What’s going wrong at West Ham?
The general consensus among Football Twitter analysts seems to be that West Ham are a much stronger team than their league position suggests and that their results will catch up with their performances. As far as I can tell, that seems to be predicated on them jumping onto Fbref, sorting the Premier League table by xG Difference per 90 and seeing West Ham in 8th.
But that doesn’t line up with just how poor West Ham have been in the last 12 months under Moyes. So I poked around a bit and there’s more to their performance data than the bean-counting of xG.
The TL;DR of it all is that West Ham’s underperformance of their expected goals is born from their style of play. And that is dictated by their manager.
Spotkick Spotlight
If you sort the Premier League by Expected Goal Difference per 90 (i.e. the average difference between a team’s xG for and their opponents’ xG for in every game), the table has West Ham as the 8th best team in the league. Stop there and everything is all sunshine and rainbows.
But the first major issue with this metric is that it doesn’t strip out xG created from penalties, which is relevant because West Ham have taken 6 pens this season (only Fulham have taken more with their total of 7) and only conceded 3 penalties in return. Fbref now uses Opta data and according to Opta’s Analyst site, penalties are given a “constant value corresponding to their overall conversion rate (0.79 xG)”. That penalty differential is worth 0.11 xG per 90 (0.79xG x 3/21 games played) which accounts almost entirely for West Ham’s xG superiority over their opponents this season (0.16 xG p/90).
Whether it’s fair to remove penalties from the equation is up for debate. After all, you have to have possession of the ball in the opposition’s penalty area to be given a pen and if you’ve been fouled in the box, chances are you were on your way to getting a shot away, at the very least. It’s questionable whether winning pens is a skill or a product of fortune, but generally speaking the number of penalties a team wins isn’t repeatable season-to-season and is rarely a reflection of a team’s attacking skill.
To illustrate that point: last season Liverpool and Arsenal took 8 penalties each in the Premier League; this season Arsenal have taken 1 and Liverpool are yet to receive one. And, for what it’s worth, West Ham’s 6 so far this campaign is the same number they took in the entirety of the last league season and seems especially unlikely to be a reproduceable strategy given they attempt the second fewest dribbles per game — and dribbling would surely be linked to penalty winning skill/likelihood — and draw the third fewest fouls in the whole league.
If you do decide to leave out penalties, West Ham are virtually dead even with their opponents when it comes to the quality of the chances they create. And that becomes a problem due to Moyes’ game model.
Attacking Moylaise
In short, Moyes is playing an extremely reductive brand of football, aiming to make the game as marginal as possible while hoping to be stout defensively and clinical when chances do arise; it’s typical bunker in and nick one on the break stuff. West Ham’s 21 league games this season have featured the 2nd fewest total goals (44 along with Everton — Chelsea’s games have featured one goal fewer) and the 3rd lowest total xG (49.9 xG, behind only Palace and Southampton). So games they’ve been involved in have seen very few goals and very few quality chances — if I were Sky or BT, I’d be steering well clear.
It’s not for a lack of trying, though. West Ham are rated the 7th best team for getting shots away…
and their games in general have seen the 8th highest total number of shots. The problem is the quality of those created chances. Unsurprisingly, the high volume of shots coupled with the low quality of those shots mean that West Ham games see the lowest xG per shot in the Premier League.
And that chance quality is pretty much evenly distributed within games; West Ham average 0.08 xG per shot (joint worst in the league) and hold their opponents to 0.09 xG per shot (joint second best in the league). That expected goal difference per 90 superiority is purely a product of the volume of West Ham’s shots, not their quality.
More blocks than Legoland
There’s plenty to suggest that West Ham’s execution of their game plan under Moyes is to blame here. They have been an absolutely miserable side as an attacking force. Ranking lowest in the league xG per shot looks more miserable when you consider the outcomes of the shots they take:
West Ham hit the target with their shots less often than any other side in the league, with barely one in four ending up within the frame of the goal. Almost 10% less often than a league-average side. Monumentally bad stuff.
You can put this down to the individual players responsible for taking those shots (which would be troubling considering how much has been spent on attackers during Moyes’ tenure) but there’s something else at play here. At least that’s my take when we look at number of blocked shots taken:
Those numbers tell a story. West Ham have had the 4th highest number of shots blocked in the league and just look at the teams they’re surrounded by there: 5 of the current top 6 (+ Liverpool). Those teams are regularly playing against compact low blocks with lots of bodies between the ball and the goal. Opponents do not set up like that against West Ham and yet their shots are blocked a similar amount. It gets even worse when you factor in that the top 5 teams take considerably more shots per game than West Ham do — over 1 in 3 shots (~35%) that West Ham players take are blocked, comfortably the worst proportion in the league.
That’s probably related to the distance West Ham are shooting from rather than simply a reflection of poor shooting technique:
West Ham are one of 6 teams whose average shot location is from outside the penalty area, with Villa being the only team who shoot from further out. Averages can be misleading though (10 shots from 18 yards has the same average as 5 shots from 36 yards and 5 shots from 1 yard) so it’s worth having a quick look at their shot maps from the last 5 games:
Barring the Everton and Newcastle games where we scored non-headed goals from set pieces, virtually every shot West Ham have taken post-World Cup restart from closer than the penalty spot has been a header. The rest are mostly hopeful long range pings. There’s a possibility they’re getting a bit stitched up by data collection here (I’ve long felt that headers are misrepresented by xG figures — the xG for a shot works on the basis that the identity of the shooter is irrelevant but that’s not the case for headers, where the height and size of the shooter compared to the defender(s) influences the quality of the chance) but even so, there are very few open play shots from close range. This isn’t the output of a team that’s going to reliably score from open play.
And it’s not like it’s a case of the wrong players getting ideas above their station and getting a rush of blood to the head when the crowd screams for them to shoot:
Pull out Cornet, Ings, and Lanzini, whose per 90 numbers are inflated by playing off the bench, and the top 5 shot takers here are exactly the players you want to be shooting at goal in this team.
It’s easy to understand why the club were looking at bringing in another goalscorer in the January window, given the horrendous rate at which the existing squad have hit the target, but the distance the players are shooting from and the fact that so many shots are so easily blocked says to me that the players aren’t being coached in a way that creates good goalscoring opportunities from open play with regularity. This is a team relying on individual brilliance to score goals rather than a collective system for generating chances. Those individuals have faltered and the team’s attacking output has fallen off a cliff as a result.
Game state or complete state?
At this point, it’s worth delving slightly deeper on the xGDiff again. Averages, as previously stated, can be misleading but looking at the distribution of the xGDiff within West Ham’s matches this season gives a fuller picture:
11/21 Premier League games this season have had an xG difference of more than 0.5 (i.e there was more than a 50/50 chance that the superior team would’ve scored an extra goal). In 7 of those 11 games, West Ham have been the side with the majority of the xG; in 3 of those 7 games West Ham have been awarded a penalty. Put another way, there have only been 4 games so far this season where West Ham have been significantly more likely to win than their opponents based on the quality of open play chances created.
Those 4 games were against Tottenham (H), Southampton (A), and the two fixtures against Everton. In all of those fixtures, excluding the most recent Everton game, West Ham have conceded the first goal (after 20 mins against Southampton, after 34 mins against Tottenham*, after 53 mins against Everton). Finding themselves behind early, they’ve been forced into a situation where they need to chase the game, hence the greater attacking impetus evident in those matches. And against Spurs and Everton, they only managed 0.6 xG more than their opposition, barely scraping into a position of superiority.
*(It’s worth noting that Tottenham’s opener was an own goal and, as such, wouldn’t have an xG value attached to it. So the numbers are skewed in West Ham’s favour for that one.)
The same was true against Forest — they were 1–0 down at half-time and then only managed to be 0.8xG better than them in a match where Rice missed a penalty, which accounts for the entirety of the difference in chance quality between the two sides.
And it’s here that I want to come back to the number of West Ham’s shots that get blocked. At face value, there’s no reason for them to have a similar numbers on this front as the elite teams:
My initial reaction was that “opponents do not set up like that against West Ham”. But what if they do? Why would that be the case?
One explanation is that it’s caused by going 1–0 down so often at the start of games. Troublingly, they’ve conceded the first goal in 14 of their 21 league games, which is obviously a massive issue when your approach is dependent on being defensively resolute. 12 of those 14 games have seen them 1–0 down at half-time, with 8 of those goals coming before the half hour mark.
You can argue that there’s an element of bad luck about that. Using the same metrics as earlier, West Ham have been impressive defensively: middle of the pack for the number of shots they allow but the quality of the shots they let their opponents take is the joint-2nd worst, mostly because they keep their opponents at arm’s length by forcing them to shoot from the edge of the box (only Brentford and Spurs do better on that front).
Defensive titans or Titanic defence?
On a surface level West Ham profile as a borderline elite defensive side who rarely give up good chances. Yet they’ve conceded the first goal in two thirds of their league games. What’s going on? Are they just extremely unlucky?
Are their opponents just finishing their chances incredibly well?
At first blush it seems like that’s not the case — they’re slap bang in the middle for opposition conversion %.
What about the shots that have resulted in them going behind?
If we put the pens and the OG from the City/Brighton/Spurs games to one side, that leaves 11 games where West Ham have conceded the first goal. In 4 of those games the goal was an Opta-defined “big chance” (>0.3xG) with another 2 being a better than 1-in-5 chance. The 4 low value goals they conceded were turned into substantially higher quality by the shooter’s execution of the shot. It’s a bit of a mixed bag, with a handful of games where they’ve been on the end of some hot finishing.
There’s an important question to ask here about whether shot data is a reflection of strategy and quality or more of a reflection of the circumstances of the game. If you look at the numbers in a vacuum, West Ham measure up as a top level defensive side who restrict their opponents to very few quality chances on average.
With added context though, you can ask whether West Ham’s defensive performance is a result of finding themselves behind so early, so often. In over half their games they’ve conceded at least once high quality chance in the opening stages and consistently been punished for it. Do teams feel the need to force the issue and continue to manufacture high quality chances when they’re already ahead? Or do they just settle and limit their attacking output knowing they’ve got a lead to hold onto?
Viewing things from that perspective perhaps explains the underwhelming attacking performance. Even so, given they frequently find themselves 1–0 down — regularly having at least an hour to chase the game — how have they only managed to create the better chances than their opponents 7 times? It’s somehow actually even worse than that suggests. For instance, take the draws against Liverpool and Leed, which look fairly even on the xG front (the xGDiff was dead even against Liverpool and -0.2 vs Leeds). In both games West Ham conceded early and were only bailed out by winning penalties. Overlook those and they were vastly inferior than their opposition from open play despite having over an hour to get back on terms.
This is where all those blocked shots start to make a bit more sense. If you apply the same game state framework to West Ham’s own attacking output, a pattern begins to emerge: West Ham give away a big chance early on that gets converted; the opposition play more conservatively, attacking less often (reducing the quality of their shots) and focusing more on retaining a compact defensive shape (reducing the quality of West Ham’s shots); West Ham don’t have the ability to break them down so start rattling shots off the shins of opposition defenders. This pattern reduces reduces games into a frustrating attributional affair for everyone involved; it’s coastal erosion football.
It’s an especially crippling pattern for a team whose success in recent seasons has been so dependent on playing in transition — if teams don’t have to push forward to score against you, they’re leaving far less space for you to attack on the counter.
The Blame Game
None of this is to absolve David Moyes, however.
If you want to be charitable, on the defensive end you could say that the concession of high quality chances early on is linked to the injuries the club has had to deal with, especially at Centre Back — new signing Nayef Aguerd has only featured in 5 Premier League games so far since joining in the summer. In his place, Thilo Kehrer and cat botherer Kurt Zouma have played 16 and 15 PL games respectively, both of whom were expensively acquired by Moyes. If they aren’t good enough to execute his game plan, then that’s a failure of the club’s transfer strategy under Moyes.
On the attacking end, there have likewise been issues with Scamacca’s minor injuries/troubles settling and Antonio’s physical decline, but the problems run deeper than that.
Solving problems is a major part of a manager’s job and it’s one that Moyes has consistently failed at this season. His team have regularly gone behind and there’s been no strategy for finding solutions to stubborn defences anywhere to be seen. This isn’t a team that has been coached to create space in the final third with clever player movement or one that has been taught patterns of build-up play designed to disrupt the opposition defensive structure. All this side has in their attacking arsenal is playing on the counter and slinging crosses into the box.
There’s nothing inherently wrong with wanting to play a sit-deep-and-break style of football — there have been plenty of teams who have been successful by being difficult to beat and ruthless in front of goal — but it’s a passive and reactive one, one that can be miserable to watch when it’s dysfunctional and is particularly galling for fans considering the club have spent the best part of £200mil on (predominantly) attacking players who were supposed to inject superior technical quality into the side. The whole point of these signings was to evolve the playing style and provide the team with more options to develop a game plan in possession. That has been entirely absent so far.
Even if you want to characterise the team’s defensive woes as a product of bad luck, a glaring issue with Moyes’ tactical approach has been exposed: if you’re trying to restrict the number of quality of chances in the game and banking on taking chances yourself when they arise, it’s very easy for the fine margins to fall the other way — it only takes some wonky finishing on your end and one to hit your defender on the arse and deflect in to see you come out worse in a dead-even match if you don’t seek to oppose yourself in an attacking sense. It collapses as a viable strategy when you go 1–0 down because the dynamics of the game shift in your opponents favour.
West Ham’s games have featured such low volumes of goals and quality chances because their opponents know that once they’re ahead, West Ham aren’t good enough to break them down. And therefore they don’t have to continue attacking because one goal is usually enough to beat them.
If you want to take the surface level numbers, the Irons are the worst open play attacking team in the league and somewhere around the 4th or 5th best defensive team in the league. That should, theoretically, average them out as 12th or 13th place overall and they are currently 5 points adrift of that. A deeper look at the context surrounding those numbers suggests that the defence isn’t as good as it seems and the dreadful attacking performance is largely caused by game state.
But even so, the way Moyes sets his team up means they are very rarely positioned to score more goals than their opponents on the balance of play and if you do that, you’re liable to have bad luck stack up against you.
And for a club that laments how fortune’s always hiding, that’s not a smart approach.
