West Ham and the false 4–4–2 switch
After Manuel Pellegrini saved his job with a 1–0 win over Southampton this weekend, most of the credit for the result was ascribed to the manager’s decision to switch to a 4–4–2 system with Michail Antonio upfront alongside Sebastien Haller.
Only half of that is true though.
Pellegrini did decide to pair Haller and Antonio together, but the reality is that we’ve been playing a 4–4–2 shape for the vast majority of the season, as can be seen in the average position maps of our last few games:
Burnley (A):
This one is skewed because Mark Noble (16) got injured and had to be substituted after 20 mins, which meant that Robert Snodgrass (11) moved in to play centre mid and Noble’s replacement Yarmolenko played on the right wing. These maps don’t show average position of substitutes.
But the rest of the general shape is there: a back four (with Cresswell positioned slightly higher than the rest). Two CMs. Felipe Anderson (8) on the left, tucked inside the full-back. Two central attackers, with Pablo Fornals (18) marginally further forward than Sebastien Haller (22).
Spurs (H):
An early sub disfigures the team’s shape in this one too — Anderson (8) was taken off for Michail Antonio at half-time, which meant that Snodgrass (11) had to change positions again, this time shuttling out to the left wing.
We started with a bit of a strange shape anyway. It was sort of halfway between a 4–4–2 and a 4–3–3, with Snodgrass initially playing a hybrid role where he’d defend on the left-side of a midfield 3 but would attack as a second central striker. In the second half he and Antonio who would alternate who would play from the left and who would play through the middle, with the Antonio/Haller pair looking more dangerous.
Chelsea (A):
Probably the most clear example of the shape of the team. Two clear banks of four (although Ogbonna’s touches were a bit deeper than the rest of the defensive line) and Michail Antonio (30) and Felipe Anderson (8) through the middle as a staggered front two.
Wolves (A):
Again, substitutions skew this map and make it look more unclear than it actually was. Pellegrini brought on Nathan Holland for Noble late in the second half, which meant that Felipe Anderson ended up playing the last 20 minutes as a central midfielder. He had quite a lot of touches after moving into that role, so his dot is more central than it should be considering where he played for the majority of the match.
Still a basic 4–4–2 shape, although Fornals’ position on this map reflects how he drifted out to the left wing to combine with Anderson throughout the first ~70 mins or so.
Arsenal (H):
The full-backs are higher and the wide players are further up the pitch — a reflection of Arsenal’s passivity during this game — but the structure of the team is still pretty clearly a 4–4–2.
Now compare the general shape of our team from these games to the shape of the team on Saturday:
Southampton (A):
If anything, this one is more of a 4–2–3–1 than all the times we’ve supposedly played that formation. Snodgrass (11) and Fornals (18) are more narrow than usual because they swapped wings at half time, naturally dragging their dot more central, but the real difference is how much deeper Haller is playing in relation Michail Antonio — his positioning is essentially the same as Anderson’s from the Chelsea game.
That’s not to say that there weren’t differences in how we played against Southampton — we passed the ball longer out from the back and the dynamic between Haller/Antonio is inherently different from the dynamic between Haller/Fornals as Fornals and Antonio play football in completely different ways — but to chalk the win up to a change to a 4–4–2 formation is false. We operated in the same structure as we have for the rest of the season but just executed a different gameplan within that structure.
I know this is pedantic, but I think formation notation is mostly bollocks. In particular, I think the distinction between 4–2–3–1 and 4–4–2 is a waste of time. The reality is that in most cases, those two systems end up functioning the same way.
Take this weekend’s game between Manchester United and Everton. If you ignore Duncan Ferguson’s back 3 misdirect, on paper most people would have described Man United’s starting lineup as 4–2–3–1 and Everton’s as a 4–4–2. But have a look at the average positions map:
If you put Lucas Digne’s (12) position (subbed off after 25 mins; barely touched the ball in his own half but took a corner and several throw-ins deep in opposition territory while he was on the pitch) to one side, the shape of the two teams are virtually identically, particularly up top where Martial/Lingard are playing as close together as Richarlison/Calvert-Lewin.
Or take another team from this weekend. Burnley under Dyche are probably the most stereotypically 4–4–2 team in a lot of supporters’ minds but the map from the Newcastle game shows their literal number 10, Ashley Barnes, playing as a number 10:
I think referring to all of these shapes as 4–4–2 would be far simpler. Most of them end up defending in a very obvious 4–4–2 out of possession anyway.
Pellegrini has rightly been slated this season for how poorly organised we are defensively, a long-standing problem that has been dragged to the surface and exposed by Łukasz Fabiański’s injury. But my bigger criticism is how poor we’ve been going forward — I can tolerate us being shit defensively if that’s a consequence of us being a threatening, committed attacking force, but that hasn’t been the case at all.
A few people have suggested that the performance on Saturday highlights how poorly coached we’ve been by Pellegrini throughout the rest of the season: if we’re capable of playing like that by going longer, why haven’t we tried to play that way more regularly?
I’m not sure that’s an entirely fair argument to make against Pellegrini. He’s only had the benefit of having Haller and Antonio available simultaneously for 6 Premier League matches this season.
Our results in those 6 games? Won 3, lost 3. We’ve won 4 league games in total all season. The 3 games we lost when they were both in the squad were against City, Spurs, and Arsenal.
They have only started of 2 those games together though. However, one of the matches they were both available but didn’t start together was the Spurs match, the first game for Antonio after his injury so he clearly wasn’t capable of playing a full 90. Two of the others were against Chelsea and Arsenal. We’d identified a specific gameplan against Chelsea that involved Antonio playing as the most advanced forward, with a different profile of player to Haller behind him. That approach yielded a result and therefore you can understand the logic behind trying that again vs Arsenal.
It’s also worth noting that in two of the games where they didn’t start together (Watford (3–1 win) and Spurs (3–2 loss)), we scored twice on both occasions after Antonio came on as a sub and played alongside Haller.
It’s clearly a partnership that works and is in fact the only time we’ve had a functioning attack this season, which was immediately obvious to anyone who knew what type of player Haller was/anyone who had a vague idea of the type of football Frankfurt played when Haller was in the team. It therefore feels a bit harsh to criticise Pellegrini for not playing this way more regularly considering he’s mostly been deprived of playing these two in tandem due to injury.
With that being said, playing this kind of direct football can’t possibly have been the plan going into the season, can it?
Michail Antonio is the only player in the squad who is capable of playing as this sort of runner in behind and relying on him to stay fit for your attack to function is a disaster waiting to happen — his fitness issues have been so poorly mismanaged by the club in the past that you pretty much have to just make your peace with him missing 10+ games a season through injury. If this was the design, it’s yet more evidence of how poorly run the club is.
And if this was the blueprint for how we wanted to attack, why was another player with Antonio’s (admittedly quite unique) profile not targeted?
This feeds in to why I think that, while Sebastien Haller is evidently a good player, he wasn’t the right player for this West Ham squad.
Prior to signing Haller, the club had spent somewhere in the region of £80 million in the last two summers on Felipe Anderson, Andriy Yarmolenko, and Pablo Fornals. The problem with playing those 3 + Haller in the same team is that they’re all most comfortable with the ball at their feet and all want to play in front of the opposition defence, rendering our attacks stagnant and predictable.
Considering that we were already financially committed to Anderson and Yarmolenko, we therefore should’ve actually signed a centre forward who would complement their playing style or, if we were intent on bringing in Haller, we should’ve signed someone else more akin to Antonio rather than chucking cash at Fornals.
So while criticising Pellegrini’s coaching might be slightly unfair, the relative effectiveness of this more direct style of football is an indictment of the club’s transfer policy under the Chilean.
As welcome as the victory was, the performance against Southampton by no means solved all of our issues. We were a bit better but I didn’t think we were especially good. Still, we got a result and it was a small beam of light in an otherwise dark period for the club. Hopefully it’ll be the start of a move towards the end of the tunnel.
