This Is What I Know About the Premier League
The Top Six is dead, KDB is back, and the title race is two games long
|Ryan O'Hanlon||Oct 29, 2019||6|
A professional analyst once told me that 10 games is right around the time in the season when we can start to make some conclusions about who teams and players are. Under- and over-performance has started to harden into, well, performance. And so, we’re gonna make some conclusions. Er, I’m gonna make some conclusions. I don’t want you to be implicated by the potential inaccuracies that lie beneath this italicized paragraph. Here are six things I know about the Premier League after 10 matches.
1. The Top Six doesn’t exist. We perhaps thought this might be the case coming into the year, but the betting markets didn’t agree. There was an 11-point gap in Sporting Index’s points projections between Chelsea in sixth and Everton in seventh. Well, after Friday’s 9-0 demolition of Southampton, Leicester City aren’t just newcomers who have nestled into the top six; they’re now the favorites to finish fourth:
The current top-six in the table includes both Leicester and Crystal Palace, while the expected-goals top-six (from FBRef) flips out Roy Hodgson’s team for Sean Dyche’s Burnley:
There’s currently a four-point gap between fifth-place Arsenal and fourth-place Chelsea, and then there’s a four-point gap between Arsenal ... and 14th-place Brighton.
2. There’s a clear third-best team. It just isn’t Tottenham. Give a look back at those Sporting Index numbers, and you’ll see that Spurs were the only team other than Manchester City and Liverpool with a number above 70. It started at 73, crept up to 74 after an opening-weekend win over Aston Villa in which they outshot their opponents by an absurd 31-to-7 margin, and it has since plummeted all the way down to 63. That’s tied with Watford, who have five total points, for the biggest drop since the season began. The betting market pegs Tottenham for a sixth-place finish and the third-most points over the rest of the season. That suggests that this team is better than its results suggest. Except, as Richard Whittall recently wrote, this isn’t a Jurgen-Klopp-at-Dortmund situation where a top team is functioning perfectly well in every phase of the game other than converting shots and stopping them. Nah, Tottenham just haven’t been very good. Their goal-differential is currently plus-1, which is tied for sixth-best in the league, while their xG differential is minus-3.3, which is ... sixth-worst. Given the preseason expectations, the past success, the roster talent, and Mauricio Pochettino’s track record, I don’t think you can write off the chance that Spurs get their shit together and launch up the table, but each passing week is making that seem more like an act of imagination rather than any kind of reality-based projection.
3. No, it’s Chelsea. Manchester United have the third best xGD, but that’s aided by the fact that they’ve taken twice as many penalties as anyone else in the league. On top of that, they’re way down the table, in part, because they never score any of those penalties:
Leicester, on the other hand, have a better goal differential than every team in the league other than Manchester City. That, however, doesn’t seem likely to stick, since their xG differential is more than FIFTEEN GOALS WORSE. Jamie Vardy, Youri Tielemans, Ayoze Perez, Ricardo Pereira, Wilfred Ndidi, and Ben Chillwell all have at least double the number of goals compared to expected. If that continues, they’ll coast to a Champions League place and maybe even sniff that title race. If it doesn’t -- and it likely won’t! -- they’re gonna be a dogfight for fourth.
Meanwhile, since getting blown out by United on the opening weekend and then drawing with Leicester in the following match, Frank Lampard’s team has won six of their last eight Premier League matches. They’re topping their Champions League group. And although I had plenty of reservations about Lampard’s ability to manage a top European club -- let alone one that had just lost a player who was the team’s quarterback, wide receiver, and running back -- it’s looking pretty good so far. Oh, and most of this elevated performance came before the team’s 64-million-Euro signing was integrated into the starting XI. Funnily enough, Christian Pulisic’s hat trick against Burnley over the weekend wasn’t at all the result of what he’d been doing when he was coming off the bench -- getting into high-value goal-scoring positions -- but it certainly evened out all of the chances he’d missed over the first few months.
4. The best English striker is ... Tammy Abraham. This should do the trick:
The scariest thing about this chart? While the guy in red is 26, right in the middle of his peak years, the guy in blue is 21. (Want more on Harry Kane’s decline? Here ya go.) If Abraham ages like most players do, then all of those bars are eventually gonna reach up even higher.
5. The best player in the league is Kevin De Bruyne. This might seem like it’s always been true, but in De Bruyne’s first season with City, N’Golo Kante was still willing whatever team he played for to the title. In his second year, Mo Salah put together either the first- or second-best individual attacking season in the history of the league. And last year ... De Bruyne only played 974 league minutes.
This year? Well, he’s lapping the field. So far, De Bruyne’s got eight assists, while no one else in the league has more than four. He’s also ahead of everyone else in total goals+assists and total non-penalty goals+assists. He’s got the most expected assists in the league -- total and per 90 -- he’s creating the most chances, he’s playing the most through balls, he’s completing the most passes into the box, and he’s also, per Stats Perform data, playing the most passes that lead to the pass that leads to a shot. Here’s the top 10, in case you’re curious:
On top of all that, De Bruyne is third in the league in non-penalty expected goals+assists per 90 minutes, which is perhaps the best indicator of sustainable attacking performance. And here is where I remind you that he is ... a midfielder. Everyone else in that top 10 of NPxG+A/90 list is a winger or a striker.
Here’s a quote from a paper I’ve enjoyed and cited before:
To help fill the gap in objectively quantifying player performances, this paper proposes a novel data-driven framework for valuing actions in a soccer game. Unlike most existing work, it considers all types of actions (e.g., passes, crosses, dribbles, take-ons, and shots) and accounts for the circumstances under which each of these actions happened as well as their possible longer-term effects. Intuitively, an action value reflects the action’s expected influence on the scoreline. That is, an action valued at +0.05 is expected to contribute 0.05 goals in favor of the team performing the action, whereas an action valued at -0.05 is expected to yield 0.05 goals for their opponent.
And here’s how that framework valued KDB in his last full season for City:
However, as the authors of the paper wrote:
Strikers Harry Kane and Mohammed Salah perform a relatively low number of actions but their actions are highly valued on average. Midfielders Kevin De Bruyne and Paul Pogba perform more actions albeit with a lower average value per action. Philippe Coutinho, Eden Hazard, Riyad Mahrez, Anthony Martial, Raheem Sterling, and Son Heung-min fall in between these two archetypes, hitting
Except, this season De Bruyne is still doing all of that high-frequency-low-value stuff that comes with being a midfielder, and he’s coupling it with all the high-value stuff, too. In 2018-19, his xNPG+A/90 was 0.52; this season it’s 0.97. He’s started 11 games in the Champions League and Premier League combined: City won 10 and drew the other (in which De Bruyne created nine chances -- the highest single-game mark of the season.) The two league matches he didn’t start? City lost both.
6. The title is gonna come down to two games. Some quick, back-of-the-napkin-except-I’m-not-using-a-napkin-and-just-writing-it-as-I-go-because-efficiency-is-king-in-the-modern-economy-math: City’s goal differential is plus-23, and Liverpool’s is plus-15. One goal of GD is typically worth about 0.6 points. So if City and Liverpool perform at roughly the same level over the 26 remaining games in which they don’t play each other, we’d expect City to win about 12 more points than Liverpool: 0.8 goals per game, times 26 games, times 0.6 points. With Liverpool six points ahead of City, that’d put City six points ahead of Liverpool. Or, in other words: the exact number of points that’ll be up for grabs in the two City-Liverpool games.