West Brom, Doncaster & Crystal Palace vs West Ham
I didn’t write anything about these games as they happened because they were either midweek or FA Cup games against significantly weaker opposition, so I just didn’t quite get round to it, but thought I’d do a little summary of my thoughts on them and the future. They were 3 wins, 2 of them very comfortable (I don’t care that Palace finished 3–2, it was a comfortable win) and 1 of them fairly comfortable.
We conceded 3 goals in the 3 games, but once again our defence was in reality fairly rock solid. The three goals were:
- A shot from outside the box
- A shot from outside the box
- A 94th minute consolation goal
In the two premier league games the combined xG was 5.48 to 1.09. We won those games 5–3, primarily due to good finishing by the opposition. I don’t have the xG for the Doncaster game, but shots were 20-4, and shots on target 12–2. That’s dominant across all 3 games, and we’re somehow now in 5th and in the FA Cup 5th round. We’ve been playing well for a while now, and it’s weird.
An effective system?
A common theme through these games, and really every game since the Chelsea game, has been the strength of our defence. We don’t concede many shots, we don’t concede much xG, we don’t concede many goals. As I’ve said before, I think a big part of this is the Soucek — Rice double pivot, but I think that this might actually be a significantly smarter system than I initially realised.
From what I can tell (and I’d need to rewatch games properly looking out for this to be 100% sure) what I think we have what is actually quite a fluid system, which is simultaneously very defensively solid. If you just took the players at face value, you’d say we have six “defensive” players — the back 4 and the two CMs — and four “attacking” players. This could very typically form a defensively solid side that is a bit crap going forwards — think Big Sam. However, this system seems to be more effective in attack than you’d expect *while not compromising its defensive stability*.
What I think makes this work is that we are able to add bodies when going forwards, without losing our defensive structure. In essence 4 players will get forwards for every attack, but they’ll regularly be joined by anything from 1–3 other players, depending on the situation. This means that we have bodies forwards, but always have a solid three at the back, made up of the two CBs and at least one of the CMs or FBs (although often more). I think the biggest advantage of this isn’t that it’s defensively solid, although that’s obviously great, it’s that it’s actually quite effective going forwards.
One of the big things I’ve learned about attacking strategy over the last few years is that you want your attack to be well coached, so that players know where they should be and where their teammates should be, but also with an element of randomness and unpredictability, so that your opponents don’t know where people are going. To me it feels like this system Moyes has found, where Benrahma is the key progressor and provider, while anyone except Ogbonna and Dawson can join the attack, is a very effective way of achieving this.
https://twitter.com/MC_of_A/status/1354147671179722752?s=20
The future
We’re 5th. 20(ish) games in and we’re 5th. If you play every game in hand right now the worst case scenario is that we’re 7th. After 20 games! This isn’t some freak run of form at the start of the season, this is legit. 538 have us projected to finish on 60 points in 7th, with a 12% chance of top 4, 22% chance of top 6, and only a 28% chance of finishing outside the top 10. That is far better than my wildest dreams going into this season — and there are two reasons I’m optimistic for the longer term future.
The first is that, unlike the last time we did anything like this under Bilic, we aren’t massively outdoing any metrics. Often when teams have a good season, it’s on the back of running hot. Every chance seems to go in, and opponents randomly miss more chances than usual. That helps propel you to a good finish, but isn’t a good indicator of how you’ll perform the following season. If you’re just playing well though? That’s a very good sign.
FBRef has us at +4.8 xGD compared to +6 GD. Understat has us at 31.05 xPoints, compared to 35 points. These are overperformances, but they really aren’t that big. In addition to this, Understat has our xG as 30.07 (actually 30) and our xA as 23.89 (actually 24). That’s bang in line with our performances! It’s repeatable!
Compare this to the Payet season — Understat had us with an xG of 54.40 and an xGA of 55.48. We scored 65 goals, and conceded 51, an overperformance of 15 goals, and meaning we got 62 points compared to our xPoints of 49.80. One of the big reasons it fell apart for Bilic was that as well as getting worse in terms of underlying numbers, we regressed to the mean — we stopped overperforming our metrics. That’s not something I’m worried about right now, and it’s nice.
The second reason I’m cautiously optimistic is that this doesn’t quite rely on one player doing *everything* in attack. We aren’t reliant on a Payet. Antonio has started half of our games, and he’s our only good striker. Masuaku has started half of our games, and our initial system was reliant on him — he got hurt and we found a new system. That’s a good sign for the future, as we shouldn’t need to worry too much about any one player getting injured or leaving.
The caveat to that is the two CMs. That’s the position I’m nervous about. If one of them gets hurt, I think our level will drop off significantly. We don’t have any cover there, as Mark Noble can’t run anymore, and so if one of them gets injured (or bought by Chelsea) I don’t know how we handle that. Hopefully that and centre-forward are the two positions we’re looking for cover in, as we have no depth. We maybe don’t need it for the rest of this season as we push for a top 7 finish, but we definitely need it for next year.
Overall? I’m actually optimistic about us for the first time in a while… Now watch us sell Rice and not buy another CM, and it all fall apart.
I don’t really know why I’m writing this — it’s lockdown and I’m bored I guess. I’ve considered trying to write some kind of stats based analysis of West Ham for a while now, so this felt like as good a time to start as any. I’m going to try to write these semi regularly, but who knows if that will last. Let me know what you think!
