The international break always comes at the worst time, huh? The blistering second half of Manchester City–Liverpool seemed like the perfect scene-setter for a potentially epic title race between three of the best soccer teams in the world. Chelsea are alone in first, but Liverpool and City were both back to their pre-pandemic best, and they both showed it on Sunday. Let’s go!

Or not: no club soccer for two weeks, everyone go scatter the globe, and hopefully you don’t get hurt! However, one of the few positives of this international break is that it gives us a chance to take a step back. We’re seven games into the Premier League season: What’s changed? What’s stayed the same? What do we know now that we didn’t know before the season began? Using FiveThirtyEight’s projection system, let’s take a look at how everyone’s end-of-season expectations have shifted -- up, down, or even sideways.

The Title Race

The big shift here is that Manchester United are all but out of it already. Of course, there’s a reasonable question as to whether or not they were ever actually in the title race to begin with, but we’re not looking back -- only forward. Although Ole Gunnar Solskjaer and Co. are currently even with Manchester City, one back of Liverpool, and two back of Chelsea, their title odds have fallen from 10 percent to start the season, down to four percent as of Monday morning.

So, uh, what happened? For starters, they just haven’t played all that well. Their expected-goal differential is currently sixth-best in the league, at plus-2.9. That’s closer to Arsenal’s second-worst number (minus-5.0) than to Manchester City’s league-best mark of plus-11. On top of that, they’ve produced those unspectacular performances without playing any of the big three or even “Big Six” clubs. Meanwhile, Liverpool, City, and Chelsea have each played each other already. Throw in a couple poor performances in the Champions League, and United’s SPI rating, which represents how often a team would be expected to take all three points from an average team on a neutral field, has fallen from 85.7 to 83.4. Things get rough right after the international break, too:

𝐀𝐅𝐂 𝐀𝐉𝐀𝐗 💎 @TheEuropeanLad

Manchester United with a tough Premier League schedule after the International break: • Leicester (away) • Liverpool (home) • Tottenham (away) • Manchester City (home) • Watford (away) • Chelsea (away)

October 2nd 2021

188 Retweets2,820 Likes

Despite all that, United’s end-of-season projection has barely budged: down one point to 67. That speaks some to the initial skepticism held by this model and most other math-based analyses of Manchester United’s underlying team strength, but also to how good their supposed rivals have been. Chelsea’s SPI is basically (88.7, began at 88.8) right where it was to start the season, while Liverpool’s has risen to 89.8 from 88.7 and City’s has shot up to 93.5 from 92.1. Same goes for their projected point totals and title odds, too:

The model isn’t taking into account the fact that Liverpool will lose Mohamed Salah and Sadio Mane, who are nos. 1 and 2 in the Premier League for combined non-penalty xG and expected assists so far this season, for the Africa Cup of Nations in January. Mainly due to that, the betting markets basically flip Liverpool and Chelsea from where FiveThirtyEight has them. My guess is the right projection is somewhere in between; they’re ultimately only going to miss two Premier League matches because of the tournament.

Quibbles about Liverpool’s likelihood of lifting the Premier League trophy aside, it really does seem like a three-team title race now. Manchester United needed to either A) build up a points lead from their easy early-season schedule or B) show some kind of underlying performance level that suggested they could consistently reach the same heights as their closest competitors. Through seven games, they haven’t done either one.

The Mid-Table

We could do a top-four talk, too. FiveThirtyEight gave United a 50-percent chance of finishing top four before the season, and that has dipped ever so slightly down to 48 percent.  However, the betting markets have United finishing 14 points clear of fifth place -- and where there’s a discrepancy that big I tend to side with the people who actually have money on the line. If there were a true top-four challenger, like Leicester over the past couple seasons or Arsenal and Tottenham back when they weren’t set-to-simmer minor disasters, then United might seriously be in trouble. It’s by no means a given that they finish top four; the team still hasn’t come anywhere close to performing at a level commensurate to its talent, but that talent is still better than anyone trailing behind them.

As for those trailing behind, Everton have improved their projected point total by six points, while Spurs and Arsenal have dropped by four and five, respectively. Here’s what the projected race for the Europa League spots looks like right now:

5) Everton: 56 points (plus-5 from preseason)
6) Tottenham: 56 (minus-4)
7) Arsenal: 55 (minus-5)
8) West Ham: 54 (plus-2)
9) Brighton: 54 (plus-3)
10) Aston Villa: 53 (plus-3)

When given more than a week or two to coach a team, Rafa Benitez continues to be one of the surest bets around for getting a set of players to perform to at least their average combined potential, if not better. Everton’s schedule has been soft, but they’re currently third among all teams in xG differential. Spurs waffle between godawful and decent on a day-to-day basis, and although the results have piecked up after an 0-3 start, Arsenal have only created more xG than their opponent in one match across their current four-match unbeaten run. Brighton are the opposite of this year’s Brighton, which helps explain only the minor point bump from their hot start: same number of points as Manchester City despite just a plus-3 goal differential and a negative xG differential. That’s not how regression to the mean works! West Ham would be my pick for the pure fifth-best team in the league, but the Europa League might already be starting to have an effect on a squad that’s not used to playing this many games; their only two losses this season came the weekend after a Europa League game. Aston Villa are just ... there. Among all these teams, they’d be the only fifth-place finisher who would actually surprise me.

Speaking of surprises, of opposite sorts: Brentford and Leicester City. The Bees have improved their projection by twice as much as any other team in the league: vaulting up from 38 points to 50. Coming into the season, their relegation odds were 36 percent. Now? Just five. They’ve banked a ton of points (12) and have done so with sustainable performances (plus-2.2 xGD). It’s hard to have a better start to the season than that, and it’s hard to have a worse start than Leicester. The warning signs were there last season with their poor underlying numbers ... and then they got out-shot by Norwich, the clear worst team in the league. They’ve gone from a legit top-four challenger in each of the past two seasons to, I don’t know, barely a top-10 team? Maybe worse? The scary part is the only attacker who’s producing anything is Jamie Vardy, who turns 35 in January. And without Jonny Evans (one of the most underrated players in ... Premier League history?) and Wesley Fofana, the bottom has fallen out on the other end. Their points projection has plummeted from 57 to 49.

Oh, and there’s Wolves: playing a much more fun, conducive-to-Adama-Traore-doing-cool-shit style, and probably finishing somewhere near the top of the bottom half of the table again. They’ve gained two points -- 46 to 48 -- on their preseason expectation.

Relegation Zone

Points are less important here than your points relative to the other teams -- like the title race in reverse. Here’s a similar chart to the one two sections above, including all the teams with at least a 10-percent chance to be relegated:

We’ll start with the good: Palace! They’ve shaved 16 percent off their relegation odds and added five points to their end-of-season projection. Patrick Vieira has really done a decent job so far; the’ve conceded as many xG as Liverpool so far this season, and they’ve found just enough goals from the offseason arrivals of Odsonne Edouard and Conor Gallagher. Another month or so of solid play -- they’re close to even by xGD -- and they should be able to clear away from drop-zone. Southampton, meanwhile, are the same as they ever were: somehow the most volatile and most predictable pressing team in the world, at the same time. And Leeds find themselves pulled into the outskirts of the relegation battle because they’ve conceded 115 shots, tied for the league-high with Tottenham and just one more than Arsenal. (See the problems here?) However, they pummeled Watford over the weekend, and much like Palace and Southampton, a month of good play could lift them from the danger zone.

As for Watford, they just fired their coach, Xisco Munoz, despite winning the same number of points as Palace and one more than Leeds so far. Fifteenth with a minus-three goal differential: not bad for the team that was supposed to finish last, right? Well, the problem is that they haven’t played anyone. Their next eight matches: Liverpool, Everton, Southampton, Arsenal, Manchester United, Leicester, Chelsea, Manchester City. Godspeed, Tinkerman.

Newcastle and Burnley each have three points in the exact same manner: three draws, four losses. Only Norwich, Arsenal, and Watford have worse xG differentials than Newcastle, and Burnley is just one spot and about half a goal ahead of the Magpies. Burnley are still Burnley, somehow combining to win the ball up the field more often than any team in the league ...

The Analyst @OptaAnalyst

@johnspacemuller 73 overall - three more than Liverpool. Just like Liverpool, they've attempted 10 shots from these.

October 4th 2021

4 Retweets16 Likes

... but also pressuring the ball less often -- measured by opponent touches per pressure -- than any team in the league. Launch it long, try to win it back ASAP, and if that doesn’t work, drop into a shell. It’s worked in the past, but their roster has never really been turned over, and their average age (weighted by minutes played) has been creeping upward ever since they finished in seventh in 2018. Remember that? Wild, man. Anyway, now they’re the oldest team in the league, with an average age of 29.2, meaning the average Burnley player is on the downwide of the age curve. Given how Burnley typically floated around the range of the 15th-best team in the league for a while, “everyone got old and therefore worse” is a pretty good way to get relegated. 

Newcastle, meanwhile, just remain bad. They’ve been more fun this season -- Allan Saint-Maximin has been genuinely good, and the games are open -- but overall it’s still just as bad as it’s always been under Steve Bruce: don’t press, can’t consistently attack, can’t consistently defend. At least Burnley has an identity. Last year’s loan savior Joe Willock, shockingly, has not continued to score with half of the shots he takes.

Then there’s Norwich, with one point from seven. Consider them a symbol for what the first seven games of the season indicate: enough games to significantly shift a team’s fortunes and alter our assumptions about team strength, but nowhere near enough to definitively change anything. The defending Championship champs are 70 percent for relegation now, but hey, that means they’ve also got a 30-percent shot at staying up.