actions and interceptions. Lamine Camara posts a similar ball-winning profile in Ligue 1 at a lower reported fee. Bournemouth’s own rating (1721, 16th in the world) and +28 opponent adjustment help explain a £100m ask; Camara is the value alternative with more adaptation risk.

Liverpool are reported to be weighing a midfield addition, with Bournemouth’s Alex Scott and Monaco’s Lamine Camara the two names most often paired. Separate reports from Germany and from Fabrizio Romano put both on Liverpool’s shortlist for a central midfielder who can start, not just cover. Scott is also linked with Chelsea, with fees around the £90m to £100m mark floated for the England international. Camara was close to Chelsea on deadline day before that deal collapsed, and is discussed at a lower fee. Nothing is agreed. The Index question is simpler than the rumour mill: what does roughly £100m buy, and which profile actually fits?

Transfer windows invent false binaries. Scott or Camara is useful as a comparison of profiles, not as a claim that Liverpool must pick one of these two men by Friday. Plenty of other midfielders will enter the conversation. The reason these two are worth a full Index piece now is that the reporting has converged, the fee talk for Scott has reached nine figures, and Bournemouth’s own numbers make that fee intelligible rather than absurd.

Start with the club Scott would leave. Bournemouth are not a selling club in distress. They are rated 1721, 16th in the world and eighth in the Premier League on our scale. Their opponent adjustment this week is +28 rating points — one of the biggest lifts in the division — even though their schedule ranks only 15th of 20 for toughness. Results have been unkind: they are -4.9 points behind what their football deserved. Paying nine figures for their midfielder is a tax on how good the side around him has been underneath the table.

Alex Scott: Premier League-ready ball-winner

Scott is 23, right-footed, on a contract to June 2028. In 2025-26 he played 3,143 Premier League minutes across 37 games. Among the league’s central midfielders he ranked in the 82nd percentile for successful defensive actions (8.13 per 90), the 86th for interceptions and the 72nd for duels won. He is not a high-volume creator: expected assists sat around the median. He is a midfielder who wins the ball and keeps the side facing the right way.

This season’s sample is small, 599 minutes over seven games, so we hold it loosely. The early signals are stronger, not weaker: he is in the 90th percentile for duels won and the 70th for progressive runs among Premier League midfielders with 270 minutes. He already knows Andoni Iraola’s pressing rules from Bournemouth. That is the premium. You are not buying adaptation time. You are buying a player who has already done the job in this league for a coach who is now at Anfield.

Lamine Camara: the value case

Camara is 22, on a contract to June 2029 at Monaco. In 2025-26 he played 2,033 Ligue 1 minutes. He ranked in the 85th percentile for defensive actions, the 89th for interceptions and the 78th for duels won in the French midfield pool. He also created more than Scott: expected assists in the 81st percentile. Progressive runs were a relative weakness.

The step-up test is the useful one. Place Camara’s raw 2025-26 per-90 numbers in a pool of every central midfielder with 900 minutes in Europe’s top five leagues and he still sits in the 89th percentile for defensive actions and the 85th for interceptions. Those raw numbers would not survive the move unchanged. They do say he starts from a high base. This season at Monaco, in 683 minutes, his duel-winning and defensive actions look even stronger in the Ligue 1 pool, with progressive running still quiet.

This season’s sample is small, 599 minutes over seven games, so we hold it loosely.

What £100m is actually buying

Price is not a rating, but it forces a comparison. Reports put Scott in the £90m to £100m conversation. Camara’s collapsed Chelsea deal was reported around half that. On our club ratings, Bournemouth (1721) are a stronger environment than Monaco (1646), and Scott has already proved himself in the Premier League under the coach Liverpool now employ. That is why the market asks for nine figures. Whether the football premium matches the fee is the open question.

Liverpool, at 1772, do not need a project. They need a midfielder who can press, win duels and let Mac Allister or Gravenberch play. Scott maps onto that job with less translation risk. Camara maps onto a similar defensive profile at a lower reported cost, with more adaptation risk and a little more chance creation. Chelsea’s reported interest in Scott, and their burnt fingers on Camara, are part of the market noise. They do not change the profiles.

How the reports line up

Florian Plettenberg has named Scott and Camara as Liverpool’s leading central-midfield targets, with Felix Nmecha of Dortmund admired but not prioritised. Fabrizio Romano has separately described Liverpool tracking both players and planning a serious midfield addition for 2027, possibly as early as January if the right deal appears. Chelsea’s reported verbal progress on Scott’s personal terms, and their anger over the collapsed Camara deal in the summer, sit in the same news cycle. Two independent reporting lines put both players in Liverpool’s thoughts. That is the press bar. It is not a signing.

The Index does not broker deals. It can say what the football profiles are. Scott is the high-certainty Premier League press midfielder. Camara is the high-upside Ligue 1 destroyer with a bit more chance creation and a lower reported fee. Bournemouth’s level on our scale — and the +28 opponent-adjustment lift — is the hidden character in the Scott story: you are not prising a midfielder out of a struggling side. You are prising one out of a club whose football has been better than its points.

Iraola’s fingerprint matters more for Scott than for almost any other target. A player who has already pressed, covered and carried in that structure arrives with less of the usual six-month tax. That is worth money. Whether it is worth the thick end of £100m is a club decision about wage shape, squad age and alternative uses of the fee. Our job is to show that the premium is about club context as much as about the midfielder.

The verdict

Scott is the fit. Camara is the value. £100m is the Bournemouth tax. If Liverpool want the midfielder who already speaks Iraola’s language and has a full Premier League season of top-fifth ball-winning, Scott is the cleaner answer, and the fee reflects Bournemouth’s own level as much as the player. If they want a younger, cheaper version of the same defensive engine and will accept a season of adaptation, Camara is the rational alternative. Paying nine figures only makes sense if the coaching continuity is the point of the deal.

None of this is a transfer prediction. Fees move and January windows invent their own logic. The Index contribution is narrower: if Liverpool spend near £100m on a central midfielder this cycle, the Bournemouth context explains why Scott can command that money, and Camara’s Ligue 1 profile explains why a lower fee for a different skill mix remains a coherent alternative. We will revisit both when the next builds land.

What would change our mind? If Scott’s duel and interception numbers fall out of the top third over a full half-season, the premium shrinks. If Camara’s progressive running stays in the bottom fifth after a move, he becomes a pure destroyer and needs a different partner. Until then, the Index reads this as a choice between certainty with a markup and upside with a discount, not a choice between good and bad midfielders.

The short answerLiverpool’s opponent adjustment is +9 rating points; Chelsea’s is +7. Chelsea’s schedule ranks 2nd of 20 in the Premier League for toughness, Liverpool’s 12th. Liverpool still rate higher on our Index. Chelsea remain lucky on results versus the chances their football deserved this season.

Liverpool host Chelsea in the Carabao Cup fourth round at Anfield on Wednesday, 20:00 UK. On the league table Chelsea have looked healthier than their football; Liverpool have looked more ordinary than their rating. Once every game is judged against the opponent, both sides move a little — and Chelsea, not Liverpool, are the ones who have faced the harder Premier League run.

That is the Index angle for Anfield. Not a score prediction. A schedule correction.

Raw table talk has been unkind to both clubs in different ways. Liverpool’s rating holds them third in the Premier League and 7th in the world. Chelsea sit sixth in the league on our numbers and 13th globally, with results running ahead of performances. Put the opponent adjustment next to the luck figure and Wednesday becomes a meeting of peers under different kinds of pressure, not a favourite against a crisis club.

Both rise a little once you count the opponents

Liverpool’s opponent adjustment is +9 rating points across 7 league games. Chelsea’s is +7 across 8. Neither is a transformation. Both say the same quiet thing: these sides have not been flattered by soft fixtures. Chelsea’s schedule ranks 2nd of 20 in the Premier League for toughness; Liverpool’s ranks 12th. The harder run belongs to the visitors.

Ratings and luck

On our ratings Liverpool are 1772, third in the Premier League and 7th in the world. They are rated steady. Chelsea are 1736, sixth in the league and 13th in the world, with a falling trend. The gap between them is under a quarter of a goal a game.

Chelsea’s season is the mirror image in one respect. Opponent adjustment lifts them a little, but their results have run ahead of their performances: they are 3.2 points above expected, even after the Spurs defeat. Fourth in the table is a kinder place than the underlying numbers.

Recent nights that colour the tie

Chelsea arrive with the Spurs defeat still fresh. Tottenham had plenty of the ball at Stamford Bridge and won 2-0. That match does not define Chelsea’s season, but it sharpened the questions: a defence conceding better chances than last season, a set-piece record that is now a pattern, and a front line that can look short of a reference point. Cup football will ask whether the shape travels.

On our ratings Liverpool are 1772, third in the Premier League and 7th in the world.

Liverpool’s own recent league nights include the City defeat and the grind at Brentford. The numbers say those results sit inside a solid performance level, not a crisis. Andoni Iraola’s side still completes plenty of passes in advanced areas and still shows up in the box. The Carabao Cup is a chance to rotate without changing the identity. The Index cares less about who starts than about whether the chance quality holds when the midfield pairing changes.

Head-to-head history in cups is a thin guide. What matters for Wednesday is the meeting of two sides whose ratings already include the opponent adjustment. Liverpool still lead. Chelsea have had the harder league schedule and still trail on the Index while sitting high on the table. That is the tension under the tie.

Rotation will muddy the single-game signal, and that is fine. The Index is not tipping a cup scoreline. It is stopping a hard-schedule Premier League club being read as a crisis side because the table and the narrative have drifted, and stopping a lucky table position being read as a performance peak. Liverpool and Chelsea both need that correction before Wednesday is framed.

On the ball, Liverpool still look like a side that belongs in the top four of our rating. Chelsea still look like a side whose points have run ahead of their chances, even after Spurs. The opponent adjustment does not flip either verdict; it sharpens them. Chelsea’s harder schedule makes their luck note more interesting, not less. Liverpool’s mid-table schedule toughness means their rating is not being carried by soft fixtures.

For Iraola, the Carabao Cup is also a chance to blood the press structure without the full Premier League stake. For Chelsea, it is a chance to show the Spurs night was a bad evening rather than a level. Our numbers will not decide that in ninety minutes. They do say both clubs arrive better than a casual glance at recent results would suggest, and that Liverpool still hold the clearer edge on the Index.

What would change our mind? If Liverpool’s opponent adjustment turns negative over a softer run of fixtures, the “solid underneath” reading weakens. If Chelsea convert their luck surplus into a sustained chance surplus against tough opponents, the lucky label comes off. Wednesday is one more data point, not a verdict on either project.

Iraola’s side have already shown they can hurt big clubs when the press lands. Chelsea have the talent to match that if their chance creation catches their points. The Index puts both clubs above the mid-table noise without pretending either is settled.

Wednesday also sits inside a crowded calendar. Rotation will change the names; it should not change the season-to-date reading. Liverpool rate seventh worldwide with a modest positive opponent adjustment. Chelsea rate thirteenth with a similar adjustment and a much harder schedule rank. Those two facts are the preview.

Readers who want a score tip will not find one here. Readers who want to know whether either club is being misread by the table will: Liverpool are not a soft-schedule mirage, Chelsea are not a finished product, and the gap between them on our Index is real but not a rout.

Put simply: treat Liverpool as a top-ten Index side with a fair schedule, treat Chelsea as a top-fifteen Index side with a hard schedule and a luck surplus, and treat Wednesday as a cup night that can still go either way without rewriting either season story.

That is the whole point of bringing schedule and opponent adjustment into the preview. Without them, Chelsea look like a mid-table story and Liverpool look like a stumble. With them, both clubs look like Index sides with unfinished business — and Anfield on Wednesday looks like a proper test rather than a brand mismatch.

The verdict

Liverpool are better. Chelsea have had the harder league run. Anfield on Wednesday is not a free hit for either coach. Liverpool still lead the Index between the two. Chelsea’s luck note remains. Neither side is where a glance at the table alone would put them.

Wednesday will not settle either club’s season. It will add one more data point. If Liverpool’s chance quality holds with rotation, the Index will keep treating recent league results as noise around a solid level. If Chelsea’s chance quality stays flat while the table looks healthier, the luck note remains. The Carabao Cup is a fair stage for that test.

For readers who only want one sentence: Liverpool are better than Chelsea on our Index; Chelsea have faced the tougher Premier League schedule so far; both get a small lift once opponents are counted. Everything else in this preview is scaffolding for that sentence.

The short answerMarcus Linday was one of the Eredivisie's best two-way midfielders in 2025-26, in the 95th percentile for duels won and the 88th for progressive passes. Ranked raw against Big Five midfielders, his ball-winning would still be near the top 10%. Expect a dip after the step up. Bournemouth, Freiburg and Stuttgart fit.

This summer SC Braga made four or five offers for Marcus Linday, depending on which report you read, rising to about €6.5m. Heerenveen said no every time. The 23-year-old Swede told the club he wanted to go. He stayed, and his contract runs to June 2029. The market's view of him is clear from that saga: a good Eredivisie midfielder, worth a mid-sized Portuguese club's money. Our numbers suggest the market has him one level too low.

This is a step-up report. The question isn't whether Linday is good in the Netherlands, because he plainly is. It's whether there's evidence he can do the job at a club in one of Europe's top five leagues, what that jump would cost him, and which clubs would suit him.

The jump in his first full season

Linday arrived from Västerås in Sweden in January 2025, and his first half-season was steady rather than special: 1,446 league minutes, roughly mid-table among Eredivisie central midfielders on most measures, with weaknesses at intercepting (the 17th percentile) and in the air (the 9th).

Then 2025-26 happened. Over 2,493 minutes in 32 league games he moved into the top fifth of the league's midfielders for almost everything a modern box-to-box or holding midfielder needs. He was in the 95th percentile for duels won, the 86th for defensive duels won, the 82nd for successful defensive actions, the 88th for progressive passes and the 91st for progressive runs. Even the weaknesses improved: interceptions rose to the 72nd percentile and aerial duels to the 75th.

That combination is the point. Plenty of midfielders win the ball and give it simply to a centre-back. Plenty of others move it forward but leave a hole behind them. Linday did both in the same season. He made 9.64 successful defensive actions per 90, won 57% of all his duels and added 7.22 progressive passes per 90.

The step-up test

Being good in the Eredivisie is not the same as being good in the Premier League or the Bundesliga. On our ratings, the average Eredivisie club is about 206 points weaker than the average club in the top five leagues. That's a big gap, worth more than a goal a game. So we ran a simple, deliberately harsh test. We took Linday's raw 2025-26 per-90 numbers and placed them in the pool of every central midfielder with 900 minutes in the Premier League, LaLiga, Serie A, the Bundesliga and Ligue 1 last season, 347 players.

His raw numbers would still rank in the 91st percentile for defensive actions, the 88th for interceptions, the 93rd for duels won, the 90th for progressive runs and the 83rd for progressive passes.

Plenty of midfielders win the ball and give it simply to a centre-back.

Be careful how you read that. Those raw numbers would not survive the move unchanged. Stronger opponents keep the ball better, win more duels and give a midfielder less time to pick a pass. Expect his percentiles to drop, perhaps sharply, in his first season. But the starting point matters. A midfielder whose raw ball-winning sits around the top 10% of the Big Five pool has room to fall and still be a useful starter. One sitting at the median does not.

What role he would play

Profiles like Linday's are easy to misuse. Put him in a number 10 role and you lose what makes him valuable, because he doesn't create enough chances. Ask him to sit alone in front of the defence as a pure screen and you waste his carrying. His numbers point to a specific job: one of two central midfielders in a 4-2-3-1 or 4-4-2, or the more mobile of the three in a 4-3-3. He would press the ball, win it back and carry it 20 or 30 yards before releasing a better passer.

That's also why the type of club matters more than the league. A possession-heavy side that rarely loses the ball would underuse his defending. A deep, low-block side would underuse his running. The sweet spot is a team that presses high and plays quickly once it wins the ball, where a midfielder who wins duels and immediately moves the ball forward turns turnovers into attacks.

His size helps. At 186cm he has the frame for Premier League and Bundesliga physicality, and the jump in his aerial duels from the 9th percentile to the 75th suggests he has learned to use it. The question for a scout watching him over the next two months is simple: are the duels and the carrying back to last season's level after the knee injury? If they are, the case is strong.

The caveats

Three things should worry a buyer. First, he isn't a creator. His expected assists were only in the 58th percentile in the Eredivisie, so he'd need a playmaker beside him. Second, the record is short: one excellent full season, half a decent one and a small sample this season. Third, fitness. He hurt his knee in training in late August, missed three league games and only returned to Heerenveen's starting line-up in October, so our 2026-27 sample is just 367 minutes over four games. In that small sample his ball-winning and pass completion look like last season. His aerial duels and the accuracy of his forward passes don't, yet.

Three realistic destinations

These are our suggestions based on fit and level, not clubs reported to be interested.

Bournemouth (1695, 9th in the Premier League on our ratings). On our style measures, they are one of the most active defensive teams in the league, with very high possession-adjusted interceptions. A midfielder who wins duels and carries the ball forward is exactly what that system needs. They have a record of buying from smaller leagues and selling on.

Freiburg (1599, 5th in the Bundesliga). It's a classic step-up club: well coached, patient with players from smaller leagues, and a Bundesliga intensity that would test his duels without throwing him into the Premier League straight away.

Stuttgart (1561, 6th in the Bundesliga). A high-tempo side where two-way midfielders get minutes. It would be a gentler first step than the Premier League. For context, Braga are rated 1481, so all three of our suggestions sit well above the club that bid this summer.

Our verdict

Step-up readiness: real, with a sample-size warning. The market has priced Linday as a €6–7m player for a Portuguese club. His 2025-26 numbers, even placed raw in a Big Five pool, look like a Big Five squad midfielder with starting upside. He combines ball-winning and progression, which is rare at his price. The honest risk is that the evidence rests on one season. If he gets back to last season's level over the next two months, his asking price will rise quickly. A Big Five club that moves in January, or plans for the summer as Heerenveen's directors have suggested, would be buying before the market catches up.

The short answerPSG beat Lyon 2-0 with second-half goals from substitutes Senny Mayulu (72') and Mika Godts (84'). Lyon defended deep and stayed level for an hour. Our ratings had PSG about a goal a game better beforehand; the result matched the hierarchy more than the old points level.

Paris Saint-Germain beat Lyon 2-0 at the Parc des Princes on Monday night and went top of Ligue 1. Senny Mayulu scored in the 72nd minute and Mika Godts in the 84th, both after coming off the bench. For an hour the game had the shape our preview expected: PSG on the front foot, Lyon defending deep, and the gap on the pitch narrower than the gap on the rating until the substitutes arrived.

This is the morning-after. The Index question is not whether PSG deserved to win. They did. It is what the night said about two sides that had been level on points, and about a PSG squad that is carrying injuries and still finding different routes to the same result.

Our preview had them about a goal a game apart on rating while level on points. The night played out in that spirit: Lyon competitive, PSG persistent, and the hierarchy asserting itself when Enrique could change the picture without changing the system. A morning-after piece should be honest about what one match can and cannot prove. It can prove that PSG’s depth is currently a weapon. It cannot prove that Lyon’s rise is over.

How the night unfolded

The first hour was tight. PSG had more of the ball and more of the territory. Dominik Greif, Lyon’s goalkeeper, kept the score at 0-0 with saves from Ferran Torres and others. Lyon’s best moments were on the break: Matvey Safonov had to deny Julien Duranville and Zachary Athekame around the half-hour. The scoreboard was still 0-0 deep into the second half.

Luis Enrique changed it. João Neves made way for Godts on 57 minutes; Mayulu replaced Ferran on 65. Seven minutes later Mayulu struck from the edge of the box, a low right-footed finish after a combination with Godts, to make it 1-0. On 84 minutes Godts cut inside from the left and finished for 2-0, his first Ligue 1 goal in a PSG shirt, after a high regain involving Lucas Beraldo and a ball from Quentin Ndjantou. The Parc had its win. Lyon had competed for an hour and faded when PSG’s bench entered the game.

What our numbers said beforehand

Before kick-off, PSG were rated 1833, 5th in the world and first in Ligue 1. Lyon were 1600, 44th and fourth. The gap is about 1.3 goals a game. PSG’s opponent adjustment this week is +39 rating points — the toughest schedule in Ligue 1 — while Lyon’s is −27 against one of the easiest. Both have taken about the points their football deserves. Monday did not invent a hierarchy. It confirmed one that the table had been hiding by keeping them level on points.

The bench as a tactical fact

The interesting part for anyone watching PSG week to week is not that they beat Lyon. It is who scored. With Khvicha Kvaratskhelia out long-term and other absences in the squad, Enrique is rotating a deep attacking group. Mayulu and Godts are not emergency options; on Monday they were the difference between a draw and a statement. That is how a rating this high survives an injury list: the second wave is good enough to break a deep block that the starters could not.

João Neves made way for Godts on 57 minutes; Mayulu replaced Ferran on 65.

Lyon’s plan was coherent. They sat in, trusted Greif, and looked for Duranville and Athekame on the break. It worked for 70 minutes. It asked their legs to hold against wave after wave, and once PSG’s fresh attackers arrived the block stretched. Fonseca had already rotated for Europa League duty before this fixture; the cost showed late.

Where that leaves both

PSG go top. Their trend rating is still falling, which is a warning light rather than a verdict, and the injury list is real. But a fifth straight league win, an eighth in all competitions, and a clean sheet against the side that had matched them on points is the behaviour of a champion squad, not a fragile one. Lyon stay in the conversation on rating and drop a chance to keep pace on the table. Their own trend is rising. One away night at the Parc against this PSG does not undo that.

What the night confirms

Club and league reports agree on the shape: 0-0 past the hour, Lyon in a deep block, Greif busy, Safonov’s saves on the break, then 1-0 at 72 and 2-0 at 84. The late game is PSG territory, Lyon stretched, and the Parc celebrating a first spell on top of the table.

None of that requires inventing a tactical essay the pictures do not support. PSG dominated territory without an early breakthrough. Lyon stayed in the game through organisation and a goalkeeper on a good night. The substitutes changed the energy and the finishing. That is the whole shape of the match, and it is enough.

The preview, revisited

Before the game we noted that PSG and Lyon had been level on points while sitting about a goal a game apart on rating, and that PSG’s within-league dominance is enormous once you separate it from Ligue 1’s weaker baseline. Monday did not rewrite that essay. It illustrated it. A deep block can suppress a favourite for an hour. It rarely suppresses a favourite a goal a game better for ninety, especially when that favourite can change the attack without changing the structure.

The verdict

PSG’s hierarchy, delivered from the bench. The game was even on the scoreboard until the 72nd minute and never even on the underlying season numbers. Mayulu and Godts finished what the starters had asked Greif to survive. For PSG the night answers the depth question in the short term. For Lyon it confirms the gap we had them facing before kick-off: competitive for an hour, short for ninety against a side a goal a game better on the rating.

Ligue 1 will keep producing nights like this for PSG: deep blocks, a goalkeeper in form, and a late breakthrough that looks inevitable only in retrospect. Monday’s 2-0 does not change the framework that separates PSG’s within-league dominance from the league’s European baseline. It fills that framework with a concrete example of how Enrique’s bench currently turns a stubborn night into a statement win.

What would change our mind? If PSG start needing late magic every week against mid-table sides, the falling trend becomes the story. If Lyon create the better chances in the return meeting, Monday was a game state, not a level. Until then, the Index reads this as a strong favourite eventually telling the truth on the scoreboard.

Method note for the morning-after: the rating and luck figures above are season samples through Sunday morning 25 October. The match narrative is anchored to the confirmed scorers and times and the publicly reported shape of the night. That is enough to say PSG’s hierarchy held. The Index reading does not need a full shot map to stand: season form and the result tell the same story.

The short answerArsenal are No. 1 again, ahead of Bayern and Barcelona, after we began judging every game against the opponent. PSG, Aston Villa and Bournemouth rose most; Lille, Porto and Lyon fell most. Lazio are the luckiest side in Europe's top five leagues, and they have also had one of the easiest schedules.

Barcelona have dropped from first to third in our world ranking this week, and the Clásico has nothing to do with it. Sunday's 2-1 defeat by Real Madrid is not in these numbers yet. What has changed is how we judge the games that are.

From this week, every performance in the ranking is weighed against the opponent it came against. Creating five good chances against the bottom side is no longer worth the same as creating five against the champions. That one change reshuffled the top of the table: Arsenal are back at No. 1, PSG made the biggest jump among the elite, and some of the clubs that looked lucky or unlucky now look a bit different.

So this week's question is simple. Once you account for who everybody has played, who is better than their league position, who is worse, and who has simply had the kinder fixture list?

A note on the numbers

Our ratings use results to the morning of Sunday 25 October 2026. That means Saturday's games are in, but Sunday's afternoon and evening games in Europe, including the Clásico and Manchester United's 3-0 home defeat by Bournemouth, will be added next week. Ratings sit on one world scale, where 1500 is an average club from Europe's top five leagues and about 180 points is worth a goal a game.

Two new figures appear in this piece and on every club page. Schedule strength is the average rating of the opponents a club has faced this season. Opponent adjustment is how far a club's rating moved once those opponents were taken into account. Our luck figure is unchanged: it compares the points a club has won with the points its chances in those same games would normally earn. We call that second number the points the football deserved.

The top of the table

Arsenal did not do anything new to go top. Their rating is 1910, exactly what it was a week ago. Bayern München moved up one place to second on 1900. Barcelona fell 16 points to 1897, which means the top three are now separated by 13 rating points, less than a tenth of a goal a game.

Barcelona's fall is a schedule story. They have 27 points from nine league games, a perfect record before Sunday, but their opponents rank 17th of 20 in LaLiga for toughness. Once each game is judged against the side they were playing, their rating fell by about 12 points, which means a perfect start against modest opposition counts for slightly less than it did. They are still the highest-rated side in Spain, with the best deserved points rate in LaLiga. The point is narrower than it sounds: the margin between them and the clubs just below was always thinner than nine straight wins suggested.

Two new figures appear in this piece and on every club page.

The rest of the top ten: Manchester City are fourth (1850), PSG fifth (1833), Real Madrid sixth (1817), Liverpool seventh (1772), Benfica eighth (1760), Brighton ninth (1751) and Inter tenth (1749).

The risers

Among clubs in the world top 50, three gained more than 25 points.

PSG (+37, 6th to 5th). No club in Ligue 1 has faced a tougher run of opponents than PSG. Judged against that schedule, their performances are worth about 39 rating points more than they looked, which means a method that treated every opponent alike was slightly underrating them. They are now a clear fifth in the world.

Aston Villa (+27, 14th to 11th). This is the most interesting rise of the week, because it comes with a warning attached. Villa have had the 5th-toughest schedule in the Premier League, so the adjustment helped them. But on the points their football deserved, they are bottom of the Premier League: 6.4 from eight games. They have 11. That gap of 4.6 points makes them one of the luckiest sides in the league. Villa are rated as the 5th-best team in England because earlier evidence, mostly last season, still carries most of the weight this early in a campaign. Hard fixtures explain some of their poor numbers. They do not explain all of them.

Bournemouth (+26, 19th to 16th). Bournemouth had 6 points from seven games before Sunday, against football worth 10.9. Their opponents were slightly easier than the league average, yet the adjustment still lifted them. The adjustment works game by game, so this is about how well they played against each opponent, not just who the opponents were. Then they won 3-0 at Old Trafford, a result that is not in these numbers yet.

Just behind them, Monaco rose +25 to 33rd and Benfica +21 to 8th. Further down the table, Olympiacos jumped from 99th to 52nd and Braga from 105th to 63rd, after this week's update revised their league numbers upwards.

The fallers

Lille (−38, 50th to 61st). Lille have 14 points from seven games, 5.5 more than their football deserved, and they are only 11th in Ligue 1 on deserved points. The adjustment took a further 36 points off their rating, the biggest cut for any club in the top 50. Their schedule has not been easy, so this is not about weak opponents. Judged game by game against the sides they were playing, their performances were worth less than the raw numbers said.

Porto (−36, 18th to 28th). Porto have 27 points from nine league games, and nobody in Portugal has played a tougher schedule. Yet their rating fell, by about 28 points from the adjustment alone. The adjustment looks at how well Porto played against each opponent, not just who the opponents were, and on that measure the performances were less special than a perfect record. It is a reminder that a hard fixture list and a strong performance are two different things.

Lyon (−28, 37th to 44th). Lyon have had the second-easiest schedule in Ligue 1, and their rating dropped by 28 points once that was counted. Al Nassr fell by the same amount, to 53rd.

Two Premier League sides also slipped. Brighton dropped from 7th to 9th (−27), and Manchester United from 10th to 14th (−22). Both deserve a closer look.

Manchester United and Brighton: what the schedule says

The easy explanation for United's rating is that it is propped up by an easy run of games, or by last season. Neither fits the numbers. United's opponents have averaged 1581, the 7th-toughest schedule in the Premier League. On the points their football deserved, they are 6th in the league, and they are rated 7th. The rating is roughly where this season's football puts them.

The adjustment still cost them about 20 points. Judged game by game against those opponents, United's performances were worth less than the raw totals suggested. Before Sunday they had 7 points from seven games against football worth 11.7, so they were 4.7 points short, and our model still called them unlucky. Sunday's 3-0 home defeat by Bournemouth is not in these numbers. Next week will tell us whether it was more bad luck, or the first sign that the performances are dropping towards the results.

Brighton are the opposite case. Their opponents have averaged 1485, the weakest schedule in the Premier League by a distance; the league average is 1572. They are 9th in the world and 3.1 points ahead of their football. None of that makes Brighton a bad side. It does mean their next run of games is the real test.

The Luck Table

The Luck Table compares the points each club has won with the points its football deserved. This week we have added a second question: what kind of opponents were those points won against? A club that is lucky against easy opponents is flattered twice. A club that is unlucky against hard opponents is underrated twice.

Flattered twice: Lazio and Manchester City. Lazio have 22 points from eight games and football worth 12.8. That gap of 9.2 is the biggest in Europe's top five leagues. Their opponents have averaged 1413, the second-easiest schedule in Serie A (the league average is 1448), and the adjustment cut their rating by about 21 points. They are only 9th in Serie A on deserved points. Manchester City are 5.7 points ahead (21 from eight against 15.4 deserved), with the third-easiest schedule in the Premier League. City are still the second-best side in England on our numbers. Their points total just runs ahead of it.

Lucky, but against good sides: Arsenal, Atlético Madrid and Aston Villa. Arsenal are 5.3 points ahead, but they also have the best deserved points rate in the Premier League and the 6th-toughest schedule. Atlético are 6.7 ahead, second only to Lazio, with the 6th-toughest schedule in LaLiga. Villa, as above, are lucky on points and short on chances, but their fixtures have been hard. For all three, some of the points cushion is real.

Underrated twice: Nottingham Forest and Athletic Club. Forest have 6 points from eight games, against football worth 14.7. That 8.7-point shortfall is the biggest in the top five leagues, and it has come against the 4th-toughest schedule in England (opponents averaging 1602). Forest are 4th in the Premier League on deserved points. Athletic Club are next: 9 points against 16.2 deserved, with the 2nd-toughest schedule in LaLiga. Both tables are wrong about them, and both have earned an easier run.

Unlucky, without the excuse: Juventus. Juventus have 11 points from seven games against 16.9 deserved, a gap of 5.9. Their schedule has been slightly easier than average, yet their football has been the second-best in Serie A on deserved points. That is the cleanest case of points and performance pulling apart, and it usually corrects itself as a season goes on.

Across the top five leagues, our model now labels 23 clubs lucky and 21 unlucky, out of 96. The rest are getting roughly what they deserve.

What to watch next

Next Tuesday's update adds last Sunday's games, including the Clásico and United's defeat by Bournemouth, along with this week's fixtures. The first test of this week's moves is whether PSG keep their gains and whether Lazio and City start to give some points back. Forest and Athletic will look for the points their football has already earned.

Every club's rating, schedule strength and luck figure is on its club page, and the full table is in the world club ranking. How the ranking works explains the method in plain words.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerReal Madrid beat Barcelona 2-1 at the Camp Nou because they refused to sit deep. They pressed in moments, broke quickly and defended their box in numbers, with Jude Bellingham driving that work off the ball. Kylian Mbappé scored twice, including an 87th-minute winner. Barcelona now lead Atlético by two points and Madrid by three.

Real Madrid won 2-1 at the Camp Nou on Sunday night, and Kylian Mbappé will get the headlines. He scored both goals, the first a ricochet off his knee four minutes before half-time and the second an 87th-minute tap-in that was his 100th goal for the club. But the result was built long before either finish. José Mourinho's side came to Barcelona to play, not to survive. They pressed in moments, ran at every chance, defended their box in numbers, and in Jude Bellingham they had the player who did more than anyone to make that plan hold together.

Barcelona had won all nine of their league games before Sunday. They had 62% of the ball, 18 shots to Madrid's ten, and a stadium that expected a nine-point lead by the end of the night. They leave with a two-point lead over Atlético Madrid and a three-point lead over Real. A title race that looked close to settled in October is open again.

Toe-to-toe, not backs to the wall

The usual script for a team visiting the league leaders at the Camp Nou is a low block and a long wait for a counter. Madrid did not follow it. The first shot of the night was theirs, after a loose Pedri pass let Arda Güler break inside four minutes. Through the first half, Mourinho's side took turns with Barcelona: Güler forced Joan García into a save, Anthony Gordon tested Andriy Lunin, Raphinha shot over and then curled wide after a neat move with Gordon. It was end-to-end, and Madrid were comfortable with that.

Mourinho said afterwards that his players had shown "a lot of courage to play out from the back against an opponent that presses hard". That was the first half in one line. Madrid did not simply clear their lines. They kept the ball for short spells, enough to stop Barcelona building rhythm, and then they went quickly when space opened. Hansi Flick said much the same thing from the other side: "Against the ball they gave us not so much space."

The opening goal was lucky, but it was earned. Marc Cucurella released Yan Diomandé on the right, Diomandé drove a low cross across the six-yard box, and Gerard Martín's attempted clearance hit Mbappé and went in. The bounce was fortunate. The move that forced Martín to clear was not. Madrid had worked the ball through Barcelona's press and into the space behind their left side, which is exactly where they wanted to go.

Bellingham's night was won without the ball

Bellingham did not score and did not register the assist for either goal. On the ball, his night was mixed. Early in the second half he broke on a two-against-one with Mbappé to his side and could not find the pass, and Martín shut the chance down. Later he had another opening to release Mbappé and did not take it either.

His contribution was elsewhere. In the 33rd minute it was a big Bellingham tackle in midfield that started the move ending with Güler's cross and Mbappé's overhead kick over the bar. When Barcelona came back at Madrid after half-time, the tackles around the Madrid box kept coming, with Bellingham and Aurélien Tchouaméni in the middle of them. Madrid defended their box with most of the side behind the ball, and Bellingham was part of that block rather than a spectator to it, giving Madrid a midfield that was willing to defend for each other. Against a side that had been scoring heavily all season, Madrid restricted Barcelona to one goal.

Madrid had worked the ball through Barcelona's press and into the space behind their left side, which is exactly where they wanted to go.

Then, with three minutes left, came the moment on the ball that mattered. Under pressure in a tight space, Bellingham turned his way out and released Eduardo Camavinga, who sprinted clear, slipped past Pau Cubarsí and pulled the ball back for Mbappé to finish. One of Bellingham's few clean actions in possession all night decided the Clásico. But Madrid were only in a position to win it because of everything he had done without the ball before that.

How Barcelona got back in, and why it was not enough

Barcelona were better after the break. João Cancelo and Alejandro Balde pushed on together, Barcelona committed more players forward, and the game tilted. On the hour Lamine Yamal took a corner short to Pedri, who slipped a clever reverse pass into Raphinha, and his low shot beat Lunin. Just before the equaliser, Cancelo had caught Diomandé on the calf with his studs and stayed on the pitch, which Mourinho called a clear red card.

For 25 minutes Barcelona pushed. Lunin made a superb one-handed save from Martín's header from a corner. Yamal skipped past two challenges and clipped a shot just wide, then saw another effort blocked by Dean Huijsen's face. It was not Yamal's night. He had a glaring chance to equalise before half-time and scuffed it wide at the near post, and Cucurella was in his face from the first minute.

The trouble with pushing was what it left behind. The more Barcelona committed, the more space Madrid had to run into, and Madrid never stopped looking for it. Madrid created four big chances to Barcelona's two on the night. The winner was the clearest of those: one turn, one run, one pass and the Camp Nou fell silent.

What the Index said before kick-off

Our numbers had warned that this match was closer than the table suggested. Before Sunday, Barcelona had 27 points from nine games, but their football had been worth 23.7 points, 3.3 fewer. They were still the best team in Spain by performance, and the best-rated club in the world, but they had also been riding some good fortune. Last season they finished 14.9 points above what their football deserved.

Madrid were the opposite. They had 21 points from nine games, and their football had been worth 21.2. That is a team getting what it deserves, no more and no less. The only side in LaLiga producing better football than Madrid was Barcelona. A gap of six points between them was always larger than the gap in the quality of their play, and Sunday closed half of it.

Our ratings tell the same story. Barcelona are rated 1913, first in our world ranking of 3,834 clubs. Madrid are rated 1808, 5th in the world. Around 180 points is worth a goal a game, so the gap is a little over half a goal. That is a real difference, and over a full season it matters. Over one night at the Camp Nou, it can be wiped out by a counter-attack and a cutback.

The club profiles in the Index add one more layer. Madrid's manager score is the highest in LaLiga. That reflects Mourinho's record over time, not one result, but Sunday was the kind of night that record is built on: a plan that fitted his players and was carried out for 87 minutes. Barcelona's profile shows a squad that ranks among the best in the world for its first XI but sits near the bottom of the league for lifecycle, which measures where a squad's key players are in their careers. Flick said his side "missed the freshness". A heavy week that included a Champions League defeat at Paris Saint-Germain will not have helped.

What it means for the title race

The table now reads Barcelona 27, Atlético 25, Real Madrid 24, all after ten games. Do not read too much into Atlético's position just yet. Before Sunday they had 6.7 points more than their football deserved, the biggest surplus in LaLiga. Their results are real, but their performances suggest they have been fortunate to collect so many.

For Barcelona, the lesson is not that they have suddenly become a worse team. Our ratings still have them as the best side in the world, and one defeat moves a rating only a little. The lesson is that their perfect start had been flattering them slightly, and that Madrid are a genuine rival rather than a side waiting to fall away. Two defeats in a week, in Paris and at home to Madrid, will test them, and the answer will come in the matches that do not get this kind of attention.

For Madrid, Sunday is proof of something the numbers had been saying quietly all season. Their football has been worth almost exactly what they have earned, and that football is the second best in Spain. With Mbappé finishing and Bellingham doing the work that does not show up on the scoresheet, they went to the home of the league leaders and beat them at their own game. The gap is three points now, and Madrid look like a team ready to close it.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerNot on the season so far. Before Sunday, United's football was worth 1.67 points a game, the sixth-best rate in the Premier League, and they were 4.7 points short of what it deserved. But the 3-0 home defeat to Bournemouth was a poor performance, not bad luck, and the Fernandes row is a risk no model measures.

Manchester United lost 3-0 at home to Bournemouth on Sunday, and it could have been worse. Michael Carrick dropped his captain, Bruno Fernandes, and never brought him on. Rayan, Marcus Tavernier and Evanilson scored, United went eight games without a win in all competitions, and Old Trafford booed its own players off. United are 16th, a point above the relegation zone, with seven points from eight league games.

So here is the question that matters for Carrick and for whoever makes the decision about him. Is this really a 16th-placed team? Or, after a fortnight like this, is it something worse? The Index has a clear answer on the first eight games, and a warning about the ninth.

How Bournemouth took Old Trafford apart

United had the ball: 63% of it, 513 accurate passes to 264 and 17 shots to ten. None of that told the story. Bournemouth were the more dangerous side from early on, and when they scored it was because United handed them the chance.

The opener came in the 38th minute. Ayden Heaven made a wild challenge on Evanilson, was booked, and Justin Kluivert's free-kick from the right was allowed to drop to Rayan, unmarked at the far post, who volleyed it in. Five minutes later Leny Yoro slipped near his own area, Adrien Truffert took the ball and squared it, and Tavernier slid it into an empty net. Carrick defended both players afterwards, calling the first a dubious free-kick and the second "just a slip".

The half-time change was Matheus Cunha for Youri Tielemans, not Fernandes. It did not help. Evanilson had a goal ruled out for offside within seconds of the restart after a loose Harry Maguire pass, and Rayan wasted another good chance. Lisandro Martínez came on for Heaven, and soon afterwards a long ball from Lewis Cook sent Evanilson racing towards goal. Senne Lammens came out, Evanilson got there first, went round him and curled a superb finish in from a tight angle on the right. That was the game.

The detail that should worry United most is the type of chance each side created. Many of United's 17 shots came from outside the box. Bournemouth, with ten, scored three and missed three more big chances. Possession without penetration is a pattern that the numbers can hide for a while. On Sunday it was obvious.

Bournemouth deserve their share of the credit. They arrived with three straight wins in all competitions, including a 3-0 win over AC Milan on Thursday, and left with four. Before Sunday they were in the same position as United, 4.9 points short of what their football deserved. Now, with nine points from eight games, their results are starting to catch up.

The half-time change was Matheus Cunha for Youri Tielemans, not Fernandes.

What the numbers said before Sunday

This is where the Index pushes back against the table. Before Sunday, United had seven points from seven league games. Our expected-points measure, which looks at the quality of the chances a team creates and concedes and asks what that would normally earn, valued their football at 11.7 points. That is 1.67 points a game, the 6th best rate in the Premier League.

Put simply, United were 4.7 points short of what their football had earned. Only Nottingham Forest and Sunderland were further behind their performances, and Bournemouth were just ahead of United on the same list. That is why our preview called this a meeting of two teams the table had got wrong. On Sunday one of them finally got what it deserved, and it was not United.

The ratings say the same thing. United are rated 1750, the 5th-highest in the Premier League and 10th in our world ranking. Their style profile shows a team that takes a lot of shots, one of the highest volumes in the league. That is a side that has been creating plenty, not one being overrun every week. On the evidence of the first seven games, United's underlying level was closer to the top six than the bottom four.

So is it worse than 16th?

Not on the evidence of the season so far. But Sunday was a bad performance, not just a bad result, and that matters. Until now, United's problem had been finishing and fine margins: good chances missed, soft goals conceded. Against Bournemouth the chances themselves were poor, and the defending was worse. One match will not move a rating far, because ratings are built to be patient. If performances like Sunday's become the norm, the numbers will follow the table down. If they do not, the table is likely to move back towards the numbers.

History is useful here. Two seasons ago United took 42 points from football our model valued at 52.2, ten points short. Last season they swung the other way, collecting 71 points from football worth 64.4. Results have moved around far more than the quality of the football. This season, for the first seven games, they were back on the wrong side of the line.

The club profile in the Index shows where the real weak points are. United's first XI ranks highly in the league, but their star spread, which measures how evenly quality is shared across the squad, is near the bottom, and their manager score sits in the bottom half. A team that leans heavily on a few players is exposed when one of them, in this case the captain, is suddenly out of the side.

The Fernandes question

That is the part of Sunday the numbers cannot measure. The build-up to the game was dominated by Fernandes's public criticism of the team's tactics after last Wednesday's 4-1 defeat at Como in the Champions League, and Carrick's decision to leave him out. Carrick said the call was about changing "the course of play". He also said: "What's next? I really don't know."

A dressing room in that state can turn a team that deserves more into one that gets less. On Sunday, for the first time this season, United played like the side the table describes. Whether it was a one-off or a turning point will depend on what happens off the pitch this week as much as on it.

What it means for Carrick

We have looked at this before. Across 6,876 head-coach spells, teams tend to peak long after the point when most managers are sacked, and the clubs that judge managers on performances rather than results usually make better decisions. Before Sunday, United's performances were running well ahead of their results, which is the strongest argument for patience.

Sunday weakens that argument, but does not destroy it. The case against Carrick now rests less on the underlying numbers and more on the dressing room, the captain and the mood in the stands. The case for him still rests on seven league games in which his team played like one of the better sides in the country and did not get the points. If United's football keeps looking like the first seven games, their position will improve. If it looks like Sunday, the table will turn out to have been right.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerThe weekend's biggest stories were Real Madrid's win at Barcelona, United's 3-0 home defeat by Bournemouth and Inter's late comeback. On our numbers, Bayern and Inter are stronger than the table shows, while Lazio, Palmeiras and Nashville have collected well over the points their football has deserved.

Real Madrid won the Clásico, Manchester United lost again, Inter came from behind again, Bayern and Rennes went top, and on the other side of the Atlantic Palmeiras and Nashville kept collecting points their football has not quite earned. Here are the seven stories from the weekend that matter most, each one set against what our ratings and expected-points numbers said going in.

This is the first edition of The Monday Index, our new weekly column. Every week we pick the five to seven stories that tell you most about where the season is heading, in Europe and beyond, and ask the same question of each: was the result a surprise, or had the numbers seen it coming?

A note on the numbers

Our figures run to Sunday morning. That means they include Saturday's games, including Lazio's, Dortmund's, Palmeiras's and Nashville's wins, but not Sunday's. When we say "before Sunday", that is what we mean. Expected points estimate what a team's chances created and conceded would normally earn. A side with many more points than expected points has been finishing better, or getting more breaks, than its football alone would suggest. That does not mean it will stop. It means the gap is worth watching.

1. Madrid went to the Camp Nou and won the argument

Real Madrid beat Barcelona 2-1. Kylian Mbappé scored in the 41st minute, Raphinha equalised after the hour, and Mbappé won it in the 87th after Jude Bellingham worked his way out of a tight spot and released Eduardo Camavinga. Barcelona still lead LaLiga, but by two points from Atlético Madrid and three from Madrid.

Before Sunday, Barcelona had 27 points from nine games, about 3.3 more than their football deserved. Madrid had 21, almost exactly what theirs was worth. The two clubs were first and second in our LaLiga ratings and Barcelona were rated 1913 to Madrid's 1808. A Madrid win at the Camp Nou was not a shock on those numbers. What made it notable was how they did it. We broke it down in our tactical analysis of the Clásico.

2. United are no longer just unlucky

Manchester United lost 3-0 at home to Bournemouth. Michael Carrick left Bruno Fernandes out after the captain's public criticism of the team, and United are 16th with seven points from eight games, eight matches without a win in all competitions.

Before Sunday, United's football was worth 1.67 points a game, the sixth-best rate in the Premier League, and they were 4.7 points short of what it deserved. That has been the main case for patience with Carrick. Sunday weakened it, because this was a poor performance rather than another unlucky one. Our full piece, are United really a 16th-placed side?, looks at whether the underlying level is about to catch up with the table, or the other way round.

Before Sunday, Barcelona had 27 points from nine games, about 3.3 more than their football deserved.

3. Inter keep leaving it late, and Lazio keep getting away with it

Inter were heading for defeat at San Siro when Alex Jiménez put Fiorentina ahead in the 65th minute. Marcus Thuram equalised in the 91st, then won the penalty that Piotr Zieliński scored in the 98th. Inter are level with Lazio on 22 points at the top of Serie A. Fiorentina are 18th.

These are two very different leaders. Inter are the highest-rated side in Serie A and 9th in our world ranking, with the third-best deserved points rate in the league. Lazio, including Saturday's win, had 22 points from football worth 12.8. That gap, 9.2 points, is the biggest in Serie A, and on deserved points per game Lazio were only 9th. Fiorentina, meanwhile, were 4.9 points short of what their football deserved before Sunday, and late goals like these are how a side ends up in 18th. If the title race comes down to the two sides level on 22 points, the numbers back Inter clearly.

4. Bayern were the best team in Germany before they went back top

Bayern won 3-1 at Freiburg to go back above Borussia Dortmund on goal difference. Dayot Upamecano and Serge Gnabry scored before half-time, Tom Bischof's own goal gave Freiburg hope, and Michael Olise sealed it deep in stoppage time. Vincent Kompany was not there. He was taken to hospital after collapsing on Friday, and his staff took charge.

The table had flattered Dortmund, not Bayern. Before Sunday, Bayern had 14 points from six games but football worth 2.74 points a game, the best rate in the Bundesliga, and they were 2.4 points short. Dortmund, after Saturday's win, had 17 points from seven, 2.8 more than their 2.02 points a game deserved. Bayern are rated 1899, 3rd in the world. The gap between the two clubs on underlying performance is much bigger than goal difference suggests.

5. Rennes are top, and the sides behind them have been lucky

Rennes beat Strasbourg 3-0 to go top of Ligue 1, a point ahead of Lille and Monaco, who drew 1-1 with each other. Boulaye Dia scored first and Sebastian Szymański added two. Paris Saint-Germain and Lyon, two points behind Rennes, meet on Monday.

Rennes are not a statistical mirage, but they are not a title favourite on our numbers either. Before Sunday they were 7th in our Ligue 1 ratings and 7th on deserved points per game, and 2.5 points ahead of their football. The more interesting story is just behind them. Lille were 5.5 points ahead of their football, the most in the league, and only 11th on deserved points. Monaco were second on the same luck list. PSG remain the highest-rated side in France and 6th in the world. A crowded top of the table is the result of a lot of good fortune, and Monday's game in Paris could start to sort it out.

6. Palmeiras lead Brazil, but Flamengo have been better

Palmeiras beat Red Bull Bragantino 2-0 on Saturday, with goals from Mauricio and Giay, and went top of the Brasileirão on 69 points, two clear of Flamengo. Flamengo have a game in hand, at home to Atlético Mineiro on Monday night, and a win would put them back on top.

Both clubs are well ahead of their football, by a long way. Including Saturday, Palmeiras had 69 points from football worth 49.4, 19.6 points more than they deserved and the biggest gap in the league. Flamengo were second on that list, 13.8 points ahead. But Flamengo have the best deserved points rate in Brazil and the highest rating, 1566. Palmeiras are 3rd on deserved points. With six rounds to go, the club at the top of our numbers is not the club at the top of the table. To put Brazilian football on one scale: Palmeiras are rated 1524, about the level of an average club from Europe's top five leagues.

7. Nashville's Shield was won by finishing, and Vancouver know it

Nashville SC won 2-0 in Montréal on Saturday, with goals from Sam Surridge and Hany Mukhtar, to clinch the Supporters' Shield for the first time. It was their 21st league win of the season and took them to 70 points.

Nashville are the highest-rated side in MLS, 1521, and their season deserves the trophy. But their points tally runs well ahead of their football. Including Saturday, they had 70 points, 15.7 more than expected, the biggest gap in the league, and they were only 3rd on deserved points per game. The best football in MLS, on our numbers, has been played by Vancouver, whose 2.08 deserved points a game leads the league but who have 55 points, 9.5 fewer than that football was worth. The regular season rewards points. The playoffs reward the football in each game. If the two sides meet there, the gap between them is smaller than the table suggests.

The week ahead

Two games still to come this weekend could change the picture. PSG play Lyon on Monday night with the chance to go top of Ligue 1, and Flamengo can return to the top in Brazil if they beat Atlético Mineiro. On Tuesday our weekly update adds every Saturday and Sunday result to the numbers, and The Luck Table will show which clubs gained and lost the most against what their football deserved. The Monday Index returns next Monday.

All our coverage of the season is in one place: see the world club ranking, our guide to Europe's luckiest and unluckiest teams, and every edition of The Monday Index.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerSpurs deserved their first win at Stamford Bridge since 2018. They had 69% of the ball, 15 shots to seven and dominated the air. Chelsea's set-piece weakness and Alonso's no-striker gamble cost them. Even so, the Index still rates Chelsea above Spurs, and Chelsea remain ahead of what their football deserves.

Tottenham had won one of their previous 36 league visits to Stamford Bridge. On Saturday night, in torrential rain, they won the 37th 2-0, and the scoreline undersold them. Roberto De Zerbi's side had 69% of the ball, 15 shots to Chelsea's seven, 27 touches in the Chelsea box to 17, and 17 aerial duels won to seven. The goals came late: a Mike Penders own goal from a corner in the 77th minute, and a solo run and finish from Mathys Tel in the 87th. In between, Cole Palmer was sent off for a second yellow card.

This is a morning-after piece, so it comes with a health warning. Chelsea's goalkeeper punched the ball into his own net, and their best player was sent off. Both things change a match. But what had happened before the 77th minute matters more for what comes next for both clubs, and that is where this piece starts.

How the game played out

Spurs had the ball in Chelsea's half for long spells, with their full-backs high and their midfielders passing around Chelsea's front line. Chelsea's attacks were mostly single runs and early balls, not sustained possession.

The opener came from a Spurs corner. Penders, making his first league start because Emi Martínez fractured his jaw against Everton last week, came for the ball with Jan Paul van Hecke challenging him. The ball came off his hand and went in, despite Maxence Lacroix's attempt to keep it out on the line. Chelsea's own report, the BBC and Spurs' report all describe it that way. Penders is only the second Chelsea goalkeeper to score a Premier League own goal, after Petr Čech in 2009.

The second goal belongs to Tel. He picked the ball up inside his own half on the left, held off Josh Acheampong, accelerated past him, cut inside Lacroix and hit a low right-footed shot into the bottom corner from around the edge of the area. It was, according to the club, his 50th league appearance for Spurs and a sign of how much space ten-man Chelsea had left behind their midfield by then.

The set-piece problem is now a pattern

Treat the own goal as a one-off if you like. Chelsea's record at set pieces says you should not. They have now conceded eight set-piece goals in the Premier League this season, double any other club. A goalkeeper on his league debut, in the rain, against a tall Spurs side that won 17 aerial duels to Chelsea's seven, was always going to be tested. Spurs kept putting balls into his area, and on the 77th minute it worked.

Our numbers add a broader point. Before Saturday, Chelsea were conceding 1.84 expected goals a game in the league, up from 1.46 across last season. Expected goals measure the quality of the chances a team allows, so that is a defence that has got clearly worse, not just unlucky. They were hit for 2.64 expected goals at Arsenal and 2.47 at Brentford. Xabi Alonso's side began the season in a back three and have moved to a back four for the last three league games. On Saturday Spurs created little in open play for long periods, but they never had to defend for their lives either.

They have now conceded eight set-piece goals in the Premier League this season, double any other club.

Alonso's selection gamble

With João Pedro injured, Alonso chose not to play a recognised centre-forward. Morgan Rogers played through the middle, 19-year-old Geovany Quenda made his first start on the wing, and Danny Welbeck began on the bench. The idea is easy to see: more runners, more pace, and a front line that could press Spurs' build-up and break quickly.

It did not work. Without a reference point up front, Chelsea's long balls had nobody to aim at, and Spurs' centre-backs were rarely stretched. Seven shots was Chelsea's lowest total in a league game this season and their fewest at home since January. Chelsea's own report notes that Palmer and Jordan Henderson forced good saves from Antonín Kinský in the first half, and the Index does not dispute that Chelsea had moments. But they were moments, not a plan.

Palmer's evening summed it up. He was booked for barging into Mateus Fernandes after Fernandes had stopped one of his driving runs. Sky's commentators argued it could have been a straight red. Nine minutes later, three minutes after the opener, he caught Conor Gallagher late in midfield and Chris Kavanagh showed him a second yellow. Chelsea's most creative player had spent much of the night chasing a game that was not coming to him.

De Zerbi's Spurs are starting to look like his teams

For Spurs this was the second win in six days, after Monday's 3-1 against Coventry, their first league win of the season. Before that week they had gone six league games without a win. They had been beaten 3-0 at Brentford on the opening day, conceding chances worth 4.89 expected goals, and they had conceded 1.97 expected goals a game before Saturday, worse than last season's 1.47.

What changed at Stamford Bridge was control. De Zerbi's teams are built to keep the ball deep, draw the opposition press and then play through it. Spurs averaged 32 possessions in attack a game before Saturday, well up on last season's 25, so the style was already arriving. On Saturday it arrived in full. Almost 70% possession away at a top-four side is not a fluke of game state, because the score was 0-0 for 77 minutes.

The defensive side is as important. Holding Chelsea to seven shots means Spurs defended largely by keeping the ball. They won the aerial battle 17 to seven, and Kinský needed to make four saves all night. If Spurs can defend like that against Chelsea, their awful start looks less like the real level of the team and more like a side still learning a new system.

What the Index says now

Here is the part that cuts against the mood. Even after this defeat, Chelsea are 3.2 points ahead of what their performances deserve, one of the bigger surpluses in the league. They have 13 points from eight games, but their football has been worth 1.22 points a game, 12th best in the Premier League. Saturday did not make Chelsea a worse team. It brought their results closer to what their football has been worth all season.

Spurs are almost exactly where they should be: +0.3 points against expectation, from 1.09 expected points a game. That matters because last season they finished 8.3 points below their expected total, one of the unluckiest seasons in the league. Two wins in a week have not made them lucky. They have stopped being unlucky.

Ratings move slowly by design, and one match does not overturn months of evidence. Chelsea are rated 1634, 12th in our world ranking, and still 51 points above Spurs at 1583, 25th. Around 180 points is worth a goal a game, so the gap is about a third of a goal. The difference is direction. Chelsea are rated as falling, Spurs as steady.

What comes next

Chelsea host Liverpool in the Carabao Cup on Wednesday, without Palmer for the next league game through suspension, and with a goalkeeping decision to make. The bigger fix is structural: a defence conceding clearly better chances than last season, a set-piece record that is now a pattern, and a front line that needs a centre-forward. They are fourth, but the numbers say that position is ahead of the football.

Spurs are 12th, four points behind Chelsea, after the best week of De Zerbi's time in charge. The Index does not read too much into one result, even one this emphatic. But the shape of Saturday's win, more than the scoreline, is the strongest evidence yet that the possession game he was hired to build is starting to work. If it holds, Spurs' table position will start to catch up with their football rather than the other way round.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerPalmeiras lead the Brasileirão by two points, but Flamengo have a game in hand and are the stronger team on the evidence. They are Brazil's highest-rated club, have the league's best underlying numbers, and depend less on luck. Palmeiras have the division's biggest surplus of points over chances.

Palmeiras went top of the Brasileirão on Saturday night. Abel Ferreira's side beat Red Bull Bragantino 2-0 at Arena Barueri, with a first-half penalty from Maurício and a second from Agustín Giay, to reach 69 points from 32 games. Flamengo have 67 from 31 and play their game in hand against Atlético Mineiro at the Maracanã on Monday night. With six rounds left for Palmeiras and seven for Flamengo, the title is a two-club race: third-placed Cruzeiro are 15 points behind.

The table says it is close to level. The Beaumont Index says Flamengo are the stronger team, and by a clear margin. They are the highest-rated club in Brazil at 1435, 84th in our world ranking. Palmeiras are second at 1394, 125th in the world. The gap of 41 points is worth about 0.23 of a goal a game, which is a lot between two teams separated by two points.

One caveat first. In Brazil, our expected-points figures come from each club's season-long average chance quality, not match-by-match shot data. They are a solid guide to which teams create and concede the better chances, but they are rougher than our figures for Europe's top leagues. Read the sizes below as indicative.

Same points, different football

Per game, the two clubs have taken points at almost the same rate: 2.16 for Flamengo, 2.16 for Palmeiras. Underneath that, the picture is different. On our estimate, Flamengo's chances created and conceded are worth 1.84 points a game, the best in the league. Palmeiras's are worth 1.71. Over a season that is a gap of four or five points of underlying performance.

The goals tell the same story. Flamengo have scored 61 and conceded 25, a goal difference of +36 from 31 games. Palmeiras have scored 56 and conceded 22, +34 from 32. Palmeiras's defence is excellent, but Flamengo create more and score more, and they have done it with a game fewer played.

The difference between results and football is luck, or at least things that do not usually last. Palmeiras have about 14 more points than their chances suggest, the biggest surplus in the division. Flamengo are also ahead of their chances, by about 10, which is normal for a dominant side. Teams that win a lot tend to convert slightly more and manage leads well. But a margin as large as Palmeiras's is hard to maintain over the last six games.

Three seasons of the same rivalry

This is the third year in a row that these two clubs have been at or near the top. In 2024, when Botafogo won the title, both finished close to their expected totals: Flamengo +2.7, Palmeiras +2.1. In 2025, Flamengo won the league with 79 points to Palmeiras's 76, and both ran ahead of their chances by similar amounts, +7.9 and +7.0.

Flamengo have scored 61 and conceded 25, a goal difference of +36 from 31 games.

So a modest surplus is what the top two in Brazil usually produce. What stands out in 2026 is the size of Palmeiras's: about twice what either club managed last year. That does not mean Ferreira's team are not good. They are the second-best side in Brazil on our ratings, 45 points clear of third-placed Athletico Paranaense on rating. It means their position at the top owes more to fine margins than Flamengo's does.

The run-in

The fixtures matter. Palmeiras go to Santos in a Paulista derby on Wednesday 28 October, then have the first leg of a Copa do Brasil semi-final against Vasco. Flamengo face Atlético Mineiro on Monday and Vasco in a Rio derby on Thursday 29 October. The two leaders still have to meet, at Palmeiras's Nubank Parque, in what could be the game that decides the title.

The Index does not make predictions, and this is not one. But if the question is which of the two teams is better, the evidence points one way. Flamengo are the top-rated side in the country, they have the best underlying numbers in the league, and they have a game in hand that would put them back on top with a win. Palmeiras are top because they have taken more from their chances than anyone else. That kind of edge tends to shrink.

What Monday means

The arithmetic of the game in hand is simple. A Flamengo win over Atlético Mineiro would put them on 70 points, one ahead of Palmeiras with both clubs on 32 games. A draw would take them to 68, a point behind. A defeat would leave them two behind with the same number of games played. Atlético Mineiro are no soft touch: they are sixth on 50 points, sixth in Brazil on our ratings too, and almost exactly where their chances say they should be.

Whatever happens on Monday, the direct meeting still to come at Nubank Parque gives the race a natural decider. A two-point swing over seven games is nothing. What the numbers add is a view on which side is more likely to keep producing results, and on that, Flamengo's case is stronger.

How good is Brazil's best?

A word on scale. Our ratings put an average club from Europe's top five leagues at about 1500. Flamengo, at 1435, sit about 0.36 of a goal a game below that average, which on our scale is the level of a lower-half side in one of those leagues. Palmeiras are about 0.59 of a goal below. Those are performance ratings adjusted for league strength, not a judgement on any individual player, and Brazil's league produces more talent than its club ratings suggest because so much of it is sold to Europe young.

Within Brazil, though, the top two are in a class of their own. The gap from Palmeiras to Athletico Paranaense in third on rating is bigger than the gap between Flamengo and Palmeiras.

The race behind them

Cruzeiro and Athletico Paranaense are level on 54 points in third and fourth, with Fluminense fifth on 51. Athletico's results match their chances almost exactly, while Fluminense have about 8 more points than their football suggests, the third-biggest surplus in the league. Cruzeiro are a little ahead of theirs. If the chasing pack shuffles over the last six rounds, Fluminense are the side with the most to lose.

At the bottom

The relegation fight has its own luck story. Internacional are 16th with 35 points, just above the drop zone, but on our estimate their football has been worth about 50. That shortfall of around 15 points is among the largest in the division, and a side with those numbers would normally be in the top half. Corinthians, who dropped into the bottom four after losing 3-2 at Vasco on Saturday, are about 10 points short of their chances, and Remo, 19th, about 10. Botafogo and Vasco, both mid-table, are also well below what their football deserves.

For Flamengo and Palmeiras, the next three weeks will settle a title. For Internacional and Corinthians, they may settle whether bad luck sends a big club down.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerSanfrecce Hiroshima are Japan's highest-rated club and have the best underlying numbers in J1, but are a point off the top and short of what their chances deserve. Kashima won the 2025 title far ahead of their chances. This season their luck has evened out, and they are eighth.

Twelve rounds into Japan's first autumn-to-spring J1 season, FC Machida Zelvia and FC Tokyo lead on 27 points, with Sanfrecce Hiroshima a point behind. The Beaumont Index has a clear view on who the best team in that group is. It is Sanfrecce, rated 1390 and 132nd in our world ranking, the highest-rated club in the country. They also have the best underlying numbers in the league this season, and fewer points than those numbers deserve.

At the other end of the luck story are Kashima Antlers, champions of the 2025 J1 season and runners-up in the one-off spring competition. On our numbers they have been the most fortunate big club in Japan over the past year and a half. This season, for the first time in a while, their results look roughly like their football, and they are eighth.

A note before the numbers. In Japan, our expected-points figures come from each club's season-long average chance quality, not from match-by-match shot data. They are a good guide to who is creating and conceding the better chances, but they are rougher than the figures we use for Europe's top leagues, especially after only 12 games. Treat the sizes below as indicative and the direction as the story.

Sanfrecce: the best team, a point off the top

Sanfrecce have 26 points from 12 games: eight wins, two draws and two defeats, with 29 goals scored and only 10 conceded. That is excellent. Our estimate says their football has been worth about 31 points, the highest expected total in the league, so they are around 5 points short of what their chances deserve. With the same luck as Machida, who have exactly the points their chances suggest, Sanfrecce would be top.

The pattern is not new. Across the 2025 season and the spring 2026 transition season, which we treat as one long campaign of 58 games, Sanfrecce finished about 11 points below their expected total. In 2024 they were close to fair, +1.9. A club that keeps creating more than it converts, or keeps conceding goals its defence should not, has sometimes found a structural flaw. More often it has been unlucky in the finishing at both ends, and the gap closes.

Our ratings are built to look through that kind of noise, and they rate Sanfrecce clearly above the rest. They are rated as rising. Their margin over Machida, the current leaders, is 32 rating points, worth about 0.18 of a goal a game. In a 38-round season with 26 still to play, that is a real edge.

Kashima: the champions whose luck has turned

Kashima won the 2025 J1 title on the final day, their ninth league title and first in nine years, under coach Toru Oniki. Leo Ceará finished as the league's top scorer with 21 goals. In June they reached the final of the one-off 100 Year Vision League, the spring competition that bridged the move to an autumn calendar, and lost 5-2 on aggregate to Vissel Kobe.

A club that keeps creating more than it converts, or keeps conceding goals its defence should not, has sometimes found a structural flaw.

It was a remarkable 18 months, and on our numbers it came with an enormous amount of help. Across those 56 league games, Kashima took about 119 points from football that our estimate values at about 83. That surplus of around 36 points is the biggest in Japan for that period by a distance. Some of it is genuine skill that a season-average model misses: a striker like Leo Ceará turns half-chances into goals, and a team that wins tight games late is doing something right. But no club sustains a gap that size for long.

The decline has arrived. Kashima have 20 points from 12 games this season, sixth on rating but eighth in the table. Their results are now almost exactly in line with their chances, +0.9 on our estimate. Nothing about their football has collapsed. What has gone is the surplus, and without it, Kashima are a good side in a crowded group rather than the team that always finds a way.

The rest of the top group

Machida are the most honest leaders you could ask for. Their 27 points match their expected total almost exactly. They are rated second in Japan and rising, and their 30 goals scored and 10 conceded are the best goal difference in the league. If Sanfrecce are the best team, Machida are the closest challenger, and nothing in their numbers suggests a fall is coming.

FC Tokyo, level with Machida on 27 points after Sunday's games, are rated third in Japan and rising, with results in line with their chances. Kashiwa Reysol, the 2025 runners-up, have 25 points and are the one leading club whose results are clearly ahead of their football: about 3 points on our estimate. They are worth watching for a slide.

Vissel Kobe, the spring champions, are fifth on 23 points with results bang in line with their chances. They also ran ahead of their expected points in the 2025 and spring 2026 period, by about 9 points, a smaller version of the Kashima story.

The unluckiest club in Japan is in the drop zone

The biggest gap in the league is at the bottom. Avispa Fukuoka have 7 points from 12 games and sit 19th, inside the three relegation places. Our estimate says their football has been worth about 16 points, which would put them comfortably mid-table. That shortfall of about 9 points is the largest in J1. Shimizu S-Pulse, 13th on 14 points, are also well short of their chances.

The fortunate clubs are further from the spotlight. Urawa Red Diamonds and Kawasaki Frontale both have around four points more than their chances suggest. Neither is in the title race, but both are sitting a little higher than their football deserves. With three relegation places and a long winter break to come, Fukuoka's numbers suggest they are better than their position, though that only helps if the points start to arrive.

What it means for the season

There are 26 rounds left, and the J1 title will not be decided until 2027. The Index does not make predictions, but it does say where the evidence points. Sanfrecce have been the best team in Japan on underlying performance for most of the past two years, and they have repeatedly finished with fewer points than that performance deserved. If their luck merely evens out, they should be the team to beat.

Kashima are the opposite case. Their title was real and their football was good, but the margin by which their results outran their chances was extraordinary. This season they are being paid exactly what they earn, and that leaves them eighth. Neither club has changed much. The luck has.

The same lesson runs through every league we measure. A big gap between results and performance is usually temporary, and the clubs on either side of it tend to drift back. In Europe this autumn the clearest examples are Viking and Nottingham Forest. In Japan, they are Kashima and Sanfrecce.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerNashville SC won the 2026 Supporters' Shield with 70 points and three games left, but they have the biggest surplus of points over chances in MLS. Vancouver Whitecaps have the best underlying numbers and are rated top on the Beaumont Index, despite having 15 fewer points.

Nashville SC won the 2026 Supporters' Shield on Saturday, beating CF Montréal 2-0 to clinch the best regular-season record in MLS for the first time in their history. They have 70 points from 31 games, 15 more than anyone else, and three games left to chase Inter Miami's single-season record of 74 points from 2024. Four more points would equal it. Five would break it. Their rate of 2.26 points a game is already the best of any Supporters' Shield winner since the league began in 1996, ahead of Miami's 2.18 in 2024.

It has been a superb season, and B.J. Callaghan's side deserve the trophy. But the Beaumont Index has a different answer to the question of who is the best team in MLS. On our ratings it is Vancouver Whitecaps, leaders of the Western Conference, rated 1407 and 112th in the world. Nashville are second at 1404. The gap is 3 rating points, which is nothing. The gap in the table is 15 points.

The regular season is not over. Clubs have three rounds left, running until 7 November, before the playoffs begin on 20 November. That is enough time for the luck in this table to start evening out, and in MLS it often does. A caveat on the figures that follow: in MLS, our expected points come from each club's season-long average chance quality rather than match-by-match shot data. They are indicative, and the sizes should be read with some care.

Nashville: brilliant, and fortunate

Nothing here takes away from what Nashville have done. They have conceded 23 goals, the fewest in the league, and their designated-player trio of Sam Surridge, Hany Mukhtar and Cristian Espinoza have 63 goal contributions between them. Their football is genuinely excellent: our estimate values it at about 57 points, among the best totals in MLS.

The trouble is that they have 70 points, about 13 more than that. It is the biggest surplus in the league. Some of it will be quality the estimate cannot see. A side that concedes so few goals often has a goalkeeper and back line performing above the average for the chances they face. But a gap that size usually includes plenty of good fortune.

The striking part is the history. In 2024, Nashville finished about 19 points short of their expected total. In 2025, they were about 11 short again. Two seasons of bad luck have turned into one of very good luck. The swing from last year to this is about 24 points. Their football has improved, and our ratings say they are rising, but the table has exaggerated how much.

Vancouver: the best football in the league

Vancouver's numbers are the mirror image. On our estimate, their chances created and conceded are worth about 65 points this season, the highest total in MLS by some distance. They have 55. They lead the Western Conference by a single point from FC Dallas, when their football says they should be clear.

The trouble is that they have 70 points, about 13 more than that.

And unlike Nashville, their luck has not turned. Vancouver were about 14 points short of their chances in 2024 and 7 short in 2025, when they still took 63 points. Three seasons in a row of finishing below their numbers is unusual. It may say something about their finishing or their goalkeeping. It also means that, on performance, they have been better than their record for a long time.

Our ratings, which are designed to look through this kind of noise, rate Vancouver as rising and the best team in North America's top league. They have scored 63 goals and conceded 32 in 31 games, the second-best goal difference in MLS behind Nashville.

The rest of the contenders

Inter Miami are third on our ratings and level on 55 points with New England in the East, behind Nashville. Miami's results are almost exactly in line with their chances. They score more than anyone, 73 goals, and concede plenty too. New England are the other big overachiever: about 11 points ahead of their chances, second only to Nashville, with football our estimate rates in the bottom half of the league.

At the other end, several clubs with decent underlying numbers are scrapping for wild-card places or have fallen out of contention. Philadelphia Union, last year's Shield winners, are about 9 points short of their chances. Seattle Sounders are about 8 short. Minnesota United, who were in the playoff places last year, are about 12 short and outside the top nine in the West. CF Montréal, bottom of the East, have the biggest shortfall in the league.

The Western race

The fight for the top seed in the West is tighter than in the East, and the numbers favour the leaders. FC Dallas, a point behind Vancouver on 54, have about 6 more points than their chances suggest. St. Louis City, on 51, are about 4 short of theirs and rated as rising. LAFC, sixth on 46 points, are about 5 short, but they are the one leading club our ratings mark as falling.

With three games each left, nobody can be written off. But if results drift towards performance in the final rounds, Vancouver and St. Louis are the sides with room to gain, and Dallas the side with the most to lose.

How good is MLS's best?

For readers used to European football, some scale helps. Our ratings put an average club from Europe's top five leagues at about 1500, with about 180 points worth a goal a game. Vancouver, at 1407, sit about 0.52 of a goal a game below that level. Nashville and Miami are close behind. These ratings are adjusted for league strength, so they compare MLS's best with Europe's directly. On that measure, the top of MLS sits a level below an average side in Europe's big five leagues, and close to the best clubs in Brazil and Japan.

What the run-in will tell us

The Index does not make predictions. What it can do is say where the evidence points as the regular season ends. Nashville have the Shield, the top seed in the East and home advantage through the playoffs, all earned. Whether they also have the record depends on three games, starting at home to Philadelphia on Wednesday 28 October.

But the playoffs are short, knockout football, where the luck that built a 15-point lead offers no protection. If the question is which team has played the best football in MLS this season, the answer is Vancouver. If the question is which team's results are most likely to come back towards the pack, it is Nashville. The playoffs, from 20 November, will show which of those matters more.

Across every league we measure, the same lesson holds. Clubs well ahead of their chances usually drift back, and clubs well behind them usually recover. In MLS this autumn, those two clubs are the two best teams in the league.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerPSG and Lyon are level on 14 points, but PSG are rated about 0.9 goals a game better and are at home. They are fifth in the world on our ratings because Ligue 1 is weaker than the Premier League. They dominate their own league more than Arsenal or City dominate theirs.

Paris Saint-Germain against Lyon was meant to be Sunday night's game. It was moved to Monday, 19:30 UK time at the Parc des Princes, after the security situation around Saturday's demonstrations in Paris. The extra day helps both squads a little. It does not fix PSG's injury list, and it does not change the most interesting fact about the fixture: two clubs level on 14 points from seven games, with our ratings putting them about a goal a game apart.

This preview also answers a question readers have asked since we published our world ranking. If PSG are, to many people, the best team in Europe, why does the Beaumont Index have them fifth?

Why PSG are fifth in the world

Our ratings put every club on one scale: goals per 90 minutes better or worse than an average club from Europe's top five leagues. Two things go into that number. The first is how far a club is above its own league. The second is how strong that league is against the rest of the world, which we measure from every cross-league match, including European competition.

On the first measure, PSG are outstanding. They are 1.22 goals a game better than an average Ligue 1 club. Arsenal are 0.90 better than an average Premier League club, and Manchester City 0.64. Of the six highest-rated clubs, only Bayern München (1.63) and Barcelona (1.46) dominate their leagues by more.

The second measure is where PSG lose ground. The Premier League rates about 0.80 goals a game stronger than Ligue 1 on our numbers, based on how English clubs perform against everyone else. Being a goal and a quarter better than Lens, Lyon and Marseille is a huge achievement. It is not the same as being a goal and a quarter better than Brighton, Liverpool and Chelsea every week.

Since this week we also count each club's own results in continental competition, not just its league's. That change moved PSG up our ranking, because their recent Champions League record is the strongest evidence that they are better than their league suggests. Tuesday's 4-2 win over Barcelona is part of the reason. Put together, PSG are rated 1691, 5th in the world, behind Arsenal, Bayern, Barcelona and Manchester City, and just ahead of Real Madrid.

Ratings come with uncertainty, and PSG's plausible range runs from 3rd to 16th. That width tells you something too. A club that plays most of its matches in a league with fewer strong opponents is simply harder to measure precisely. The evidence that PSG belong among the very best is strong. The evidence that they are clearly the best is not there yet.

Being a goal and a quarter better than Lens, Lyon and Marseille is a huge achievement.

Their rating is also falling. Our trend measure has PSG slipping by about 0.45 goals a game, and that was before the latest injuries. Khvicha Kvaratskhelia tore an anterior cruciate ligament against Barcelona and is expected to miss months. Ousmane Dembélé and Marquinhos have thigh injuries, and Nuno Mendes is out with a muscle problem. All four are expected to miss Monday's game, according to reports in France and the UK. Between them they have started most of PSG's biggest matches this autumn, so Luis Enrique's side will look different.

PSG's season: earned, not lucky

PSG have 14 points from seven league games and their football has been worth 1.94 points a game, almost exactly their actual rate of 2.00. That pattern goes back years. Across 2024-25 they finished +0.3 points against expectation, and across 2025-26 +2.7. For three seasons, PSG's league results have been an honest reflection of their football. There is no cushion of luck to run out.

Their form is not spotless. Their recent league run includes a draw and a defeat, and the 3-1 win over Strasbourg last weekend came at the cost of Mendes's injury. But the Barcelona result showed the squad's depth: Lucas Hernández, Vitinha, Fabián Ruiz and Désiré Doué all scored. Doué has scored in his last three matches.

Lyon: a good start with some help from last season

Paulo Fonseca's Lyon are also on 14 points, and top of the table on goal difference going into the weekend. They beat Nice 2-1 last Sunday through Pavel Šulc and Tanner Tessmann. They are rated 1520, third in Ligue 1 on our numbers and 42nd in the world, and rated as rising. Their football this season has been worth 1.83 points a game, so their position is mostly deserved.

Last season is the warning. Lyon finished +7.2 points ahead of their expected total in 2025-26, a big surplus for a season. This season they are only +1.2 points ahead, which is close to fair. They have become a team whose results match their chances, and that is a good sign for how long this start can last.

Fonseca also rotated heavily for Thursday's 3-0 Europa League defeat at Hoffenheim with this game in mind, starting Steeve Kango and Rémi Himbert for the first time in this run. That says a lot about how much Lyon want this match. Keito Nakamura is suspended after his red card against Nice, Lois Openda is injured, and Nicolás Tagliafico and Mohamed Ouédraogo are also out. Šulc is a doubt with a knee problem.

Where the game is decided

PSG are expected to line up in their usual 4-3-3, with a front three built around Doué on the left, Ferran Torres through the middle and Maghnes Akliouche on the right, according to French reports. That is a mobile, rotating line rather than one with a fixed reference point, and it relies on Vitinha and the midfield to keep the ball high up the pitch. Illia Zabarnyi is expected to partner Willian Pacho in place of Marquinhos.

Lyon's best route is the counter. With Fonseca's side likely to sit deeper than they do against most Ligue 1 opponents, the space behind PSG's full-backs is the obvious target, particularly with Achraf Hakimi pushed high on the right. Ernest Nuamah's pace on the break is the kind of threat that can punish it, and Corentin Tolisso's late runs from midfield give Lyon a second route to goal.

The verdict

On our numbers, PSG are about 0.9 goals a game better than Lyon before injuries, and they are at home. That is a big gap between clubs level on points. Missing four regular starters narrows it, but PSG's depth is exactly what the Barcelona win demonstrated. The Index does not predict scores. It does say that the table currently flatters the closeness of this pair more than the football does.

As for the bigger question: PSG are an elite team, rated fifth in the world and the highest-rated club outside England, Germany and Spain. If they keep winning in the Champions League, our continental-results term will push them higher, which is how it should be. A ranking has to be earned in matches against the best, not in the opinion polls.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerAston Villa beat Manchester City 2-1 by defending deep and compact, pressing hard early and breaking quickly. Nicolas Jackson scored both goals, the first from a City error in build-up. City dominated the ball and created more, but less than usual, and lost at Villa Park for the fourth straight league visit.

For the first quarter of an hour at Villa Park on Saturday lunchtime, Manchester City looked like a team that had forgotten how to start a football match. They took so long over one goal-kick, more than the eight seconds now allowed, that the referee gave Villa a corner for it. In the sixth minute they tried to play out from the back under Villa's press. Marc Guéhi took a heavy touch, John McGinn pounced and played Nicolas Jackson in, and Jackson turned inside Elliot Anderson before finishing high past Gianluigi Donnarumma.

That should have been the cue for a long, slow City squeeze. In a sense it was. City had 63% of the ball, made 528 passes to Villa's 314, and put seven shots on target. Zion Suzuki made six saves in the Villa goal. And Villa won 2-1, which means Enzo Maresca's side, winners of all seven league games, lost to the team with the lowest expected-points rate in the Premier League.

So how did a side that had been playing like a relegation candidate beat the league's form team? The answer is partly shape, partly nerve and partly something that Villa have been doing for a season and a half.

The shapes: mirror images, different intentions

On paper the two teams were identical. Unai Emery picked a 4-2-3-1. So did Maresca, who has used that shape in all seven league games this season, after City used at least seven different formations across the previous twelve months of league football. When two 4-2-3-1s meet, the game tends to be decided by what the double pivots do and by which full-backs win their duels.

Villa's double pivot, Boubacar Kamara and João Gomes, had the better of City's midfield by most accounts. Villa did not try to compete for the ball. They tried to deny City space between the lines and break when they won it. The Guardian's match report described Villa defending "with commitment and discipline to limit the space available", and the numbers back that up. Villa had 37% of the ball and committed 21 fouls, four more than City, which means they were willing to stop moves before they started. Four Villa players were booked, including the right-back Matty Cash.

City's build-up came mostly down the flanks, through Jérémy Doku on the left and Iliman Ndiaye on the right. That was where they looked most dangerous all afternoon, and that was where Villa doubled up, putting two players on each winger whenever they could. It slowed City down but did not stop them. Ndiaye cut in from the right and curled a shot against the foot of the post, and he had a penalty claim waved away. Suzuki kept turning shots away from Ndiaye, Cherki and Doku. Doku's runs on the left caused trouble but never quite produced the goal. City's chances were good but not overwhelming. Their expected-goals total was 1.81, below their season average of 2.42 and, tellingly, below the 1.84 that the average Villa opponent had managed this season.

Read that last sentence again. Villa had been conceding plenty of chances before Saturday. Against City, they conceded fewer than usual.

Villa's double pivot, Boubacar Kamara and João Gomes, had the better of City's midfield by most accounts.

The press: brave early, smart late

Villa's start was the most aggressive part of their afternoon. They went at City's build-up for the first 15 minutes, and the opening goal came straight from it: City trying to pass through the press, a miscontrol under pressure, a quick pass and a finish. The official City report described Villa as flying "out of the traps".

After that Villa dropped deeper and stayed compact. City kept the ball in Villa's half for long spells. The risk for Villa was obvious, and in the first half they rode it with saves from Suzuki and some last-ditch defending. The reward was obvious too. With City committed forward, every Villa turnover was a chance to run at a back line with space behind it, and both match reports noted the threat Villa carried on the break.

Both of City's first-half bookings came from that dynamic. Guéhi was booked in the 32nd minute for dragging down Jackson as Villa broke; Rayan Cherki was booked six minutes later. At half-time Maresca took both off, sending on Rico Lewis and Phil Foden, which means City spent the second half with a reshaped defence against the striker who had been causing the trouble.

Key duel: Jackson against City's centre-backs

The Guardian's report called Jackson against City's defence a mismatch, and it was hard to argue. He scored both goals and drew the foul that led to a booking. Emery said afterwards that it was how Jackson worked and defended, as well as the goals, that impressed him. He was replaced with eight minutes left.

His second goal was the clearest example of how Villa planned to hurt City. Emery sent on Ibrahim Mbaye for Alysson Edward in the 59th minute. A minute later City lost the ball, Jackson moved it on to Mbaye, and Villa broke. Mbaye powered down the right and pulled a low ball back to Jackson, whose first-time shot crept under Donnarumma's dive. It was Villa's second shot on target of the match. Both had gone in.

That is the hard part to explain if you only watch City's possession. Villa scored twice from chances worth 0.64 goals, which is less than half their season average. Donnarumma will not enjoy watching the second again. City's equaliser was arguably the best goal of the game: a short free-kick, a cross from Ndiaye, and Erling Haaland heading in seven minutes after half-time, his ninth league goal of the season by City's count. Late on, Ian Maatsen cleared a goal-bound Haaland header to protect the lead. Both Villa goals came from City turnovers in transition, which is what Emery had built the plan around.

The habit Villa can't kick

Here is the multi-season part, and it's the most Bill James thing about this game. Last season Villa were about 14 points ahead of what their performances deserved, the kind of gap that usually closes quickly. This season, before Saturday, they were still 2.4 points ahead, despite the league's lowest expected-points rate. Saturday added another two goals from not much.

Is that luck or skill? Probably both. Sustained overperformance usually means a mix of good finishing, a keeper having a good spell and a habit of scoring first and then defending well. Villa did all three on Saturday. But the underlying trend is a worry. Compared with the previous twelve months, Villa are facing 14.9 shots a game, up from 11.6. They are crossing less and getting fewer shots from corners. They are making more interceptions, which suggests a team spending more time without the ball.

Villa Park, City's least favourite away day

History mattered here too. This was Villa's fourth straight home league win over City, a run that started in December 2023, and every one of them has been by a single goal: 1-0, 2-1, 1-0, 2-1. Villa have also won the last three league meetings anywhere, including a 2-1 win at the Etihad in May.

There is a pattern inside that run. City tend to dominate the ball at Villa Park and Villa tend to score on the break, keep their shape and live with the pressure. Every big club has a ground where it seems to play worse. For City, Villa Park has become one.

What it means for City

City came into the game as the Premier League's luckiest team on our numbers, 7.6 points ahead of what their performances deserved. A perfect start built on performances worth 1.92 points a game was always going to meet a day like this. Saturday was not a bad performance by City's standards, just a slightly below-par one by their numbers, against a goalkeeper and a striker having very good days.

City were rated 1731 before kick-off, the second-highest rating in the league. One defeat won't change that much. The thing to watch is whether the defensive errors that cost them at Villa Park keep coming. Two goals from mistakes in one afternoon is a warning, not a crisis.

What it means for Villa

For Villa, this was Emery's team at its most Emery: compact, disciplined, organised to win the moments that matter. The question is whether they can win more of the ball, more often, without living on Jackson's finishing. The ratings, which do not yet include this game, have them at 1626, 9th in the league, and trending down.

One afternoon does not change a trend. But it does show what Villa can still do when the plan works. City had the ball. Villa had the goals. On Saturday, that was enough.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerBarcelona are rated 3rd in the world and Real Madrid 6th, a gap worth a little under half a goal a game. Barcelona have won all nine league games, lead Madrid by six points, are at home and have won five of the last six competitive Clásicos. Madrid's numbers are strong, but Barcelona's are better.

Nine games, nine wins, 39 goals. Barcelona come into Sunday's Clásico with a perfect league record and a six-point lead over Real Madrid, who are third behind Atlético. Spotify Camp Nou, 20:00 UK time. The referee had not been named when this was written. If Barcelona win, the gap is nine points before November.

The Beaumont Index rates Barcelona 1760, 3rd in the world behind only Arsenal and Bayern München. Madrid are 1680, 6th. These are two of the best teams on the planet, and the gap between them, 80 rating points, is worth a little under half a goal a game. That is a real edge. It is not a chasm, and Madrid have closed bigger ones in this fixture.

Barcelona: better than last season, which was very good

Hansi Flick's side have been extraordinary. They are creating 3.46 expected goals a game in the league, up from 2.24 last season, and conceding 0.99, down from 1.24. Opponents average 6.3 shots a game against them. Only one league opponent, Levante, has matched them on chances, and Barcelona still won that game 4-2. They beat Racing 7-2 with more than six expected goals. Last Saturday they won 5-2 at Real Betis, with Lamine Yamal, Rodri, Gabriel Jesus, Eric García and Dani Olmo all scoring.

The depth shows in that list. Raphinha is the league's top scorer with 12 goals, yet he did not start either of the last two league games, and Barcelona scored eight in them. Yamal has eight goals and six assists. Pedri runs midfield. Pau Cubarsí and Gerard Martín started in defence at Betis, with Jules Koundé at right-back. Flick has switched between a 4-3-3 and a 4-2-3-1 without the team noticing.

They are also a little ahead of their numbers. Barcelona are +3.3 points above what their chances deserved this season, and were +14.9 last season. That second number is enormous. Some of it is finishing quality, which is real; some of it will not last. Either way, nine wins from nine is not an accident.

The one warning is midweek. Barcelona lost 4-2 at Paris Saint-Germain on Tuesday, a reminder that their high line can be exposed by elite pace. Madrid have elite pace.

Real Madrid: two defeats, two very different stories

José Mourinho's Madrid have taken 21 points from nine games and are −0.2 against their expected total. In other words, they have earned exactly what their football deserved. They create 2.52 expected goals a game, up from 1.95 last season, and concede 1.22.

Barcelona are +3.3 points above what their chances deserved this season, and were +14.9 last season.

Their two defeats are worth separating. At Real Betis in September they lost 1-0 despite creating 3.59 expected goals to 1.00, as clear a case of bad luck as you will see. A week later they lost 2-1 at Atlético with 0.60 to 2.32: outplayed, simply. The first is noise. The second is the more useful one for Sunday. In a hostile away ground against a well-drilled team, Madrid struggled to create anything at all.

Mourinho has settled on a 4-2-3-1 in every league game, a change from last season, when a 4-4-2 was Madrid's most common shape. Kylian Mbappé leads the line and has nine league goals. Jude Bellingham plays behind him. Arda Güler and Aurélien Tchouaméni were the double pivot against Sevilla last Saturday, with Federico Valverde at right-back. Madrid won that game 4-1: Güler scored, Mbappé got two, one a penalty, and Diomande added the fourth.

The selection keeps moving. Against Villarreal earlier in the month, Trent Alexander-Arnold, Antonio Rüdiger and Álvaro Carreras started in defence; against Sevilla it was Valverde, Ibrahima Konaté, Dean Huijsen and Marc Cucurella. Madrid beat Leipzig 4-1 on Wednesday, the night after Barcelona lost in Paris, and arrive in the better mood.

The history: Barcelona's run

There have been six competitive Clásicos since January 2025, and Barcelona have won five. A 5-2 in the Super Cup, 3-2 after extra time in the Copa del Rey, 4-3 in the league, 3-2 in the Super Cup again, and 2-0 at home in May. The exception was last October, when Madrid won 2-1 at the Bernabéu. Go back to October 2024 and there was the 4-0 in Madrid that started it.

These games are close more often than the scores say. Three of Barcelona's five wins in that run came by a single goal. In Clásicos, the team that takes its moments usually wins, and for 18 months that has mostly been Barcelona.

Stretch the view to league meetings only and the picture is a little more even. Of the last five league Clásicos, Barcelona have won 3 and Madrid 2. Madrid's wins came at the Bernabéu. Sunday is not at the Bernabéu.

The bigger picture

The Index also tracks which way teams are heading, and here the clubs split. Barcelona's rating trend is marked as rising: their underlying numbers have improved from a very high base, and their rating has kept climbing with them. Madrid's is steady. Both clubs finished last season well above their expected points, Barcelona by 14.9 and Madrid by 10.1, so neither is short of finishing quality. The difference this season is that Madrid's surplus has gone and Barcelona's has not.

That matters for the title race as much as for Sunday. Atlético, second on 22 points, are rated 20th in the world on 1598, a long way behind both. On rating, this is a two-horse race, and Sunday is the first time the horses meet.

Where this is decided

Barcelona's high line against Mbappé is the matchup everyone will watch, and the midweek defeat in Paris will make it louder. Barcelona defend by pushing up and squeezing the pitch, which is why opponents manage so few shots. The cost is the ball over the top. Mbappé, with Endrick or Diomande running alongside, is as good as anyone in the world at punishing it.

At the other end, Madrid's midfield has to cope with Pedri and the runs of Fermín López and Yamal between the lines. Madrid concede more chances than Barcelona do, 1.22 a game to 0.99. Over 90 minutes at Camp Nou, that difference adds up.

Fatigue cuts both ways. Barcelona played on Tuesday, Madrid on Wednesday. Barcelona have the extra day of rest; Madrid have the better result to carry with them.

The verdict

The Index makes Barcelona the better team, at home, by a little under half a goal a game. Their underlying numbers this season are the best in LaLiga by a distance, and their record in this fixture lately is excellent. Madrid's are good, honest numbers that say they are a top-six team in the world, which is exactly where the ranking has them.

But a half-goal edge in a single game leaves plenty of room. Madrid won this fixture as recently as last October, and Mourinho has spent a career preparing for nights like this one. A Barcelona win makes the title race look like a procession. A Madrid win makes it a race again.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerBarcelona beat Real Betis 5-2 at La Cartuja on 17 October 2026 because they controlled almost everything that can be measured: 69% of the ball, 18 shots to six, and chances worth roughly three goals to Betis's one. Lamine Yamal won eight of his ten duels from the right wing, and Facundo Bernal's red card at 2-3 ended the contest.

The ball was still rolling towards the far corner when the first Betis fans began to sit down. Ten minutes in at La Cartuja, Lamine Yamal had rolled his foot over the ball to beat Facundo Bernal and hit a fierce strike from outside the box, and the stadium that had come to find out whether third could live with first got its answer early. It would get it four more times.

The final score was 5-2. There was a sensational Giovani Lo Celso volley ruled out, a straight red card for Facundo Bernal, and, by the end, Barcelona had nine league wins from nine. All of that will be remembered. This piece is about the bits that won't be: the shapes, the switches, the duels and the moment the game stopped being a game.

The shapes: a mirror, with one crack

On paper this was a mirror match. Manuel Pellegrini and Hansi Flick both named a 4-2-3-1. For Betis that was routine: it was their main shape in seven of their previous eight league games. For Barcelona it was a choice. In their previous eight league games, Flick's main shape had been a 4-3-3 five times. Here he went with two holding midfielders, Rodri and Pedri, behind Fermín López, with Yamal right, Karim Adeyemi left and Gabriel Jesus up front.

Put each starter's numbers where the team sheet put them (listed positions, not where they actually stood) and the match draws itself. Barcelona's back four and double pivot attempted 522 passes between them, 70% of the team's 751. Five Barcelona players attempted 60 or more. No Betis player got past 38, and that was Valentín Gómez, a centre-back. The best-rated Betis player was their goalkeeper, Álvaro Valles, on 7.5, which is what happens when you face this many shots and only let in five. The two lowest-rated players on the pitch were Betis's full-backs: Héctor Bellerín on 5.2 and Fran García on 5.3. Barcelona's three best starters were Yamal (8.6), Rodri (8.3) and Cubarsí (8.2), one from each line of the team.

The double pivot matters because of what it allowed the full-backs to do. With Rodri and Pedri both sitting, Jules Koundé and Xavi Espart were free, we'd infer, to push on without leaving the centre-backs exposed. Koundé set up the first goal; Espart attempted more dribbles than any Barcelona defender. Barcelona held their 4-2-3-1 for the whole match. Betis held theirs for about 61% of it; most of the rest, we'd infer, was the reshuffle after the red card.

The crack in the mirror was on the Betis left. Fran García, listed at left-back, had Yamal in front of him, and Abdessamad Ezzalzouli, ahead of him, is a winger whose job is to run at people: he tried five dribbles. That's a reasonable trade most weeks. Against Yamal, it was an invitation.

How Barcelona attacked: patience, not speed

Barcelona weren't trying to beat Betis on the break. They had one counter-attack all match, about their season norm. What they had instead was volume: 47 positional attacks, against an average of about 35 in their previous eight league games, with 14 of them ending in a shot. They passed the ball 751 times and completed 92%. Rodri alone completed 126 of 129. Pedri completed 103, and Pau Cubarsí 106 from centre-back.

The double pivot matters because of what it allowed the full-backs to do.

Here's the interesting bit, though. For all that control, Barcelona's chances weren't unusually good by their own standards. Depending on which model you trust, their chances were worth somewhere around three goals, which is a big number, but below their season average on either of the models we use. Their 18 shots were slightly fewer than usual, and of their 16 crosses only two found a teammate. What Barcelona did unusually well was finish: five goals from chances worth about three. Rodri headed in Fermín's corner. Eric García tapped in after Cubarsí got to the ball first following a blocked Yamal free-kick and squared it. Yamal's strike and Dani Olmo's late strike were the kind of finishes no chance model expects.

How Betis attacked: almost not at all

Betis's numbers are the starker ones. In their previous eight league games they averaged about 30 positional attacks, 15 crosses, five corners and 16 shots. Here they managed 17 positional attacks, three crosses, two corners and six shots. They had 31% of the ball and completed 256 passes, about a third of Barcelona's total.

And yet the six shots were good ones. Five were on target, and the chances were worth a little over a goal. Betis's two goals both came from moments when they briefly got numbers forward: Valentín Gómez heading in Pablo Fornals' cross after a corner wasn't cleared, and Ezzalzouli racing clear to finish across Joan García straight after Barcelona's third. Lo Celso's ruled-out volley, cancelled after a review found a handball by Cucho Hernández earlier in the move, would have been a third moment. That's the honest summary of Betis's attack: not much of it, but not nothing. Joan García had five shots on target to deal with, and two went in.

Pressing and defending: Barcelona squeezed, Betis retreated

The clearest single number in this match is how far apart the two pressing styles were. We measure pressing as the number of passes a team allows its opponents before making a defensive action: lower means more aggressive. The LaLiga median this season is about 11.6.

Barcelona allowed Betis 7.4 passes per defensive action, close to their own season average of 6.5 and far more aggressive than the league typical. Betis allowed Barcelona 38.3. Their season average is about 13. Some of that is the red card, and some is simply what happens when the other side has 69% of the ball. We can't split these numbers by half, so we can't say how much of the retreat was choice. The duels suggest it wasn't only the red card.

Betis won 42% of their defensive duels, against a season average of 58%. Barcelona won 64%, against their own norm of 55%, and they won 58% of their attacking duels, against a usual 48%. That's the match in two numbers: when a Barcelona player ran at a Betis player, he usually got past. Interceptions were bang on both teams' averages, 39 for Betis and 33 for Barcelona, which tells you the Betis block was in roughly the right places. The problem was what happened when it had to engage.

The individual numbers count more strictly, so their totals are smaller, but they tell the same story and show who did the work. Betis players made 17 tackles and 14 interceptions, 31 defensive actions in all, against Barcelona's 21. More than half of Betis's came from the back four: seven tackles and ten interceptions, against four and four for Barcelona's defenders. Fran García (two tackles, three interceptions) and Diego Llorente (three and two) made five each, the most on the pitch. Bellerín made three interceptions and won none of his three duels. We'd read that as a back line that spent the afternoon reacting rather than stepping out, though without locations that is an inference.

Barcelona's defending was spread more evenly up the pitch. Their midfielders made nine tackles, the same as Betis's, and Pedri's three were the most by any Barcelona player. Their forwards chipped in two, both from Jesus, while Betis's forward line made none. Even Gavi, on for the last 11 minutes, made two. That fits the pressing numbers above, though we should be clear about the limits: these counts tell you who made a tackle, not where on the pitch it happened.

The key battle: Yamal against the Betis left

If you want one matchup to explain the afternoon, it's this one. Yamal played 70 minutes, contested ten duels and won eight. He tried eight dribbles and completed six, had three shots with two on target, scored the first goal and set up the third. Fran García, on the other side of that flank, contested 13 duels and won four, and was dribbled past seven times, more than anyone else on the pitch.

Yamal also led the match for attacking involvements: three shots, two key passes, a goal and an assist. Across the team, Barcelona players made 16 key passes to Betis's six. Six Barcelona players made two each, including both holding midfielders and Cubarsí from centre-back. Only Lo Celso and Fornals did that for Betis, and between them they set up both Betis goals.

We should be careful here. These are match totals, not a log of one-on-one contests, and Fran García was often left two against one by a winger ahead of him who was looking the other way. But the shape of the numbers is clear enough. Barcelona's best player spent the afternoon running at Betis's most exposed defender, and the Betis plan didn't change to protect him until it no longer mattered.

The other flanks were closer, and more interesting. Ezzalzouli, the winger in front of Fran García, won four of his nine duels against Koundé's side of the pitch but completed only one of his five dribbles. He still scored, because one clean run was enough. On the other side, Adeyemi won four of his six duels and Bellerín won none of his three, though Bellerín did make three interceptions. Lo Celso, on the Betis right, won three of eight duels against Xavi Espart's flank, but his two key passes made him one of Betis's two main creators. Espart was busy at both ends: 11 duels, four won, three dribbles tried and two shots, neither on target.

The other battle was in the middle. Betis's double pivot of Bernal and Dani Ceballos was meant to stop Rodri and Pedri dictating. Between them, in the minutes they played, they made 25 passes. Rodri and Pedri made 241. Pellegrini took Ceballos off at half-time for Nelson Deossa, the first sign that he knew the midfield had gone. That wasn't for lack of effort: Ceballos won four of his seven duels and made three interceptions in 45 minutes. Pedri, opposite him, won six of 11.

The scatter shows the two kinds of player on the pitch. In the top-right corner are Barcelona's circulators: Rodri at 98%, Cubarsí at 97%, Pedri at 92%, all with more than 100 passes. Bottom-left are the forwards who were asked to do difficult things with the ball. The four lowest pass-accuracy figures among the starters belonged to Cucho Hernández (64%), Yamal (66%), Ezzalzouli (67%) and Lo Celso (68%). For Yamal that is the cost of trying things; for the Betis attackers it is the cost of receiving the ball with three Barcelona shirts around them. Betis as a team completed 79% of their passes, Barcelona 92%.

The changes, and the turning point

Both coaches made half-time changes. Pellegrini's was about control: Deossa for Ceballos. Flick's was like for like at centre-back: Eric García for Gerard Martín. García would score. Jesus had a goal ruled out for offside moments before, then, six minutes after the restart, Yamal squared for him to make it 3-1.

Then came the two minutes that decided it. At 53 minutes Ezzalzouli made it 3-2, and for the first time since the 27th minute Betis had a game. Within seconds, Bernal and Yamal came together in the scramble that followed; Bernal was shown a straight red card for violent conduct. Until that moment Barcelona had been the better team without being out of sight. From that moment they were playing against ten men who had to chase a goal.

Flick's response was immediate and aggressive. At 62 minutes he sent on Olmo and Anthony Gordon for Fermín and Adeyemi, and in the same minute García finished from close range after Yamal's free-kick was blocked and Cubarsí got to the ball first and squared it. Pellegrini went for broke at 64 minutes, bringing on Rodrigo Riquelme and Troy Parrott, then Ángel Ortiz and Álvaro Fidalgo at 70. It changed nothing. Olmo's strike in the 81st minute was the fifth, and Cubarsí's second assist. After the red card the score was 2-0 to Barcelona, and Betis didn't manage a goal with ten men.

A fair judge has to weigh the red card properly. At 2-3, with 37 minutes left, Betis were in the game. But the numbers before it already favoured Barcelona heavily, and the gap on our ratings, 224 points or roughly 1.2 goals a game, said a Barcelona win by about a goal was the fair expectation anyway. The red card turned a likely win into a rout. It didn't turn a Betis win into a defeat.

What it means next

For Betis, this was the reckoning the Index had been warning about. Going into the weekend they were collecting 2.29 points per game from performances worth 1.40, the biggest overperformance in LaLiga, about six points over seven games. On the Index they are rated 1532, a good team; Barcelona are rated 1763. Bernal now faces a ban, and without him Pellegrini has to rebuild the double pivot that lost the midfield here. The tactical question for Betis isn't really about Barcelona. It's whether the left side can be protected against good wingers, because Yamal won't be the last.

For Barcelona, the more interesting test comes next, against a Real Madrid side the Index rates second in Spain. The 4-2-3-1 with Rodri and Pedri together looked like an away-game shape: control first, attack through the right. Whether Flick keeps it, or goes back to the 4-3-3 he has used more often this season, will tell us how he sees that game.

At La Cartuja, the fans who sat down early were right. They had seen enough to know. The numbers had known for a while.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerManchester United, Fulham and Burnley should not sack Michael Carrick, Álvaro Arbeloa or Nicky Hayen on this autumn's results. Across 6,876 head-coach spells, the median completed spell lasted four months but teams peaked at a median of 16.3 months, and all three clubs' performances are running ahead of their results. Judge them at 16 months, on performances.

Update, 25 October 2026: Burnley sacked Nicky Hayen on 24 October, shortly after their 4-1 defeat at Lincoln City left them 22nd in the Championship. He won one league game in his 14 matches in charge, and the club confirmed his dismissal in a statement. This analysis was written before he was sacked and is otherwise unchanged. The Carrick and Arbeloa sections still stand.

Let me begin with an argument that will make me unpopular in at least three postcodes. Michael Carrick, Álvaro Arbeloa and Nicky Hayen are, at the time of writing, all under the kind of pressure that tends to end in a statement thanking someone for their efforts. I don't know whether any of them is a good coach. Neither, I'd suggest, does anyone else. Not yet.

That isn't fence-sitting. It's a claim about evidence, and it comes from counting. Our study looked at 6,876 head-coach spells covering about 346,000 matches since December 2022. Two numbers came out of it, and when you put them side by side they describe a sport behaving rather strangely.

Number one: four months

The median completed spell lasted four months. Not four seasons. Four months. Of the completed spells in the study, about 36% were over inside three months, and almost 63% inside six. By the one-year mark, nine in ten were done.

I should be careful here, because "completed" is doing a lot of work. A spell that is still running isn't in that figure, and long-serving coaches are disproportionately still running. The four months describes the spells that ended, not every spell. But that's exactly the point. When football ends a spell, it usually ends it early.

Number two: sixteen months

The second number comes from the spells that lasted long enough to measure properly. For each one, we found the stretch where results were furthest ahead of what the squad would be expected to achieve. The median of those peaks was 16.3 months into the job. For a meaningful share, 27% of the spells we could measure, the peak came after two years.

You can argue about why. New coaches need a transfer window or two. Players need time to learn patterns that don't come naturally. Some of it is surely just that bad coaches get sacked before they can have a peak, so the survivors look better than average. I take that point seriously, and it is the main reason this is an essay and not a law. But even allowing for it, the gap between four and sixteen is too big to be an accident.

The second number comes from the spells that lasted long enough to measure properly.

Now, the three names

Michael Carrick took interim charge at Manchester United on 13 January 2026 and was made permanent in May. That puts him at a little over nine months. In the curve above he is past the point where four in five completed spells have ended, and well short of the median peak. On the Index, United are rated 1672, about a goal a game better than an average top-five-league club, and our latest run had them about three points worse off than their performances deserved. That doesn't save him from a bad run. It does suggest the bad run is less bad than it looks.

Álvaro Arbeloa was appointed at Fulham on 7 July. Nicky Hayen took over at Burnley on 10 July. Both are around three and a half months in. Both clubs were, in our latest run, behind their performances: Fulham by about three points, Burnley by nearly four. If either is sacked this autumn, it will be inside the window where the study says sackings cluster, and a full year before the window where it says teams peak.

What the peak curve actually looks like

It's worth looking at the shape, not just the middle. Of the 1,909 spells we could follow for two years, only 250, about 13%, peaked inside the first six months. Almost 40% peaked between 18 months and three years. The early months are the least likely time for a team to be at its best under a new coach, and they are exactly when most decisions get made. If you wanted to design a system for firing people just before they come good, you would struggle to beat this one.

Performances, not results

If time is the first thing football gets wrong, the second is what it measures. A results table tells you what happened; it's a poor guide to what will happen. So here's how the three clubs look on what their football has earned. Manchester United's performances were worth 1.67 points a game in our run, against 1.00 actually won. Fulham's were worth 1.01, against 0.71. Burnley's were worth 0.96, against 0.73.

Those three aren't equally reassuring, and it would be dishonest to pretend they were. Carrick's case is the strongest: a team rated 1672, playing like a top-six side, losing games it should be winning. Arbeloa's is reasonable: a Fulham side playing like a mid-to-low Premier League team and being rewarded like a bottom one. Hayen's is the weakest, because Burnley's football is only worth about a point a game even before the bad luck. But in each case the gap points the same way. Results are worse than the football, and results are what get coaches sacked.

Put the expected rates side by side and the three cases separate. United's football has been worth 1.67 points a game, the sixth-best expected rate in the Premier League. Fulham's has been worth 1.01. Burnley's, in a different division, has been worth 0.96, under a point a game before any bad luck is counted. That's the uncomfortable detail in Hayen's case. The case for patience there rests on the study, not on his team's luck turning.

The obvious objection

"But some coaches are just bad." Yes. Some are. The study can't tell you which ones, and I'm not going to pretend it can. What it can tell you is that the decision is usually being made before the evidence exists. A chairman who sacks a coach at four months is not reacting to the coach's peak, because the peak hasn't happened. He is reacting to noise, form and the crowd. Those are real pressures. They are not the same thing as information.

What a sensible club would do instead

None of this means a coach should be untouchable. It means the review should happen when the evidence exists. A club that wanted to act on this study would set its first serious review somewhere between 12 and 16 months, and would bring expected points, not just the league table, into the room. It would ask whether the team is creating more than it concedes, and whether that's improving. If the answer is no on both counts at 16 months, the decision is easy and well informed. If the answer is yes, the club has probably avoided paying for a new coach, a new squad rebuild and another four-month gamble.

That costs patience, which is the scarcest thing in football. But the alternative is making the same expensive decision, early, again and again.

The exception that proves the rule

The best counter-example in our eight leagues is in Norway. Kjetil Knutsen has been head coach of Bodø/Glimt since 1 January 2018. On the Index, his side are rated 1566, the best in Norway. One club proves nothing on its own. But if you set out to design an experiment in patience, it would look rather like Bodø.

So here's my unpopular position, stated plainly. Carrick, Arbeloa and Hayen should all be kept. Judge coaches at 16 months, not four. Judge them on performances, not results. If both are bad at 16 months, by all means, thank them for their efforts. Until then, the clock everyone is watching is set to the wrong time.

The short answerOn opponent-adjusted chances, Millwall are the Championship's fourth-best side, despite being 13th, after the hardest schedule in the league. Wrexham, Norwich, Burnley and Cardiff are also better than their position. Stoke (9th, 23rd on chances), Sheffield United, Swansea, West Brom and Blackburn are flattered.

Millwall are 13th in the Championship, and on the football they have played they are the fourth-best team in it. Once every game is judged against the quality of the opponent, their non-penalty expected goals difference is +0.63 a game, behind only Middlesbrough, West Ham and Wolves. They have also had the hardest set of fixtures in the division. At the other end, Stoke are ninth and Sheffield United 12th, but on chances they are 23rd and 24th. Both have finished almost a goal a game better than their chances.

The table and the football agree more often than not. Middlesbrough are top on both. Southampton are sixth on both, and would be second in the table without their four-point deduction. But across the 24 clubs, the average gap between league position and our adjusted ranking is 4.5 places. Ten clubs are seven or more places away from where their football says they should be.

How we built the real table

We start with non-penalty expected goals, for and against, in every league game this season. Raw numbers can mislead after 11 games, because the fixture list is still lopsided. Some clubs have already played most of the top six; others have barely met them.

So each game is judged against what that opponent normally does. If a side usually concedes 1.2 expected goals a game in its other matches and you created 2.0 against it, you get credit for the difference from the league average. The same goes in defence. The opponent's own game against you is left out, so a club is never measured against itself. The league average is 1.63 non-penalty expected goals per team per game.

We publish the schedule adjustment in full. It is applied at full weight, so the figure in the second chart is exactly what each club gains or loses for its fixtures. The range is narrow, from −0.21 a game for Bristol City to +0.18 for Millwall. It's small next to the gaps between teams, but it is enough to move clubs several places. In the Beaumont Index itself, this season counts for only about 26% of a club's within-league score at this stage, so this table is a sharper read on current form than the Index ranking.

Better than their position

Millwall (13th in the table, 4th on chances). Their raw margin is +0.45 a game, 2.09 expected goals for and 1.63 against. Allowing for the hardest fixtures in the league adds another 0.18. The problem has been converting it. In the 11 games in our chance data they scored 18 and conceded 18, a level goal difference from football that was clearly better than their opponents'. That is about 0.45 of a goal a game below what the chances suggest. Of all the clubs in the bottom half, they are the one the numbers like most.

Wrexham (17th, 8th). A margin of +0.28 after a schedule that has been tougher than most, the 7th-hardest. Wrexham have conceded only 1.29 expected goals a game, the 3rd-lowest figure in the division, but have scored fewer than their chances. Five draws from 12 tell the same story.

It is applied at full weight, so the figure in the second chart is exactly what each club gains or loses for its fixtures.

Norwich (20th, 12th). Four wins and eight defeats, no draws, from football that is almost exactly level once opponents are counted (−0.01). Norwich are a mid-table side on the evidence, not a relegation one.

Burnley (22nd, 14th). One win all season, but the 3rd-hardest fixtures, and their adjusted margin of −0.20 is closer to mid-table than to the bottom three. They have been below average, not desperate.

Cardiff (16th, 9th). Cardiff create a lot, 2.33 expected goals a game, the 2nd-most in the league, and concede a lot, 2.03. Their schedule has been the 3rd-easiest, which takes 0.15 a game off their raw figure, and they still come out ninth.

Flattered by their position

Stoke (9th, 23rd). The biggest gap in the division, 14 places. Stoke have created 1.21 expected goals a game and allowed 1.94, a raw margin of −0.73. Yet in the games in our data they scored 18 and conceded 16. That is 0.91 of a goal a game better than their chances, the largest overperformance in the league. Their schedule has been close to average, so it isn't the fixtures. It is finishing at one end and goalkeeping or wayward shooting at the other, and neither tends to last.

Sheffield United (12th, 24th). Mid-table on points, bottom on chances. They have the lowest expected goals for in the division (1.10 a game) and concede 1.82. A relatively gentle schedule, the 5th-easiest, makes it slightly worse: their adjusted margin is −0.86. Their goal difference is positive; their football is not.

Swansea (2nd, 10th). Second in the table on 23 points, and the obvious name to watch. In the 11 games in our data Swansea conceded only 8 goals from chances worth 15.3. Their attack is about average. Their adjusted margin of +0.13 is that of a play-off contender, not an automatic promotion side.

West Brom (5th, 13th). 21 points from football that has been slightly below average once opponents are counted (−0.19). Like Swansea, West Brom have conceded fewer goals than their chances against suggest.

Blackburn (10th, 17th). Blackburn's raw margin is −0.55 a game; the 5th-hardest schedule softens that to −0.43, but they are still a bottom-half side on chances with top-half points.

Where the table is right

At the top, the table is broadly right. Middlesbrough's adjusted margin of +1.59 is the best in the division by a distance, and West Ham (+0.95) and Wolves (+0.74) are second and third on both measures, give or take a place. QPR, fifth on chances, are eighth in the table. Southampton are the interesting case. Their football is the sixth-best in the league and they are sixth, but only because of the deduction. On the points they have earned on the pitch, 24, they would be second.

At the bottom, Preston and Derby really are struggling, although Preston are unluckier than anyone. They have finished 1.01 of a goal a game worse than their chances, the biggest shortfall in the league, and are 19th on adjusted chances rather than 24th. Charlton (22nd on chances) and Watford (20th) are where their football says they should be: near the bottom.

Six clubs sit in exactly the same place on both measures: Middlesbrough, Southampton, Birmingham, Portsmouth, Lincoln and Bolton. Portsmouth's place is worth a note. They have conceded only 1.18 expected goals a game, the second-lowest in the league, but their schedule has been the second-easiest, which costs them 0.18 a game in the adjustment. That is the schedule effect at work: Portsmouth's raw margin would put them higher, and the fixtures pull them back to 11th. Two clubs, Portsmouth and Wolves, have played one game fewer than the rest in our chance data, so their figures rest on ten games.

What it means for the rest of the season

A gap between results and chances is not a forecast that it will close in a particular week. But over a season, chance difference is a better guide to where a team finishes than its goal difference after 12 games. The overall link between our adjusted table and points per game is real but loose: a rank correlation of 0.55. That's what you would expect at this stage, and it is why a 46-game league can still reshuffle so much.

If the pattern holds, Millwall have the clearest case for climbing, and Wrexham, Norwich and Cardiff should drift upwards. Stoke and Sheffield United need their finishing to stay well above normal to keep their places. Swansea have the points already, so their question is whether they can keep conceding so little from what they allow.

Our verdict

Millwall are the best team outside the top 12 by a distance; Stoke are the most flattered. Middlesbrough are rightly top. Swansea and West Brom are higher than their football, Millwall, Wrexham and Norwich lower. We will update this table as the season goes on.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerDavid Raum has been in the top 4% of Bundesliga full-backs for expected assists in all six seasons since 2021-22. He's a consistent creator, not a one-season spike. He would fit United's 4-2-3-1 naturally. The risk is defensive: his duel numbers are inconsistent. Our verdict: strong fit, moderate risk.

Manchester United want a left-back, and this weekend's reports say they have decided which one. RB Leipzig's captain, David Raum, is said to be the club's top target for January. Reports suggest he is keen on the move, that Juventus are also interested and that Leipzig might take a fee of around £30m, because his contract runs out at the end of June 2027. We can't tell you whether any of that is true. What we can tell you is what kind of player United would be buying, whether his numbers are a spike or a habit, and how he'd fit the way United play.

The short version is that Raum is one of the most consistent chance creators among full-backs in European football. The longer version has a catch, and it's on the defensive side.

Six seasons, one habit

Start with the number that matters most for an attacking full-back: expected assists, the quality of the chances a player sets up regardless of whether team-mates finish them. We've ranked Raum against every Bundesliga full-back with enough minutes in each of the last six seasons, his last at Hoffenheim and five at Leipzig. His percentiles read 99, 97, 96, 97, 98 and, in this season's small sample, 99. He has been in the top 4% of the league's full-backs for expected assists every season since 2021-22.

That's the most important fact in this report, because it answers the question every transfer should start with. Is this a spike or a habit? It's a habit. Plenty of full-backs have one season where crosses land on heads and the assist count jumps. Raum has done it at two clubs, under several coaches, in seasons where his team was good and seasons where it was merely fine. Across his five full seasons he has 33 league assists from chances worth about 46. That's fewer than the chances deserved, so team-mates' finishing has, if anything, undersold him. Luck isn't flattering this record.

He's not just a crosser, either. Last season he was in the 98th percentile among Bundesliga full-backs for progressive passes, the passes that move the ball meaningfully towards goal, and the 98th for shot assists. His cross accuracy jumped to the 91st percentile, from the 13th to 24th range in his first three Leipzig seasons. That's the most interesting change in his profile: the volume was always there, and last season the delivery got cleaner.

This season the sample is small: six league games and 549 minutes, which is essentially every minute Leipzig have played. But the shape is the same. He is creating 3.44 shot assists per 90, the best rate of any Bundesliga full-back with 270 minutes, and is again in the 99th percentile for progressive passes. He hasn't got an assist yet. The chances say that won't last.

The catch: duels

Now the less flattering part. Raum's defending numbers move around a lot more than his creating numbers. His percentile for defensive duels won has gone 50, 75, 41, 58 and 17 over five full seasons, and is 26 this season. In the air he has never been above the 80th percentile and has usually sat around the middle of the pack or below.

His cross accuracy jumped to the 91st percentile, from the 13th to 24th range in his first three Leipzig seasons.

There's a nuance. Last season he made more successful defensive actions than most full-backs and was in the 94th percentile for interceptions, so he isn't passive. He reads the game and steps in. What he doesn't always do is win the one-on-one when a winger runs at him. In the Bundesliga that's a manageable weakness. In the Premier League, where the right wingers are quicker and come at you every week, it would be tested.

How he'd fit United

United have used a 4-2-3-1 in six of their seven league games this season. That matters, because it's the shape Raum knows best. Leipzig set up in a 4-3-3 in 21 of the 27 league games from last season in our match data, with Raum as the overlapping full-back outside a narrow left winger. In United's 4-2-3-1, the left-back is asked to provide width while the left-sided attacker comes inside. It's essentially the job he already does.

The bigger question is who he would replace. Luke Shaw has played 352 league minutes this season. Patrick Dorgu, with 380, is listed in our data as much as a winger as a full-back, and Noussair Mazraoui, with 404, is really a right-back who can cover the left. Against Premier League full-backs this season, Shaw is in the 87th percentile for expected assists, which is good. Dorgu is in the 95th but the 16th for progressive passes: he creates by running, not passing. Mazraoui is in the 14th for expected assists. He is a safe pair of hands, not a creator.

Be careful with that chart. Raum's percentiles are against Bundesliga full-backs. United's are against Premier League full-backs, and our model rates the average Premier League club about 145 rating points stronger than the average Bundesliga club. A 98th-percentile creator in Germany is not automatically a 98th-percentile creator in England. But the profile is the point. Raum combines Dorgu's chance creation with passing range that none of United's options has shown this season.

There's also the team context. United are 4th in the Premier League on our ratings and 8th in the world, at 1671, but they've taken 1.00 point a game from performances worth 1.67. That's 4.7 points below what they've deserved, which says the underlying team is better than the table. A left-back who turns pressure into chances is exactly the kind of signing that helps a team like that convert good performances into results.

What he'd bring, in plain terms

Three things. First, a reliable supply of chances from the left: crosses, cut-backs and progressive passes, every week, for six years running. Second, experience. He's 28, Leipzig's captain and a Germany international, so this wouldn't be a project. Third, a left foot on the left side, which United have often lacked when Shaw has been out.

And three risks. First, the duel numbers, which suggest a better right winger can go past him. Second, the jump from a league our model rates weaker to the toughest league on our scale. Third, age: 28 is the peak for a full-back, but the deal makes most sense as a three-season signing, not a long-term one.

Our verdict

Fit: strong. Risk: moderate. On the creative side, Raum is as close to a sure thing as full-backs get. Six seasons in the Bundesliga's top 4% for expected assists is not a fluke, and the system fit with United's 4-2-3-1 is natural. The risk is defensive. United would need to protect him, with a left-sided centre-back who covers the channel and a midfielder who shifts across, or accept that some opponents will target him. If the reported price is anywhere near accurate for a player with six months left on his contract, it's the kind of deal that makes sense: an experienced, consistent creator for a problem position, without a big long-term commitment.

What would change our mind? Two things to watch before January. If his duel percentages stay in the bottom third over a bigger sample, the defensive concern hardens. And if Leipzig's creating dries up around him, we'd want to know whether it's him or them. For now, the numbers say United have picked a sensible target.

The short answerEight second-tier players stand out this October: Ture Spendler (Norrby), Luciano Viano (Racing de Córdoba), Kaili Shimbo (Yokohama FC), Ilia Ivanschitz (Liefering), Pyry Hannola (Haugesund), Rio Robbemond (Jong AZ), Ahmad Al-Qaq (Sporting JAX) and Lee Dae-gwang (Paju Frontier). Each is top 6% or better for his league and position on one headline strength; past seasons separate steady from spike.

Every scouting list you have ever read has the same problem. It's a list of players you've already heard of, written by people who heard of them at roughly the same time you did. So here's a different kind of list. None of these eight plays in a top division. Between them they play in eight countries, on four continents, in leagues that are lucky to get a highlights package. And each of them is doing one thing better than almost anybody else in his position, in his league, this season.

That last phrase matters, so let's be precise about it. We compare each player only with players in the same position group in the same league, and only with those who have played at least 270 league minutes. "Best of 51" means he ranks first of 51. "Top 3%" means he ranks above 97% of them. We aren't saying any of these players is better than a Premier League regular. We're saying that, in the place where he actually plays, he stands out on something specific, and that standing out on something specific is how most careers begin.

One more thing. A single season can lie, so for every player we've gone back through his previous seasons, at whatever level he played, and asked the obvious question: is this the real level, or a spike? Each section ends with a verdict.

A word on what you won't find here: no transfer fees, no price tags, no "he'll be at a big club by summer". The Index profiles performance. It doesn't sell players.

The eight at a glance

Ages as of 24 October 2026. Minutes are league minutes; where public sites differ from our count we show the range. Clubs checked 24 October 2026. Percentile ranks for Shimbo, Al-Qaq and Lee run to 5 October, even where minutes are more recent.

Ture Spendler, Norrby: the quarterback at left-back

Ture Spendler plays left-back for Norrby in Sweden's Superettan, and on the evidence of this season he is the most ambitious passer in his position in the league. Of 51 Superettan full-backs, he ranks first for long passes and first for progressive passes, the ones that move the ball meaningfully towards goal. He's in the top 3% for passes into the final third as well.

You'd expect a full-back like that to be a liability without the ball. Spendler isn't: he's in the top 4% for successful defensive actions. That combination is rarer than it sounds. Most full-backs who play like this are wingers who have been asked to stand further back. Spendler looks more like a midfielder who has been told the left touchline is his.

He's in the top 3% for passes into the final third as well.

His sample is the biggest here, 2,340 league minutes across 26 games, every one of them a start, so this is a season-long pattern, not a hot month. In the summer, Swedish press reported that a supposed approach from abroad for him turned out to be a scam. Somebody, at least, had noticed.

The history makes his case stronger, not weaker. In Sweden's third tier in 2024 he ranked in the 85th percentile for long passes and the 66th for progressive passes. In 2025, still in the third tier, those were 99 and 94. This year, a tier higher, both are 100. His defensive numbers have climbed the same staircase: 24, then 66, then 96. Three seasons, three steps up, and the latest against better opponents. Verdict: improving, and the real level.

Luciano Viano, Racing de Córdoba: a winger in disguise

If Spendler is a midfielder at left-back, Luciano Viano is a winger at left-back, and he hasn't forgotten it. On loan at Racing de Córdoba from Talleres, Viano ranks first of 150 Primera Nacional full-backs for progressive carries. He's in the top 1% for dribbles attempted and for accelerations, and the top 4% for crosses and for passes that lead to a shot.

A pool of 150 full-backs is a big one, which makes "best of 150" one of the more striking numbers on this page. Argentina's second tier is a famously physical, famously unforgiving league for players who like the ball at their feet. Viano's answer appears to be to run with it until somebody stops him, and not many people have.

The caveat is minutes. Our count has him at 1,550; the public sites we checked have about 1,420. Either way it's a solid half-season, across 28 games, many of them from the bench.

Here the history is mostly an empty page. Before this season Viano had 206 league minutes at this level, as a winger for Arsenal de Sarandí in 2024, far too few to rank. There's nothing to compare 2026 with. Verdict: a breakout season, and a striking one, but no track record yet, so not an established level.

Kaili Shimbo, Yokohama FC: the wing-back who keeps making chances

Kaili Shimbo's headline this season looks like tidiness. Among 49 J2 full-backs, he's in the top 3% for overall pass accuracy, for long-pass accuracy and for progressive-pass accuracy. Being accurate is easy if you only pass sideways. Being accurate when you're trying to go forward, and long, is not.

But his history says lead with something else. He's in the top 6% for expected assists, our estimate of how good the chances he creates are, and that number has followed him everywhere: the 96th percentile in J3 in 2023, the 94th in J2 in 2024 and the 71st in J1 last season, against the best full-backs in the country. The accuracy is the newer story. In those three seasons his passing-accuracy ranks sat between 30 and 80. Shimbo was also named in the league's best XI for the regional round of the special spring season, so the people who watch J2 every week clearly rate him.

He's now on 1,006 league minutes from 12 games, to 24 October, but the percentiles above still stop at 5 October. Verdict: mixed. The chance creation is the steady trait, three seasons deep and across three divisions; the near-perfect passing is a spike that may settle back.

Ilia Ivanschitz, Liefering: 19, and already in charge

Here's the most eye-catching profile on the page. Ilia Ivanschitz is 19, plays central or attacking midfield for Liefering in Austria's 2. Liga, and ranks first of 34 central midfielders in that league for key passes and for through balls. He's in the top 3% for passes leading to a shot. And, slightly absurdly for a creator, he also ranks first of 34 for the share of defensive duels he wins.

The surname will ring a bell. He is the son of Andreas Ivanschitz, the former Austria international, and the Austrian press reports he's Liefering's vice-captain this season. Inherited talent is a lazy story, though, and the numbers don't need it. A teenager leading a league's midfielders for creating chances, while also winning his tackles, would be interesting whoever his father was.

The warning is sample size: 787 league minutes, nine games. Nine games of anything is a mood, not a season. But it's quite a mood.

Last season tells you this didn't come from nowhere. In 1,280 league minutes in 2025-26, listed as an attacking midfielder, he was already in the 92nd percentile for through balls and the 82nd for passes leading to a shot. His key passes have jumped from the 67th percentile to first. The new part is the defending: his defensive-duel win rate went from the 56th percentile to first, on fewer than 800 minutes. Verdict: improving. Enjoy the duels number, but don't build anything on it yet.

Pyry Hannola, Haugesund: the man you can't knock off it

Pyry Hannola is a Finnish central midfielder at Haugesund, back in Norway's second tier this season, and his headline is press resistance. Among 49 OBOS-ligaen central midfielders he's in the top 3% for dribble success: when he takes a man on, he usually keeps the ball. He's in the top 5% for passes leading to a shot and for expected assists, and the top 7% for accuracy in the final third.

Two honest asterisks. He turned 25 on 21 October, so he's the oldest player here, if only by three days. And public match logs show he's mostly come off the bench in recent weeks, which helps explain why public minute counts for him (about 1,620 to 1,680) sit below ours (1,858). If you're looking for one player whose role, rather than his form, is the question, it's Hannola.

If you want consistency, this is it. In every Nordic season we have for him, his rank for passes leading to a shot has been high: 91 in Finland in 2023, 94 in 2024, 82 in Norway's top flight in 2025 and 95 now. His dribble success was high in 2023, 2025 and 2026, and dipped only in 2024, when he was used further forward. The one season that breaks the pattern is a short loan in Poland, 650 minutes, where his creative ranks fell to 10 and 16. Verdict: steady. The open question is his role, not his level.

Rio Robbemond, Jong AZ: the No. 6 who scores

Defensive midfielders aren't supposed to score. Rio Robbemond, 21, hasn't read that memo. Among 36 Eerste Divisie defensive midfielders he's in the top 3% for non-penalty goals per 90 minutes. He's also in the top 5% for through balls, and in the top 9% for both aerial duels won and dribbles attempted.

That's a strange and rather lovely mix: a holding player who wins headers, runs with the ball, threads passes and arrives in the box. Reserve sides in the Dutch second tier are a well-worn path to AZ's first team, and Robbemond's profile is a good argument for why that system works. Like Ivanschitz, though, his sample is small: 746 minutes across ten games. Goals from midfield are the kind of number that comes and goes.

His history splits the profile in two. The goals are not new: in 2024-25 he ranked in the 98th percentile for non-penalty goals, then the 53rd last season, and 97 now. That's a real goal threat, if a streaky one. The rest is new. His through balls ranked 33 and 39 in the last two seasons and his aerial duels 18 and 23, when he was listed as an attacking midfielder; now they're 95 and 91, since the move to No. 6. Verdict: part spike. The goals are real; the No. 6 traits rest on 746 minutes.

Ahmad Al-Qaq, Sporting JAX: a striker who plays like a midfielder

Ahmad Al-Qaq plays for Sporting JAX, a club in its first season in the USL Championship, and he's a centre-forward who seems to prefer the job description of a No. 8. Among 72 USL Championship strikers he's in the top 3% for progressive passes, and in the top 7% for progressive carries, passes into the final third and dribbles attempted.

Strikers who drop deep and move the ball are useful in a particular way: they drag centre-backs out of position and give the runners behind them somewhere to go. The open question with any profile like this is whether he does enough in the box as well, and these strengths don't answer it. That's not a criticism. It's what makes him interesting to watch. He's now on 1,302 league minutes from 22 games, to 21 October, though his percentiles stop at 5 October.

This one is reassuringly dull. Last season at North Carolina FC, in the same league, he was listed as a full-back, but compared with strikers he still ranked 93, 78, 93 and 89 on the same four strengths. Now it's 97, 93, 93 and 93. Verdict: steady. Same player, same habits, a slightly better season.

Lee Dae-gwang, Paju Frontier: the forward who has suddenly started defending

And finally, the season on this list that most needs a second look. Lee Dae-gwang is a forward at Paju Frontier in K League 2. Among 44 K League 2 strikers he's in the top 3% for successful defensive actions, and in the top 5% for interceptions, adjusted for how much of the ball his side has, and for duels contested.

Nobody puts "won the ball back on the halfway line" on a highlights reel. Coaches put it on a whiteboard. A forward who defends like this lets a team press without leaving gaps behind.

Here's the twist. In Korea's third tier, in 2023 and 2024, he ranked between the 11th and 50th percentiles on exactly these measures. A tier higher, he's at 95 to 97. Players don't usually get more defensive by moving up a level, which makes this a new role, a new coach's demands or a very good few months. We can't yet tell which. He's on 1,134 league minutes from 19 games, to 17 October, and hasn't featured in his side's last two league games. Verdict: a spike. A strong season, not yet a proven trait.

How much should you trust this?

Honestly, it depends on the player, and the history sorts them more clearly than one season could. Spendler, Hannola and Al-Qaq are steady or improving across several seasons. Ivanschitz is improving, with one new trait. Shimbo is half steady, half spike. Robbemond's goals have form, but his No. 6 traits don't yet. Viano has no history to test, and Lee's history argues against his current numbers. On sample size, Ivanschitz and Robbemond have fewer than 800 league minutes this season, and three profiles have percentiles that stop at 5 October.

Percentiles also flatter a player in a small pool and are harsh in a big one. "Best of 34" in Austria and "best of 150" in Argentina are both firsts, but the second took beating far more people. And a strength is only a strength. None of these profiles tells you a player's weaknesses, and every one of them has some.

The verdict

If we had to pick one to follow every week, it would still be Ivanschitz: 19, already strong at creating last season and now leading his league's midfielders in the two hardest creative numbers. If we had to pick the safest bet that this is real, it's Spendler, more clearly than before: three seasons, three steps up, the last of them a tier higher. Hannola and Al-Qaq are the quiet winners of the history test, and Shimbo turns out to be a creator first and a metronome second. Lee is the one to watch with the most scepticism, which, on a list of players nobody is watching, is saying something.

We'll rerun all eight in January. Some will have kept it up, some won't, and we'll tell you which. That's the deal with lists of players nobody is watching. Someone has to.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerThe Eliteserien is clearly weaker than the Premier League. On The Beaumont Index, the median Eliteserien club is rated 1255 and the median Premier League club 1597. But the gap is uneven: Bodø/Glimt, rated 1566, would sit around 18th in England, while most Norwegian clubs would rate below every Premier League side.

The answer in one sentence

The Premier League is much stronger overall, but Norway's best club is genuinely Premier League standard, and the gap only becomes huge once you leave the top two.

How we measure it

The Beaumont Index puts every club on one scale, adjusted for the strength of the league it plays in. That adjustment is the important bit. Winning every week in a weaker league doesn't earn a high rating on its own; it has to be weighed against who you beat. The ratings use an Elo-style scale, where a gap of about 180 points is roughly a goal a game.

The headline numbers

• Median club: Premier League 1597, Eliteserien 1255. That gap of roughly 340 points is close to two goals a game between two ordinary sides.
• Best club: Arsenal, 1780. Bodø/Glimt, 1566.
• Weakest club: In the Premier League, 1438. In the Eliteserien, 1149.
• Overlap: Two Norwegian clubs rate above the weakest Premier League side: Bodø/Glimt and Viking.

What 340 points looks like on a pitch

On our scale, 180 points is roughly a goal a game. So a gap of about 340 points between the median clubs means an ordinary Premier League side would be expected to beat an ordinary Eliteserien side by close to two goals, on neutral ground, over a long run of games. That's a big gap. It's bigger than the gap between the Premier League and the Championship medians, which is about 286 points.

Could Bodø/Glimt survive in the Premier League?

On rating alone, probably. Seventeen Premier League clubs rate higher, which would put them around 18th: a relegation fight, but a winnable one. Viking, rated 1441, would sit 20th, above one Premier League club. After that the drop is steep. Norway's third-best side, Molde, rates 1348. That's below every Premier League club and in line with the middle of the Championship.

Why you can't compare points across leagues

A common mistake is to compare league tables directly. Bodø/Glimt's performances in our run were worth about 2.49 points a game, a higher figure than Arsenal's 1.96. That doesn't make Glimt better than Arsenal. Expected points measure how a team performs against the opponents it actually plays, and Glimt play Eliteserien opponents every week. The same is true at the top of every league: Bayern München's figure is 2.74, Barcelona's 2.63, Middlesbrough's 2.29. A rating on one scale, adjusted for league strength, is what makes the comparison possible at all.

A very top-heavy league

The Eliteserien is unusually top-heavy. Between its best club and its third-best there are 207 rating points, more than a goal a game. In the Premier League the same gap is 77 points; in the Championship, 86. Glimt sit 114 points clear of Viking, and Viking a further 93 clear of Molde. None of our other leagues has a gap that big between first and third. It also explains the title race. Glimt and Viking were level on points per game in our run, 2.43 each, but Viking were the luckiest team in all our leagues, about ten points ahead of their performances, while Glimt were about one point behind theirs.

Who Glimt's neighbours would be

On rating, Glimt would sit between Tottenham Hotspur, rated 1566, and Coventry City, 1538. It's worth looking at how those clubs are actually playing. To 18 October, Tottenham had created chances worth about 1.3 goals a game and conceded about 1.7. Coventry had created about 1.1 and conceded about 1.8. Hull City, the lowest-rated Premier League side, had created about 1.1 and conceded about 1.75. That's the company Glimt would keep: teams that concede more than they create and spend the season looking over their shoulder. It's a fight, not a free ride.

Bodø/Glimt's performances in our run were worth about 2.49 points a game, a higher figure than Arsenal's 1.96.

Where Norway's famous names sit

Brann, rated 1339, and Rosenborg, 1329, are close to Molde and close to the Championship median. Both would be competitive in England's second tier on rating; neither would rate near the Premier League. The rest of the league thins out from there, down to Kristiansund at 1149, the lowest-rated club in any of our eight leagues.

Where Norway sits among Europe's leagues

Here are the median club ratings for all eight launch leagues: Premier League about 1621, Ligue 1 about 1467, LaLiga 1465, Bundesliga about 1437, Serie A about 1426, Championship about 1335, Eredivisie 1293 and Eliteserien about 1253. Among our eight launch leagues, the Eliteserien has the lowest median rating, just below the Eredivisie (1293) and the Championship (about 1335). But it also has the widest range, 406 points from top to bottom against 217 in the Championship, because Bodø/Glimt and Viking pull so far clear. In other words, it's a league with two clubs of a much higher standard and a long middle. Bodø/Glimt's 1555 is higher than the best club in the Eredivisie, rated 1485, even though the Eredivisie's median club is stronger.

Eliteserien v Championship

This is the comparison that surprises people. On the median, the Championship is stronger, about 1339 against 1255. But Bodø/Glimt, at 1555, rate higher than every Championship club, including West Ham United at 1507. Viking, at 1441, would sit about third, behind West Ham and Middlesbrough. Below that, the two leagues overlap heavily: Molde, Brann and Rosenborg would all be mid-table Championship sides on rating, and the bottom half of the Eliteserien would rate below every Championship club.

Glimt in European company

Rated 1566, Bodø/Glimt sit above Real Betis (1532), Monaco (1507) and Athletic Club (1504) on our scale. Viking, at 84th, sit near Bologna (75th) and Venezia (96th). On this scale, Norway's best club belongs in the same conversation as good sides from Spain and France, even though the league as a whole sits 22nd in our world league ranking.

Why this matters

For fans, this is what makes Norway's title race so watchable: two of Europe's better teams separated by a goal of goal difference. For anyone watching players, it explains why a top performer at Bodø/Glimt is playing at a higher level than the league's reputation suggests, while a top performer elsewhere in Norway may be playing at a level closer to the English second tier.

How to use these numbers

Three rules of thumb. First, compare clubs, not leagues, when the question is about a specific team: the league median tells you about the typical opponent, not about Bodø/Glimt. Second, remember the scale. A 100-point gap is a little over half a goal a game, which is noticeable but beatable on the day. A 340-point gap is not. Third, read the trend as well as the number. Bodø/Glimt, Molde and Rosenborg are all rated as rising in this run; Lillestrøm, Fredrikstad and several others are rated as falling. A rating is a snapshot of a moving thing.

Caveats

Ratings carry uncertainty. Below Bodø/Glimt the ratings bunch together: Viking are rated 1442, Brann 1348, Molde 1346 and Rosenborg 1342, so small differences in form can reshuffle Norway's middle quickly. That is why, when we compare Norway's middle with England's, we lean on the ratings themselves rather than on positions. And this run uses results to 25 October 2026. We refresh it weekly.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerInter are rated the best team in Serie A and 15th in our world ranking, so they start as clear favourites at San Siro. Fiorentina are 18th but rated 13th: their chances deserved about 9 points, not four, and they finished below their numbers last season too.

There is a version of this fixture that looks like a mismatch, and the table tells it. Inter are top of Serie A on 19 points from seven games, level with Lazio and ahead on goal difference. Fiorentina are 18th with four points, one win, one draw and five defeats. Sunday lunchtime at San Siro, kick-off 11:30 UK time, Maurizio Mariani refereeing. Top against third-bottom.

The Beaumont Index agrees with half of that. Inter are rated 1625, the best side in Italy by some distance and 15th in the Index world ranking of 3,848 clubs. The surprise is Fiorentina. On rating they are 13th of 20, not 18th. That is no great compliment, but it is a long way from relegation form. The gap between those two numbers is the story of their season, and it is the part of this game worth watching.

Fiorentina: a bad season, made to look worse

Start with the number. Paolo Vanoli's side have created chances worth about 8.9 points in seven league games. They have four. That shortfall of 4.9 points is the fourth-biggest in Serie A, behind Monza, Bologna and Juventus. And this is not new. Last season they finished 6.4 points below what their chances deserved. One bad-luck season is noise. Two in a row makes you ask whether something else is going on: finishing, goalkeeping, or the habit of conceding the one chance that matters.

The individual games make the case better than the totals. Fiorentina lost 3-0 at home to Frosinone while creating 2.95 expected goals to 1.37. They lost 2-1 at home to Torino with 1.67 to 0.83. Last weekend they lost 3-1 at home to Como, a game in which they created 2.66, though Como created more. Over the season they have out-created their opponent in three of seven games and won only one of those, the 4-2 at Venezia.

They are not blameless. The 4-0 defeat at Roma was exactly what it looked like: 0.40 expected goals to 3.13. And the shape keeps changing. The match data record three different shapes in seven league games: a 4-1-4-1, a 4-3-3 and the 4-2-3-1 that started against Como. That is a coach still looking for a team. Against Como it was David de Gea in goal; Pietro Pongračić and Radu Drăgușin in the middle of defence; Nicolò Fagioli and Cher Ndour in front of them; Franco Mastantuono, Pote and Njie behind Pellegrino up front.

Over the whole season Fiorentina are averaging 1.90 expected goals a game and conceding 1.84. Last season those figures were 1.43 and 1.50. They create a little more than they used to and give away a little more too. That is the profile of a mid-table side having a miserable run, not of a side that deserves to be 18th.

Inter: the same shape, more of everything

Cristian Chivu's Inter are the opposite of all that uncertainty. The match data record a 3-5-2 in every one of their seven league games. The names change a little. The shape does not. The side that won 3-1 at Bologna last Saturday read: Josep Martínez; Yann Bisseck, Manuel Akanji, Alessandro Bastoni; Luis Henrique, Nicolò Barella, Piotr Zieliński, Petar Sučić, Federico Dimarco; Marcus Thuram and Francesco Pio Esposito. Thuram, Carlos Augusto and Bonny scored.

The 4-0 defeat at Roma was exactly what it looked like: 0.40 expected goals to 3.13.

What has changed is volume. Inter are creating 3.22 expected goals a game in the league this season, up from 2.04 last season. That is a huge jump, and you can see it game by game: 4.25 in a 1-0 win at Cagliari, 3.51 at Bologna, a 5-3 against Udinese. They have scored 21 league goals, three a game.

The cost is at the other end. Inter have conceded nine league goals and are giving up 1.4 expected goals a game, up from 0.99 last season. Napoli created 2.63 at San Siro and still lost 3-2. The only points Inter have dropped were in a 2-2 at Roma. In Europe the same openness cost them: a 4-3 home defeat by Club Brugge on 13 October. On Wednesday they tightened up and beat Shakhtar 1-0.

Inter also tend to finish ahead of their numbers. They are +3.4 points above what their chances deserved this season and were +4.3 last season. Two seasons running is a pattern, not luck, and the obvious explanation is up front. Thuram and Lautaro Martínez have four league goals each, and Inter have strikers who turn half-chances into goals.

The history: Inter's fixture, with one bad memory

Inter have won four of the last six league meetings, drawn one and lost one. The most recent was a 1-1 in Florence in March. Before that, Inter won 3-0 at San Siro last October. The defeat came in February 2025, a 3-0 in Florence that still sits oddly in this record, followed four days later by a 2-1 Inter win in Milan. Go back to September 2023 and it was 4-0 to Inter at San Siro.

Head-to-head records are a weak guide in football. Players change, coaches change and six games is a tiny sample. But they do tell you what each club expects walking out, and Inter expect to win this one.

Where this is decided

Look at the flanks. Inter's 3-5-2 gets much of its width from Dimarco on the left and Luis Henrique on the right, and Fiorentina's full-backs have to decide whether to step out to them or hold the line. Inter's two strikers also make sure a back four always has two bodies to worry about. If Fiorentina line up in a back four again, they must keep Thuram and Esposito apart, or Lautaro if he starts, without leaving Barella and Sučić room to run beyond them.

For Fiorentina the route is the one that has nearly worked all season: get the ball to Mastantuono and the runners behind him, and make the chances count. They do create. The question all season has been the last touch, at both ends. Inter's defence has looked beatable lately. Brugge scored four, Napoli two, Udinese three. This is not the San Siro fortress of other seasons.

There is fatigue to think about too. Inter played on Wednesday night in Europe and kick off before lunchtime on Sunday. Chivu has rotated within the same shape all season, and that fixed system makes rotation cheaper: a fresh wing-back slots into a role he already knows.

The verdict

In rating terms the gap is 208 points, worth a little over a goal a game on neutral ground, and this game is not on neutral ground. By the Index Inter are clearly the better team, at home, against a side that keeps losing. Nobody needs a model to say that.

What the model adds is a warning against reading Fiorentina as a relegation side. Their underlying numbers are those of a team in the bottom half of the middle, unlucky for two seasons running, playing in an unsettled shape. Teams like that do not usually stay 18th. If the turn comes, it will come in games where they create enough and finally take it. Sunday is a hard place to start, but the numbers say Fiorentina are better than their four points.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerThe luckiest team in our eight launch leagues is Viking, about ten points ahead of what their performances deserved, with Lazio close behind. The unluckiest is Nottingham Forest, nearly nine points short. Manchester City have dropped out of the lucky ten after losing at Aston Villa.

Luck is football's least popular explanation, and probably its most underrated one. Nobody wants to hear that their side is top because the ball has bounced kindly, or bottom for the opposite reason. So we have done the decent thing again and put it all in a list, now updated with every result to 25 October 2026.

The method is unchanged. For every club in our eight launch leagues, the Index compares the points they have actually won with the points their performances would normally earn. The bigger the gap, the more luck, good or bad, is involved. One health warning: the leagues are at different stages. Norway is 23 or 24 games in, the Championship 12, and the Bundesliga only 6 or 7. Small samples swing more.

Most clubs are not on these lists. Of the 154 clubs we measured, 83, more than half, are within three points of what their performances deserved. The twenty below are the outliers.

What has changed

Viking still top the lucky list, about 10 points ahead of their performances, but the margin has shrunk. Lazio are now second, only about a point behind, and Stoke City third. At the other end, Nottingham Forest remain the unluckiest club we measure, nearly nine points short of what their football has earned.

The biggest change is who has dropped off. Manchester City were on the lucky list last time. Since then they have lost 2-1 at Aston Villa, and their surplus has fallen to +5.7, now 13th among the 154 clubs. That is what luck catching up looks like: City are still rated 1733, with the second-best expected points rate in the Premier League, but the results are coming back towards the performances. Swansea City, who led the Championship on points per game, have dropped out too. Atlético Madrid and Blackburn Rovers come in. On the unlucky side, Cardiff City drop off and Willem II arrive.

The lucky ten

1. Viking (Eliteserien), +10.3. Level with Bodø/Glimt on 56 points at the top of Norway, with a game in hand, but their football has been worth 1.99 points a game against an actual 2.44. Still the most fortunate club we measure, though Lazio have closed to within about a point.
2. Lazio (Serie A), +9.2. Top of Serie A on points per game (2.75), 9th on expected points (1.60). Rated as rising, which softens it a little, but this is the biggest gap in Italy.
3. Stoke City (Championship), +8.3. 1.67 points per game from football worth 0.98, the 21st-best expected rate in the Championship. The Index has them 18th of 24.
4. Fortuna Sittard (Eredivisie), +8.0. 1.67 points per game from football worth 0.78. Still on this list from our first run.
5. Lillestrøm (Eliteserien), +7.1. 1.56 points a game from football worth 1.26, and rated as falling. One of three Norwegian entries.
6. Atlético Madrid (LaLiga), +6.7. New to the list. 2.50 points a game, second-best in Spain, from football worth 1.83. Rated 1602, the best team on either list.
7. Real Betis (LaLiga), +6.4. 2.11 points per game from performances worth 1.40.
8. Fredrikstad (Eliteserien), +6.2. 1.26 points per game from football worth 0.99, and falling. The third Norwegian club.
9. Lecce (Serie A), +6.0. The lowest-rated side in Serie A on the Index, with the 20th-best expected rate (0.43), yet 1.29 points a game.
10. Blackburn Rovers (Championship), +5.9. New to the list. 1.50 points a game from football worth 1.01, the 20th-best expected rate in the Championship.

The unlucky ten

1. Nottingham Forest (Premier League), −8.7. 0.75 points per game from football worth 1.83, the 4th-best expected rate in the Premier League. Rated 1583 and rising.
2. Preston North End (Championship), −8.1. Bottom of the Championship on points per game (0.42), with football worth 1.09. Unlucky, but also rated as falling.
3. Brann (Eliteserien), −7.8. 1.26 points a game from football worth 1.60, the 5th-best rate in Norway. Norway gives with one hand and takes with the other.
4. Willem II (Eredivisie), −7.5. New to the list. Bottom of the Eredivisie on 0.20 points a game, from football worth 0.95. Last season they were nearly 14 points ahead of their performances.
5. Athletic Club (LaLiga), −7.2. The 3rd-best expected rate in LaLiga (2.03) and 1.12 points a game to show for it. About 12 points short last season too.
6. Villarreal (LaLiga), −7.0. 4th in Spain on the Index, 9th on points per game. Performances worth 1.90, results worth 1.20.
7. Monza (Serie A), −6.9. 0.88 points per game from performances worth 1.74, and rated as rising.
8. Start (Eliteserien), −6.8. 0.70 points a game, bottom of Norway, from football worth 0.99. Not good, but not this bad.
9. Sunderland (Premier League), −6.8. Bottom of the Premier League on points per game (0.57), 9th on expected points. Rated 1563 on the Index.
10. Brest (Ligue 1), −6.6. The 2nd-best expected rate in Ligue 1 (1.95), 1.12 points a game. Rated as rising.

Viking and the Norwegian title race

Viking's case is the most interesting on the list, because it is a title race. They are level with Bodø/Glimt on 56 points at the top of the Eliteserien, with a game in hand after Glimt lost 1-0 at home to Vålerenga on Saturday. Their goal difference is the same. But on our numbers, Viking's football has been worth about 1.99 points a game, well short of the 2.44 they are taking. They were also well ahead of their performances last season. Luck does not usually last two years, so this is a club to watch closely in the run-in.

Good teams on the wrong list

The unlucky list is not a list of bad teams. Four of the ten are rated above 1500, the level of an average club from Europe's top five leagues: Nottingham Forest (1583), Athletic Club (1504), Villarreal (1545) and Sunderland (1563). Forest are the starkest case. Their performances have been worth more points per game than all but three Premier League sides, and they have collected fewer than all but two. That is not a team playing badly. It is a team losing games it is playing well enough to win, and the Index rates them as rising.

Since then they have lost 2-1 at Aston Villa, and their surplus has fallen to +5.7, now 13th among the 154 clubs.

Villarreal are the same story in Spain. On the Index they are the 4th-best side in LaLiga, behind only Barcelona, Real Madrid and Atlético Madrid, with the 4th-best expected points. Their results are 9th-best. A gap of about seven points in ten games almost always narrows. Athletic Club are the third Spanish case, and the unluckiest club in LaLiga for the second season running.

On the lucky side, the best team on either list is now Atlético Madrid, rated 1602. Their luck is real, but like City's before it, it is luck on top of a very good team.

Which league is luckiest?

Serie A, still. Across the division the average gap between results and performances, in either direction, is about 4.3 points per club after only seven or eight games. The Eliteserien is next at about 3.8, but Norway has played three times as many games, so its gaps have had far longer to build. Game for game, Italy is the noisiest league we measure. The calmest is the Bundesliga, at about 2.1 per club, partly because it is only six or seven games in.

Near misses

On the lucky side, just outside the ten: Real Sociedad (+5.8), Frosinone (+5.7), Manchester City (+5.7), Lille (+5.5) and Arsenal (+5.3). Both Manchester City and Arsenal are there, so the top two in the Premier League are each about five points ahead of their football. On the unlucky side: Cardiff City (−6.6), Juventus (−5.9), Espanyol (−5.7), Fiorentina (−4.9) and Bournemouth (−4.9).

Does luck carry over?

Here is the most useful number in this piece. We compared every measured club's gap this season with its gap last season. The correlation is about 0.03, essentially zero. Knowing how lucky a club was last season tells you almost nothing about how lucky it will be this season. Sunderland were nearly twelve points ahead last season and are nearly seven behind now. Willem II were nearly fourteen ahead and are now bottom of the Eredivisie, seven and a half short. Luck is not a trait. It is weather.

What it means

A list like this is not a prediction, but it is a hint. Gaps of this size usually close, in both directions, and City have just shown how quickly it can happen. If your club is on the first list, enjoy it and do not book the parade. If your club is on the second, the football is better than it feels, and for Forest, Villarreal and Athletic in particular, the underlying level is high. We will keep rerunning this list and tell you honestly who regressed.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerManchester United and Bournemouth meet at Old Trafford on Sunday as two of the Premier League's four unluckiest sides. United have one win from seven, as have Bournemouth, but their performances deserved far more points. United are rated slightly higher, Bournemouth arrive on a three-game winning run, and United have not beaten them in six league meetings.

On Thursday night at the Vitality Stadium, Bournemouth beat AC Milan 3-0 and the Milan manager, Ruben Amorim, walked off having lost to the team that used to give his old Manchester United side such grief. On Wednesday, in northern Italy, Michael Carrick's United lost 4-1 at Como. Sunday lunchtime at Old Trafford brings the two clubs together, and the European week tells you which way the moods are running. It tells you rather less about the football.

Because here is the twist. On the evidence of seven Premier League games, these are two of the four unluckiest teams in the division. The table has them close to each other, near the wrong end. The underlying numbers have both of them playing like top-half sides. So the question for Sunday is a slightly odd one. Which of two teams that deserve more will finally get it?

Two teams the table has got wrong

Start with the plain record. United have one league win from seven, with four draws and two defeats. Bournemouth also have one win, with three draws and three defeats. That's 1.00 points a game for United and 0.86 for Bournemouth, which is the form of clubs looking over their shoulders, not up the table.

Now the record the football deserved. Our expected-points measure looks at the quality of chances each side created and conceded, and asks what that would normally earn. United's performances have been worth 1.67 points a game. Bournemouth's have been worth 1.55. Over seven games, United are 4.7 points short of what they've earned on the pitch, which means they've lost the equivalent of one and a half wins to finishing, goalkeeping and the bounce of the ball. Bournemouth are 4.9 points short.

Only Nottingham Forest and Sunderland are further behind their performances. That's the company. Both of those clubs have stories of their own; these two have each other.

The ratings agree with the expected points rather than the table. United are rated 1671, the 4th-highest rating in the Premier League. Bournemouth are rated 1651, 7th in the league. On our scale an average club from the top five leagues sits at 1500, and roughly 180 points is worth a goal a game. The gap between Sunday's teams is 20 points, which means about a tenth of a goal a game. That is not much of a gap at all for a home side and an away side so far apart in history.

A regression game

Bill James built a career in baseball on a simple idea: small samples fool the eye, and the numbers are most useful when they tell you so. Seven games is a small sample. Luck in football is not a trait, it's weather, and the weather changes.

Only Nottingham Forest and Sunderland are further behind their performances.

There's a twist inside the twist, though. United were well ahead of their performances last season, by about 6.6 points, and are now well behind. Bournemouth were behind last season too, by about 3.3 points, which means this is the second season running that they have played better than their results. Two seasons is still not a law of nature. But if there is a club in England with a genuine grievance against the dice, it plays in Dorset.

Look at the games one by one and the pattern gets sharper. United created the better chances in three of their seven league games; Bournemouth did so in four. At Fulham, United produced chances worth about three goals and drew 1-1. At home to Brentford, Bournemouth created more than two goals' worth against half a goal and drew 2-2. At Newcastle it was a similar story, and another 2-2.

A caution on Bournemouth's best number. Against Sunderland last weekend they created chances worth 4.3 goals, but Sunderland had a man sent off in the 17th minute. Take that game out and Bournemouth still look like a side that deserves more, just not quite as much more.

How United play now: volume, width and crosses

The Carrick version of United is a busier team than the one it replaced. Compare their league games this season with the previous twelve months. They are taking 19.1 shots a game, up from 15.1, which means roughly a quarter more attempts every week. They are crossing far more, about 19 times a game against 14 before. And they are building more set-up attacks, the patient possessions that end in a shot or an entry into the box.

It's a style that piles up attempts rather than refining them. The best of it came against Ipswich, a 5-2 win built on more than five goals' worth of chances. The worst of it is more recent. In two of their last three league games, United have created less than one goal's worth of chances, at home to Tottenham and away at Leeds. The volume is still there; the quality inside the box has thinned. Then came Como, and a 4-1 defeat that made the week feel worse than the league numbers do.

United have not won in seven matches in all competitions since beating Sabah in the Champions League on 10 September. That run includes two European defeats and a League Cup exit at home to Brighton. Seven games without a win at a club like United is a crisis by reflex. On the underlying numbers it is closer to a drought.

How Bournemouth play now: the same shape, a tighter door

Bournemouth are one of the most predictable sides in England in the best sense. In every league game in our data, going back twelve months, they have lined up in a 4-2-3-1. Rose has kept the shape and changed the emphasis.

The change is at the back. Bournemouth are allowing 8.9 shots a game this season, down from 12.2, which means about a quarter fewer attempts at their goal. The quality of those chances has fallen too, from about 1.70 expected goals a game to 1.30. That makes the 5-1 defeat at Chelsea all the stranger. Chelsea's chances that day were worth less than two goals. Five went in.

Going forward Bournemouth are much as they were: about 17 crosses a game and a steady diet of set-up attacks rather than counters. They average about one counter-attack a game; most of their football is built patiently and worked wide.

And they are rolling. Three straight wins in all competitions, with 8 goals scored and none conceded, the last of them that 3-0 against Milan. The catch is the calendar. Bournemouth played on Thursday night and United on Wednesday, which means Rose's players get one day less to recover before a lunchtime kick-off in Manchester.

The fixture that keeps going wrong for United

Some fixtures carry their own weather, whatever the form book says. This one has a long cloud over Old Trafford. United have not beaten Bournemouth in the league since May 2023. In the six meetings since, there have been four draws and two Bournemouth wins, both of them 3-0 at Old Trafford.

Across those six games Bournemouth have scored 15 goals to United's 9. Last December's 4-4 at Old Trafford was the most chaotic of them, and March's 2-2 on the south coast the most recent. None of that history plays on Sunday. It does explain why nobody at Old Trafford will be treating this as a gentle return from Italy.

Where the game will be decided

Three things to watch, and none of them involves a percentage.

First, United's crossing against Bournemouth's back four. United will have more of the ball at home and, on recent form, will put a lot of it into the box. Bournemouth have been good at limiting the number of shots they face. Something gives. If United turn their crosses into chances of real quality, the luck gap starts to close.

Second, Bournemouth's legs in the last half-hour. A Thursday-to-Sunday turnaround after an emotional European night is a test of squad depth. Rose rotated against Milan, starting Di Gregorio in goal instead of Petrović and switching to a 4-4-2, so some of his first choices will be fresher than the calendar suggests.

Third, the transitions. United conceded four at Como on Wednesday. Bournemouth are not a counter-attacking team by habit, but they have scored freely in their last three games. If United's full-backs push on to feed the crossing game, the space behind them is where Bournemouth will aim.

What the numbers say, and what they don't

The numbers say these are two good sides with bad tables, and that the home side is rated slightly higher. They say Bournemouth have been tighter at the back this season and United busier in attack. They say the luck runs out eventually, for somebody, though seven games cannot tell you when.

They don't say who wins. We don't publish match predictions, and a game between two teams this close on rating is exactly the kind where a deflection decides everything. What we can say is that whoever loses on Sunday will probably deserve better than the table shows, and so will whoever wins.

That is a strange thing to be able to say about a game at Old Trafford. It may also be the most honest sentence in any preview this weekend.

Ratings and world ranks in this piece come from before our 25 October 2026 ranking update. Current figures are on each club's page.

The short answerFeyenoord lead the Eredivisie by four points and have out-created all nine league opponents. But PSV are still the Netherlands' highest-rated club, are at home, and have won the last four league meetings. Our ratings put them about a third of a goal a game better, so this looks close.

Feyenoord arrive in Eindhoven top of the Eredivisie, unbeaten after nine games, five league wins in a row, 31 goals scored. PSV are third, four points behind, with two defeats already. Philips Stadion, Sunday, 13:30 UK time, Serdar Gözübüyük refereeing. On the table, the challenger is the home side.

The Beaumont Index sees it the other way round. PSV are still the highest-rated club in the Netherlands at 1498, 50th in our world ranking. Feyenoord are 1433, third in the league on rating behind Twente and 87th in the world. The two readings are both true, and the gap between them is the most interesting thing about this game.

Feyenoord: dominant, and a little fortunate on top

Begin with what Giovanni van Bronckhorst's side do well, because it is a lot. Measured game by game on expected goals, Feyenoord have out-created every one of their nine league opponents. They average 2.83 expected goals a game and concede 0.86. They won 7-0 at PEC Zwolle, 5-0 against Utrecht and 2-0 against AZ, the side now second. Nacho Ferri has six league goals. Gjivai Zechiël has five and four assists from midfield. Anis Hadj-Moussa has five from the wing.

The two points they dropped came in games they controlled. They drew 2-2 at home with ADO Den Haag after 33 shots and 2.47 expected goals to 0.24. On opening weekend they won 1-0 at Sparta despite 3.53 expected goals. If anything, the shot data say Feyenoord should be further ahead.

So why does the Index's expected-points model, which weighs chances differently, have them +4.5 points ahead of what they deserved, the second-biggest surplus in the league? Partly because some of those wins were closer than the scores. Last Saturday's 4-3 at Telstar was exhilarating, but Telstar had 23 shots. Partly because that model rates PSV and Feyenoord almost level on expected points: 2.11 a game for PSV, 2.06 for Feyenoord. Feyenoord's lead is real. Their underlying edge over PSV is smaller than four points suggests.

PSV: the old habits, and the cracks

Peter Bosz's PSV have taken almost exactly the points their chances earned this season: +0.0. Last season was a different story. They finished 16.2 points above their expected total, a huge surplus. A season like that tends to come back towards normal, and it has.

The good PSV is very good. They won 6-1 at Utrecht and faced one shot all afternoon. They won 3-1 at Ajax. They scored five against Groningen and four against Sparta. They average 2.59 expected goals a game. Guus Til and Sem Mijnans have five league goals each and with Dennis Man alongside them the attacking midfield has goals from several directions.

Partly because some of those wins were closer than the scores.

The bad PSV concedes. They give up 1.40 expected goals a game, well over Feyenoord's figure. At Twente they lost 3-2 having faced 23 shots and created just 1.02 expected goals to 2.40. Last Saturday they lost 2-1 at ADO Den Haag, and one of the goals was an own goal by goalkeeper Matěj Kovář. On opening day they drew 2-2 at home with Fortuna Sittard. Even the 3-1 at Ajax came with 19 shots faced.

Then there is Europe. PSV won 1-0 at Porto on Tuesday, a result that changes the mood of a week. Feyenoord lost 1-0 at Real Betis on Wednesday night, a day later, and travel to Eindhoven with a day's less rest. Their Champions League campaign so far has been all over the place: a 5-1 defeat at Barcelona, then a 3-2 win over Como.

The stakes

The table explains why this matters more than a normal October game. AZ are second on 20 points, PSV third on 19, Ajax fourth on 18. A Feyenoord win would put them seven points clear of PSV before November. A PSV win would cut the gap to a single point and give AZ the chance to go level at the top. A draw would suit Feyenoord and the chasing pack more than it suits PSV. Nine games is not a season, and Dutch title races have turned on less, but a seven-point lead over PSV this early would be a serious statement.

The history: PSV's fixture, and recently so

This is where the home side's confidence comes from. PSV are unbeaten in the last six league meetings, five wins and a draw, and have won the last four. In February PSV won 3-0 at home. Last October they won 3-2 in Rotterdam. They also won 3-2 in Rotterdam in May 2025, and 3-0 in Eindhoven in December 2024. The last time Feyenoord beat PSV in a match decided in normal time was a 1-0 cup win in January 2024.

As always, read head-to-head records carefully. Squads turn over quickly in the Netherlands, and a few results in a fixture this tight say little about next Sunday. But four straight defeats in the biggest fixture of the Dutch season is a pattern this Feyenoord would very much like to break.

Where this is decided

Both sides line up in a 4-2-3-1 most weeks, so this will be decided in the spaces between the lines. PSV's double pivot of Sano and Mauro Júnior has to stop Zechiël arriving late from midfield; Feyenoord's pair, Zechiël and Vanhoutte in Rotterdam last week, will be trying to keep Til and Mijnans out of the pockets behind them. Sergiño Dest and Filip Kostić give PSV pace and crosses from full-back. That leaves room behind, and Hadj-Moussa, with five league goals, is exactly the sort of runner to use it.

Look at the shots against. Feyenoord concede far fewer chances than PSV, and PSV's habit of giving up 15 or 20 shots against decent sides is their biggest weakness. If PSV have to defend the way they did at Twente, Feyenoord's numbers say they will create enough to win. If PSV control the ball the way they did at Utrecht, the old order returns.

The verdict

On rating PSV are 65 points better, about a third of a goal a game, and they are at home. That rating carries memory: years of PSV being the best team in the country, including a big surplus over expected points last season that is already fading. Feyenoord's season so far, built on chances created and conceded, is at least as good as PSV's, and arguably better.

The deeper question is which version of each team turns up. Feyenoord at their best this season have been relentless: high shot counts, few chances conceded, goals from Ferri, Zechiël and Hadj-Moussa. PSV at their best can blow anyone in the league away, as Utrecht found out. Both have also shown the opposite in the past month, Feyenoord at Telstar and PSV at ADO.

So the numbers point towards a close game. PSV have the higher rating, home advantage and the recent history. Feyenoord have the form, the better defensive numbers and a four-point cushion they can afford to protect. The Index will not guess at a score. A draw would keep Feyenoord top. A PSV win would cut the gap to one point and tell us the old order is not finished yet.