Premier League Power Rankings: Why Man City topped Arsenal
ESPN's Premier League Power Rankings return to look at all 20 teams based on performance: Man City prove surge over Arsenal wasn't a blip.
Premier LeagueEnglish Premier LeagueManchester CityArsenal
Full story
At the start of this month, Arsenal had a nine-point lead over Manchester City atop the Premier League table. And just like that, it's gone.
After beating the Gunners last weekend and then taking down Burnley earlier this week, Man City are tied with Arsenal -- on points, and by goal difference -- with just five matches left to go.
The bookies see it as City's title to lose, too. The implied odds at DraftKings give Pep Guardiola's side a 58% chance of chasing down Mikel Arteta's team for the third time in four years.
If only there was a way to see this coming. You could've, I don't know, read this power-ranking column last month, when City vaulted over Arsenal for the top spot? OK fine -- even I downplayed that development in the column itself, but this rating system's simple model picked up on some very real changes to each team's style of play that we've seen borne out over the past month.
With just the stretch run still to go, let's take a look at what the Premier League Power Rankings are saying about what we might be able to expect over the final few weeks of the season.
- Is new Premier League ball boosting long-range goals? We asked a scientist - How bad analytics built a Tottenham team headed for relegation - Ranked: The top 50 players in the Premier League right now
How the Premier League Power Rankings work
Why rank, when we have a whole table that does that for us? Here's the reasoning, as outlined in a previous edition of the rankings:
These aren't meant to be the most accurate power rankings in the history of power rankings -- the goal here is to create something that is simple and intuitive but still has some predictive power. The value comes from comparing the rankings to the table, seeing what looks different, and trying to pry apart the reasons why.
For more detail on the methodology, please refer here. But to sum it up, there are four inputs: (1) talent, as measured by Transfermarkt's squad values, (2) underlying performance, as measured by non-penalty expected goal differential, (3) pressing intensity, as measured by passes allowed per defensive action or PPDA, and (4) buildup play, as measured by pass-completion percentage outside of the attacking third.
Here's how the rankings look with five or six games left in the season:
1. Manchester City (unchanged)
While City's performances hadn't quite caught up to Arsenal's, what the model picked up when it jumped them up to No. 1 was City's increased pressing intensity and the valuable additions of Marc Guéhi and Antoine Semenyo to the roster. The performances eventually caught up, too.
Across the entire season, Arsenal have played at a higher level than City. The points and goal differential are identical, but the Gunners have significantly better underlying numbers:
And well, if you were choosing between two teams who were tied in the standings, wouldn't you choose the one that grades out better by the most predictive individual metric in the sport? There's definitely a reasonable case for that. The Gunners, after all, just played City more or less even -- on the road. And they did it without arguably their most important player, Bukayo Saka.
But if we look at recent performance, then it's City -- nearly lapping the rest of the league. Here's each team's per-game expected goals differential since the start of February:
Normally, I'd always lean on the full-season sample of play over recent performance, but City have real reasons for material in-season improvement: the arrival of Semenyo and Guehi, plus the consistent presence of Rayan Cherki in the starting lineup. Also, Guardiola's City teams do just seem to round into form every spring -- we have nearly a decade's worth of seasons of evidence.
This isn't a sure thing by any means -- even those sparkling numbers since February would be below what we've seen from all of Guardiola's best City teams. But it's getting hard to argue against what these rankings hinted at a month ago.
2. Arsenal (unchanged)
Kai Havertz's performance against Manchester City was a nice little microcosm of what good striker play actually looks like. And that is: you do everything you can to generate chances for yourself.
You press the goalkeeper -- and score a goal. You make runs off the centerbacks -- and you get a 1v1 on a fast break. You constantly pressure the center of the box -- and you get a free header from the penalty spot.
The downside of this, of course, is that you don't always score your chances. And people are way more likely to remember the striker who missed a chance than the guy who never got the chance to begin with. If Havertz finishes off the breakaway or the late header at the Etihad last week, then Arsenal are still title favorites. But if Havertz had been healthy for the whole season, then Arsenal may have already clinched the title.
Havertz attempted five shots against City, which is more than summer signing Viktor Gyökeres has attempted in 550-plus minutes in all of his games against the Premier League's top six so far this season. Throw in linkup play, his off-ball movement, the way he sacrifices his game for his teammates, his aerial ability, and his defensive prowess, and yeah: Havertz is easily Arsenal's best center forward.
If Arsenal do end up finishing in second at the end of the season, then I'd point to injuries to Havertz, Martin Odegaard, Bukayo Saka, and Jurriën Timber as the main reasons why they let it slip away. But the main reason why those injuries have been so impactful is that the most expensive summer-spending spree of any team in the world -- a net spend of €280 million on transfer fees -- didn't really move the needle Arsenal.
Now, they definitely can still win the league. If both teams play exactly to the level of their ratings, then Arsenal will finish 0.2 points back of City. But that's not accounting for schedule difficulty, which also lands in Arsenal's favor -- although a Champions League semifinal against Atletico Madrid will likely take a toll on their Premier League performance.
But we've seen Arsenal's buildup play deteriorate over the past couple months. They've completed 80.7% of their passes outside of the final third since the start of February -- 10th-best in the league and tied with... Burnley. I don't think that elite-defense-and-set-plays approach is going to get them over the finish line.
The hope, then, is the same as it ever was: that the return of Odegaard and Havertz -- and potentially Saka and Timber -- fixes the problems that have formed without them on the field.
3. Chelsea (unchanged)
They still rate highly because they have a super-valuable squad, they press more aggressively than all but two other teams, they complete a higher percentage of their passes outside the attacking third than everyone other than City, and they have the fourth-best xG differential in the league. The fundamentals are quite good, in the aggregate.
And you could even squint and say the same thing about the Liam Rosenior era, where their xG differential in the league was plus-0.4 per game -- which is better than the season-long number and would be good enough for fourth-best in the league over the whole campaign.
But, uh, well, yeah: they've lost five straight games, they haven't scored a goal in five straight games, and they just fired their second manager of the season.
Most of the really positive performances were packed into the first few games under Rosenior, and they've been genuinely average-to-bad in four of the five losses and absurdly unlucky in the other loss against Manchester United.
But as I said last time we did this, a big part of managing is keeping your team on track when the bounces don't go your way, and Rosenior did the opposite of that:
Purple circles are shots, orange circles are goals, and the bigger the circle, the higher the expected-goals value of the attempt.
The game against Brighton was one of the worst you'll see from a Champions League club, and it all but ended their chances of qualifying for the tournament next season. Per Simon Tinsley's projections, Chelsea now have a 2% chance of finishing in the top five.
Just a reminder: this team finished in the top four in each of the previous four seasons before BlueCo took over. Come season's end, they'll have made their way through five different full-time managers and finished in the top four just once, in the four years since they bought the club.
4. Liverpool (unchanged)
The good: this is probably the third-best team in the Premier League, and some finishing variance is probably going to keep them out of the top four, but Premier League teams have done so well in Europe that now five English sides get to qualify for the Champions League. They have a five-point lead on Brighton with a game in hand.
The bad: it's late April, and we still don't really know what they're building toward.
Arne Slot changes the formation every other game. Their one true star who emerged from the summer signings, Hugo Ekitike, tore his Achilles and won't play until 2027. Their record signing, Alexander Isak, is clearly not fit and almost never touches the ball when he plays. Their previous record signing, Florian Wirtz, still struggles to consistently impact matches, and he's not even Liverpool's best player at his best position. Their greatest-ever attacker, Mohamed Salah, is leaving after the season. Their midfield, so important last season, might not work as currently constructed. Neither of their new fullbacks are clearly good enough to play for a team that wants to win every trophy. And both of their goalkeepers are currently injured.
The transition season was coming at some point, and it did, but somehow, there are more questions now than there were back in August.
5. Manchester United (unchanged)
The wheels have started to come off -- even if the results aren't showing it. Here's how United stacks up by non-penalty xG differential since Michael Carrick took over as interim manager:
And here's how it looked before he arrived:
United are very likely to finish in the Champions League places, and they're the favorites to finish third. But they have the 11th-best xG differential in the Premier League under Carrick despite only playing Premier League games during that stretch.
We've seen United ride some unsustainable luck, convince themselves it's real, and come crashing back to Earth so many times since Sir Alex Ferguson left the club. Will they do it again?
6. Newcastle United (unchanged)
Something feels untenable here.
Newcastle still rate highly because they press and possess the ball at above-average rates, they've been unlucky on the finishing front, and they have a theoretically talented squad. But I say "theoretically" because their two biggest summer signings basically never play anymore.
Nick Woltemade hasn't started a game since March, Yoane Wissa hasn't started a game since February, and neither one has scored a Premier League goal in 2026. (The Isak transfer saga really was a lose-lose for both Liverpool and Newcastle, huh?)
But at the same time, it's not like Newcastle are doing well without them, either. Eddie Howe is making a big call by benching the two most expensive signings on the roster -- except their results and their performances have been terrible without them. They're currently in 14th, and their xG differential is only the 10th-best in the Premier League.
7. Brighton & Hove Albion (unchanged)
The rankings have liked Brighton from the start, but not necessarily because they were playing particularly well. No, it was because they have a talented squad and they press and possess the ball like a big club. And it's hard to say that it was wrong to do so.
Brighton have beaten both Liverpool and Chelsea over the past couple weeks. They're up to sixth in the table. They have the fifth-best goal differential. And they have the sixth-best expected goal differential. This is just a good team, and they have an outside chance of qualifying for the Champions League, if...
8. Aston Villa (up 1)
Villa have steadily moved up the rankings since we started this, as their performances have improved and they've solidified some of their control in possession. Tinsley's model gives them a 98% chance of qualifying for the Champions League next season, and that's lower than their actual chances of doing so because they have a second route into the tournament.
As of now, Villa have minus-110 odds to win the Europa League, per DraftKings. That suggests that they're better-than-50% favorites to win the tournament. That's high, in part because it includes the vig, or the fee that bookmakers charge you to make the bet. So, let's be conservative and drop it down to 40%. Add that into Tinsley's odds, and there's a less-than-1% chance that Villa finish outside the top five and don't win the Europa League.
But what's interesting isn't what this all means for Villa but what it means for everyone else. Based on UEFA's rules, if Villa finish fifth and win the Europa League, then the sixth-place team in the Premier League will qualify for the Champions League, too. Tinsley's projections say there's a 30% chance of Villa finishing fifth. Combine that with our back-of-the-napkin Europa odds, and that means there's a 12% chance that sixth place becomes a Champions League place.
Per Tinsley's numbers, there are currently five teams with between a 25% and a 10% chance of finishing in sixth:
Things could get really weird over the next month.
9. Tottenham Hotspur (down 1)
If you're new here: the model refuses to believe that a team with Tottenham's roster could ever actually be a bottom-10 team -- let alone a favorite for relegation. They played reasonably well against Brighton last weekend, and if we somehow magically re-started the season today, I wouldn't pick them to get relegated.
But at this point, two points back of 17th place with five games to play is, at best, a coin flip.
10. AFC Bournemouth (unchanged)
Since letting Antoine Semenyo leave for Manchester City in January -- and since Semenyo immediately became a starter for Manchester City, as they chased down Arsenal -- Bournemouth have... not lost a single soccer game.
If I had lots of money and made decisions at one of the biggest soccer clubs in the world, I would seriously be considering Andoni Iraola as my next manager.
He seemed to work well within the Bournemouth model, where the coach doesn't have a ton of say in who gets signed and who leaves. His teams have massively outperformed their expectations -- first in Spain and now in the Premier League. And his Bournemouth teams have arguably been the most influential stylistic innovator in the league, with their emphasis on pressing and creating and dominating transition moments.
The one big question is if he can create a team that dominates the ball, which you really have to be able to do if you're going to win enough points over the course of a 38-game season to win the league. Neither his Bournemouth nor his Rayo Vallecano sides did that, but he greatly adjusted his approach at Bournemouth after a tough first couple months in England, so he's shown a willingness to adapt to the needs of his new environments.
The Premier League is crying out for a top team that takes more risks in possession and lives with some more chaos. I hope he gets a chance to try to create one.
11. Crystal Palace (up 1)
The main story of Palace's season, on the field, is this:
-Expected goals created: 49.19 -Goals scored: 35
That gap of more than 14 is the biggest in the league; no one else has a gap in the double digits. Had they just converted their chances at an average or even slightly below average rate, they'd be part of that group of teams hoping that Villa finish fifth and win the Europa League.
12. Brentford (down 1)
Over the past month-plus, we've seen both why the rankings have been so low on Brentford and why it's so hard to qualify for the Champions League without a high-possession gear. After getting into sniffing range of the top five with a win over Burnley, the Bees have ripped off... five straight draws.
13. Fulham (up 1)
They're scoreless in five of their last six matches across all competitions, but they're technically still in the mix for a Europa League place and maybe that sixth Champions league spot. But this really just feels like a team with nothing left to play for.
14. Nottingham Forest (down 1)
It feels like they're safe, with a five-point lead over Tottenham, but I'm not so sure. None of their remaining matches are particularly easy: away to Sunderland, Chelsea, and Manchester United, plus home games against Newcastle and Bournemouth. On top of that, they have something relegation battlers almost never need to worry about: a pair of European semifinals, against Aston Villa in the Europa League.
15. Leeds United (unchanged)
Outside of Wolves and Burnley, Leeds might be the most "on the beach" team in the league. They've got a nine-point lead over Spurs, and they have no chance of qualifying for Europe. That said, I still wouldn't want to play them if I had something on the line. Despite sitting so low in the standings and these rankings, they have a positive expected goal differential with five games to go.
16. Everton (unchanged)
Some have joked about the MLS-ification of the Premier League this season. What that incredibly niche comment means is that there's finally parity in a league that had become stratified.
There are, of course, different tiers of quality across the league, but those tiers are closer together than ever before, so actual performance quality -- creating more chances than your opponents -- gets drowned out by the more random, unpredictable, and unsustainable aspects of the sport.
Take Everton. Their xG differential is sandwiched between Nottingham Forest and West Ham. Those two teams have been in a season-long relegation battle. Everton, meanwhile, are one win away from drawing even with Brighton -- in sixth place.
17. West Ham United (unchanged)
The Hammers have been a different team since the calendar flipped to 2026:
Up to Dec. 31, this season: 14 points, minus-17 goal differential, minus-11 xG differential After Dec. 31: 19 points, even goal differential, plus-1 xG differential
They have five more points in 2026 -- despite playing five fewer games. If they play like the 2026 version from here on out, they're way better than 50% to avoid relegation. But, well, those 2025 games all still happened, and a potential reversion toward that form is lurking, somewhere in the background.
18. Sunderland (up 1)
It'll be interesting to see how Sunderland react to a successful first season back in the Premier League. It makes me think back to Bournemouth in 2022-23, who avoided relegation after hiring Gary O'Neill midseason, but decided to still bring in a new coach after the season, presumably, because their underlying performances suggested they were lucky they didn't go down. That coach was Andoni Iraola, and well, look at them now.
Now, I'm not advocating for Sunderland to move on from Regis Les Bris -- I actually think producing the third-worst xG differential in the Premier League is a successful season for a club that finished fourth in the Championship last season. But rather, I mean that Sunderland can't just go into next season without trying to make significant upgrades to the team. Run back this same level of play next year, and they'll be in a relegation battle.
19. Wolverhampton Wanderers (down 1)
20. Burnley (unchanged)
I have some nuanced analysis to share here about Wolves and Burnley: These teams are bad, and they're both going to be playing in the Championship next season.