3 Hot Takes: Man City, Man Utd, VAR all stumble

Some thoughts on the Premier League this weekend

Hi, and welcome to what I hope will be a recurring new feature. I’m just taking three of the most noteworthy stories from the weekend and giving my briefer reactions to them. This is a little less formal than my usual articles, and my thinking is I can get these out quicker than the single-topic pieces so they can run in addition to those. I do like this style of writing where it’s faster and more instinctive. Let’s get on with the topics.

Stats are from FBRef unless stated otherwise.

1. Man City have an off day at Wolves

Pep Guardiola has heavily tweaked his side in the last 18 months, but his core structure has remained the same since his first season at Bayern Munich.

It starts out as a 4-3-3 but then shifts into a 3-2-2-3, or 3-2-4-1 depending on how you view the wide players. He almost always wants to have three defenders staying back, a double pivot in front of them, two wide players hugging the touchline, and two attacking midfielders behind the striker. At Bayern, that usually meant having the defensive midfielder drop in between the centre backs. *Grandpa Simpson voice* which was the style at the time, making room for the full backs Philipp Lahm and David Alaba to both push up and become defensive midfielders. The wingers would stay wide as the sole outlets hugging the touchlines, while the two more advanced central midfielders would really push up and become duel number tens. The striker would stay as a static focal point. This is the one major tactical shift of Guardiola’s career that’s defined his post-Barcelona era.

In Guardiola’s first title-winning season at Man City, the breakthrough was having right back Kyle Walker stay deep and become the third centre back, while left back Fabian Delph would push up and become the second defensive midfielder next to Fernandinho. He’s tweaked it here and there, but that core structure has stayed the same. During the second half of last season, the big tweak was to have John Stones push up from centre back into defensive midfield, while both full backs stayed deep. He’s finding different ways of getting to the same result.

City’s spot is in a weird spot right now. Stones, Kevin De Bruyne and Bernardo Silva are all injured and Rodri is suspended, while Ilkay Gündoğan and Riyad Mahrez left in the summer. Guardiola is thus having to find solutions on the fly with players who aren’t totally integrated into his methods. Currently, he’s playing what looks like a 4-3-3/4-2-3-1 (it kind of blurs the difference) on paper, starting out like this:

Look, Share My Tactics was down when I was writing this newsletter, so this is what I have.

Then Phil Foden moves inside from the right to become a second number ten next to Julián Álvarez, while Mateo Kovačić drops a little deeper into the double pivot. Walker rolls back the years to become a genuine attacking full back bombing down the right, and we get that same Pep structure again:

There are three new signings (Kovačić, Matheus Nunes and Jérémy Doku) in that team, plus two players in different roles than they’ve typically played (Walker and Phil Foden). With that much change, it’s not a shock that City’s highly specific and complex rotations aren’t quite as instinctive as they’re supposed to be. We saw this with Wolves’ first goal. Foden has come inside, but Kovačić hasn’t yet dropped deeper, so both players are occupying the same space, while Wolves are able to surround them both and cut off passing options. Two players in the same zone and no one to receive the ball is the antithesis of Guardiola’s ideas of structure in possession.

This confusion between Kovačić and Foden quickly causes them to lose the ball. Pedro Neto wins it and then drives down the right flank, doing a good job of beating Foden and Aké in the process.

After that, it’s just a normal defending mistake from Dias and Aké more than anything tactical. But when the Death Star is fully operational, City never give the ball away in the first place. Guardiola has worked miracles in his career at turning individual defenders previously thought to have a mistake into them into some of the most reliable back lines ever. The reason it works is because situations like this are never supposed to happen.

From there, City had a really hard time breaking down Wolves’ low block. I think it was a combination of some of their most creative players missing and this team not being 100% fluent in the movements and passing patterns. They just couldn’t quite manipulate space to get openings in dangerous areas. You can see from the passmap below, from Between the Posts, how often they just passed it around the back and out to Doku, who was the only outlet in space. Doku didn’t make great use of this, and too often made the quick decision to shoot from range. I think Guardiola would say his players were too rushed, and needed to be more patient in waiting for the openings to appear rather than forcing the issue. Whatever the causes, they led to a really poor attacking performance. City didn’t take a single shot with an xG value higher than 0.09.

Source: Between the Posts

This is the first time things have properly gone wrong but, if you look for them, there have been some small warning signs. Scoring five goals from seven shots against Fulham made up for the fact that seven really isn’t very many shots at all. The red card muddied things a lot against Nottingham Forest, but even with 11 players in the first half, 0.79 xG from six shots is not scintillating stuff, and they totally ceded control of the game to Forest in the second half with a man down. That’s three of City’s last four league games with at least some attacking issues. I’m not for a second suggesting City are anything other than overwhelming title favourites, but there are some kinks for Guardiola to work out here, and they’ve happened in an easy stretch of fixtures. In their first seven matches, they’ve played just one of the other six rich teams (Newcastle). In the next eight games, they will play all of the remaining five, plus Brighton and Aston Villa. I fully expect City to figure it out and retool this team to dominate those upcoming matches. But if they do stumble in the next couple of months, the warning signs were here.

2. This is Manchester United

Not exactly the shock of the century, this one.

Both of Manchester United’s left backs were unavailable, so Erik ten Hag was forced to get creative and play Sofyan Amrabat in that role. Mason Mount was fit enough to start (but seemingly not to finish) so he came into central midfield, while Facundo Pellistri got a chance on the right. Otherwise, it was a pretty typical Man Utd lineup.

Ten Hag really did try to fix how light the team have been in midfield. As is all the rage these days, Amrabat stepped up from left back and moved into a defensive midfield role next to Casemiro, giving United that very Guardiola-esque 3-2 defensive structure. This, in theory, would also allow Mount to get forward more often.

United started the game well. They were playing at a high tempo and getting it wide to shake up Crystal Palace’s compact defence. The first warning sign came 22 minutes in, when Amrabat was a little too keen to push forward and get the attack going, giving it away in a central area from which Palace could spring a counter into a lot of open space. Palace didn’t have the quality to cause United’s defenders a real problem, though. That changed two minutes later, Amrabat again looked uncomfortable in that position and gave away a cheap foul in a wide area. No one gets a head on the cross, which means it flashes all the way across the box only for Joachim Andersen to smash it first time. I do have some sympathy here. It’s a tight angle with a lot of bodies in the way. Andersen has to hit it absolutely perfectly on the half-volley to score. The other side of that is that when he does hit it perfectly, André Onana can’t get a clear look at the shot because of all the defenders in the way. I’m not too critical of United for conceding that goal.

Before watching this game back in full, I had assumed this would be another poor second half performance after the team went behind and confidence levels dropped. I don’t quite think that’s what happened. United didn’t “drop off” as such, and I don’t think they were mentally frail here. They just didn’t have the ideas to break down a very Roy Hodgson low-block Palace performance. They recognised that the space was in wide areas, but never had the quality to do the damage out there. Pellistri’s poor impact is understandable, but this was a game that needed a big Rashford performance. He still should have been more decisive, but the left back issue probably hurt him here. Rashford usually has a conventional attacking left back in Luke Shaw or Sergio Reguilón overlapping him. This lets him cut inside and use the full backs’ decoy runs to distract defenders. Against Palace, Amrabat was coming inside to help out the midfield, which meant Rashford was out on the left on his own. He had to start wider and often do it all himself. It’s the classic “short blanket” theory of football: in helping out Casemiro, Ten Hag took something away from Rashford. Fortunately for the Englishman, Reguilón should be fit again soon, and Amrabat didn’t do enough to deserve keeping his place at left back.

Man Utd weren’t terrible against Palace. They’ve played worse this season, and could’ve taken the three points on another day. But they were hardly impressive. Their 1.3 xG created was more than Palace but, playing most of the game 1-0 down and desperate for a goal, you’d really like to see more. According to Danny Page’s xG simulator, United fail to score from these shots 26% of the time. So they were creating, but it was hardly liquid football.

But once that Palace goal went in, the minimum requirement for a decent result was to score twice. They only created about 0.9 xG in this period, and Page’s model estimates that they’d turn it around from those shots just 25% of the time. It’s not great. When the team hasn’t been playing well, I can totally understand why Ten Hag is changing the system around and trying different things. But that does come with a cost. I don’t know what I’d do, but the solutions aren’t coming easily. He needs to find a coherent side and keep things settled for a run, but I don’t know what that coherent side would look like right now.

3. VAR does a thing, everyone argues about it

Tottenham’s 2-1 win over Liverpool was a lot of things. It was absolutely terrible if, like me, you wanted a better sense of how good either of these teams are, because the circumstances gradually became so strange that it’s impossible to make real sense of the match. But at least it gave everyone plenty to talk about. I really hate talking about refereeing decisions on the internet because no one can ever be normal about them, but it’s the biggest story of the weekend, so needs must.

Somehow, Liverpool going down to nine men isn’t even the biggest controversy. The Curtis Jones red is the type of incident where people talk about “intentionality”, something the rules aren’t really built around. Diogo Jota, meanwhile, was just being a hothead getting two yellows so quickly. I think the referee is within his rights to give both reds, so let’s move on.

In case you’ve been living under a rock, here’s what happened. Luis Díaz received the ball on the shoulder of the last man and put the ball in the net, but the flag went up for offside. VAR obviously checked the decision, and the footage clearly showed that Díaz was onside, meaning the goal should stand. Somehow, the referee and assistant in the VAR room seemed to believe incorrectly the on-field decision was for the goal to stand and, when seeing Díaz was on, just said “check complete” and on we went. The referee said the goal didn’t stand when the facts show it should have. PGMOL cited “human error”.

Simply saying “check complete” is such an obviously flawed method that I’m surprised it hasn’t caused more problems. Would “decision: onside, goal should stand” be difficult? Certain aspects of VAR feel like they have been a little too matey and comfy. A lot of it stems from the level of abuse referees receive. This runs at all levels, from amateurs dealing with screaming parents at kids’ games all the way to European finals. I’m sure TV coverage and social media makes it feel more uncomfortable though. I certainly wouldn’t want to be a Premier League referee. Former ref Mike Dean admitted that he once didn’t send on-field official Anthony Taylor to review an incident because “he is a mate as well as a referee, and I think I didn’t want to send him up because I didn’t want any more grief than he already had”. The way referees are abused is a disgrace, so I understand why they feel this in-group loyalty. But if the job is to be done properly, the system needs proper accountability without this level of solidarity. I don’t know how that looks in practice but it should be explored. Refereeing in the VAR era does need fundamental reviews, but that also includes fans and the media taking the heat down in terms of how we talk about it.

We’re living in an era where we’ve never had a higher percentage of decisions called correctly. Ten years ago, good goals like the Díaz strike were ruled offside all the time. Football for most of my life was dominated by referees seeing an incident in a split second and just guessing. That’s not how it is anymore, which is what everyone said they wanted. But it doesn’t seem to be making fans any happier. The remaining wrong decisions have been magnified and driven fans to even more conspiratorial thinking. This is the football every manager moaning about referees after a game wanted, and people hate it.

Football is not an ordered universe. It’s a place where chaos ensues. Higher-scoring sports produce a larger sample size, which in turn reduces variance. In football, any number of ridiculous events can and do dictate match results. This is a game where you can watch your team dominate a game for 90 minutes and lose because someone threw a beach ball on the pitch. That’s why it’s the most popular sport in the world: it makes you want to kill yourself.

I’m always surprised that most fans seem to miss this basic point. The game is the game. If you want to see fair results and the best teams winning, go and watch another sport. We’re trying to instil order in a chaotic universe and it’s naturally infuriating everyone.