The context: Arsenal are in a massive slump all over the pitch
This is a long article, I’ll save you my usual bloviating at the start and just get stuck in with Arsenal’s team stats over the last three seasons:
Arsenal passes completed per game are on a decline from Wenger’s last season, from 526.9 per game to 437.6 per game. And since you will ask, things have gotten worse in the last 7 games under Arteta and Arsenal are only averaging 326 passes per game. Those last 7 matches have been so bad from a possession standpoint that it’s dragging down Arsenal’s season average from 481 (which is what it was last season) to 438. Still, 481 passes completed compared to 527 passes completed is a 9% decrease in passes completed.
The decline in passes completed is coupled with an even larger decline in passes completed in the opponent’s final third:
In Wenger’s last season Arsenal completed 46.3 passes per game in the opponent’s area and in Emery’s first season that number dropped to 33.6 – a decline of 27% in one season. That number took a further tumble this season where Arsenal are now below Brighton and Hove Albion in number of final third passes completed per game.
As much as we like to ridicule Manchester United as a counter-attacking team they get the ball forward a lot more often than Arsenal do this season.
If you can’t get the ball forward, you probably aren’t taking a lot of shots either. Since we aren’t getting the ball forward as much as before shots per game are also way down from Wenger’s worst ever season:
Arsenal’s attack dropped 4.6 shots per game over the last two seasons. Arsenal were 5th in overall shots per game in Wenger’s last season and they are now 14th, just a few shots per game above relegation-bound Bournemouth.
What’s worse is that the quality has dried up as well. Arsenal got 6.0 shots on goal per game in 2017/18. That was good enough to be 3rd in the League. This season, Arsenal’s 3.7 shots on goal per game is worse than every team but Bournemouth, Newcastle, Sheffield United, Watford, Crystal Palace, and Burnley.
Just to put it bluntly, Arsenal have managed to get 92 shots on goal this season but have allowed 123. Arsenal concede 1.24 more shots on target than they create per game.
And finally, if all of the other attack metrics are down, it shouldn’t be a surprise when I say that goals, expected goals, assists, and expected assists are all down as well.
This shouldn’t be a surprise if you’ve been watching Arsenal this season. We simply struggle to get shots.
One complaint that fans often leveled at Wenger is that his team played too many sideways passes, or what people deride as “crab football”. But most top teams play that way with the purpose of probing to find space, or in order to move their opponent into and out of space so that they can create a chance in behind.
When Emery came to Arsenal, most fans wanted something different from the Wenger era, and they got it. Emery’s football wasn’t the same as Wenger’s (as we will see in a second) and also (as we see from the stats above) wasn’t able to create chances. Different? For sure.
Crab at the back
First, the baseline. I took every player who played more than 900 minutes and ran their completed passes per90. In 2017/18 Arsenal’s main two players on the ball were Xhaka and Ozil.
Xhaka led Arsenal in completed passes per90 with almost 74 passes per game while Ozil came in a respectable 2nd with 66 per game. Nacho Monreal was 3rd followed by the two main center backs, Koscielny and Mustafi. Below that, the other players all made around 50 passes per game. Even Iwobi, hated by a large and vocal segment of the fanbase, completed 48.5 passes per game.
Fast forward to Unai Emery’s first season at Arsenal and things have changed radically.
Guendouzi has replaced Ozil in the completed passes chart. Ozil hasn’t been completely shut out but he is below Lucas Torreira. And what’s more interesting is that Iwobi’s completed passes numbers have dropped from 48.5 to just 38.5.
Ozil’s main replacement was meant to be Mkhitaryan but he also produced just 36 completed passes per game. That’s a massive drop in possession from Ozil’s 66 per game the season before.
And this season is just terrible. Every player’s performance has plunged. Xhaka is off 12 passes per game, Guendouzi’s down 5 per, Torreira’s down 7, Ozil’s down another 7 from the previous season:
And when we look at each player’s progressive passing (I’m using that phrase to mean passes into the final third and penalty area), the decline from 2017/18 to 2019/20 is stark once again across the team. A note: these are per90 stats and I’m only selecting players with at least 900 minutes. First the 2017/18 season (Wenger’s final):
No surprises here that Xhaka leads the pack but it’s interesting that there is plenty of technical quality in Arsenal’s attack. Ozil, Alexis (who was traded in January 2019), Jack the Lad, Ramsey, and even Monreal added a lot of forward thrust to Arsenal.
That all changes. In Emery’s first season at Arsenal a lot remains from the Wenger approach with some notable exceptions:
Xhaka dropped a bit but literally every player at Arsenal got worse in terms of progressing the ball. Ozil’s numbers dropped significantly under Unai Emery and also Arsenal traded down from Alexis to Mkhitaryan. And then the middle of the pack saw a drop as well.
And this season things got worse:
Once again we see a drop in Arsenal’s progressive passing across the board. But another significant change from 2018/19 to 2019/20 is that Matteo Guendouzi became Arsenal’s most progressive passer. Ozil’s progressive passes have dropped in half from 10 per90 to 5 per90 in just two years. And we see a significant problem over three different seasons with the right side of Arsenal’s attack and after Monreal was sold, the left side as well.
Kolasinac’s final ball is great, but his buildup play is far behind someone like Monreal. Monreal was underrated by Arsenal supporters for his role in getting the ball forward for Arsenal. And more importantly, Arsenal have struggled to recover from the sales of Alexis, Ramsey, Monreal, and Iwobi.
The passmaps confirm this problem with Arsenal and progressing the ball. Here’s the most recent one from the Burnley match courtesy Between the Posts dot net.
Compare to Liverpool’s passmap against Arsenal:
When people complained about Arsenal’s “crab football” we were playing a lot like Liverpool played against us: high up the pitch, hemming the opponent in, searching for openings. What we are seeing with Arsenal now is that we are still being held at bay but that it is happening much further back. Arsenal play crab football, we just play it in our own half.
All of this data may seem like overkill but it’s necessary to explain the context in which Mesut Ozil plies his trade at Arsenal. If I just post Ozil’s numbers from the last three years the comments will overflow with “yabbuts”: people who say “yeah but…” and then ask me to look up hundreds of data points in order to satisfy their pre-formed notion that Ozil’s numbers are false, misleading, or that I haven’t thought of some angle.
The facts are that Mesut Ozil has just five (5) assists for Arsenal in his last 4119 minutes. That’s an assist rate of one every 824 minutes, or one every 9 full games. At that rate, if Mesut Ozil played every minute of a 38 game Premier League season, he would have 4 assists.
In a fascinating twist, Ozil has led Arsenal nearly every season since he’s been at the club in total key passes. That includes these last two seasons: he has 31 key passes this season and had 45 last season.
In 2017/18 he led us with 84 key passes, in 2016/17 he led Arsenal with 100, in 2015/16 he had 146, in 2014/15 he was 3rd to Alexis and Cazorla with 69 key passes, and in his first season at the club (2013/14) led Arsenal with 76 key passes. This is an incredible record to be quite honest. However, there has been a decline:
The question is, however, how much of that decline matches the overall decline of Arsenal’s attack?
Go back up to the top of this article. There you will see that Arsenal suffered a 27% decline in final third passes per game between 2017/18 and 2018/19: Ozil’s decline was 26%. So, pretty much exactly in line with the new manager’s (FUCKING AWFUL) playing style.
Arsenal slipped further between 2018/19 and 2019/20, 12% fewer total progressive passes than the season before. However, Ozil went from 7.72 to 5 progressive passes per90. That’s a 35% decrease over the previous season. Ozil’s sample size is small but there are other data points that are worrying as well.
Key Passes or Expected Assists: why not both?
Key passes are a decent measure of a player’s involvement in the attack but total key passes alone don’t tell us how good the passes were or how likely they were to score a goal. A layoff from 30 yards out that results in a 25 yard shot through a forest of defenders and a deadly cross in the 6 yard box that results in a tap in are not equivalent. That’s where expected assists comes in.
Unfortunately, Ozil’s expected assists per90 have also sharply declined:
That said, expected assists can also be a bit misleading. One problem with the “expected” series of stats is that they simply aggregate all chances. That means that folks with high numbers of low/medium percentage key passes or shots will often pad out their “expected” stats with just bulk creation. Creating a lot of low percentage chances still can add up to several expected assists. To combat that, I use expected assists per key pass. This is roughly equivalent to saying “on average what’s the return rate on this player’s key passes?”
Kevin de Bruyne leads the League this season in total key passes, so you should expect that he leads the League in expected assists, which he also does. And when I look at his expected assists per key pass this season it’s excellent. De Bruyne averages 0.16 expected assists per key pass. His teammate Riyadh Mahrez has an xA/KP rate of 0.19 this season. That is nearly a big chance created with every key pass on average. Trent Alexander-Arnold’s pretty good with a 0.12 xA/KP this season, Salah’s rate is 0.16, and both of the Silvas at City average 0.14.
Mesut Ozil’s xA/KP in 2019/20 is 0.05. So, while he’s leading the Gunners in key passes, he’s creating shots that aren’t much better than layoffs outside the 18 yard box, folks. And this is a major decline from last season’s 0.10 which is pretty average. In 31 key passes this season, Mesut Ozil hasn’t created a single big chance for his teammates. It’s not a surprise then that his total assists numbers are low.
We know from the data in the first section that Arsenal’s attack is bad: we struggle to get the ball forward, we struggle to create shots, and we play the ball around the back a lot. So, at least part of the reason for Ozil’s decline is explained by Arsenal’s mid-table attack. And Ozil himself shows signs that he’s also being forced to drop deeper.
Twitter user SMukherjee pointed out that Ozil’s heatmaps from the 2016/17 season to the 2019/20 season show a remarkable shift in average position on the pitch. Here’s Ozil in his pomp:
That red area is Ozil’s bread and butter zone. That’s the area between the defense’s two banks of four. Ozil receiving passes in that area forces the opponents to collapse on him, opening spaces for the attack to run into and behind. Now look at this season:
The contrast is stark. Ozil is being deployed wide and forced to come very deep in order to collect the ball. It’s no wonder that he’s struggling to create.
Now look at Lacazette’s heatmap this season:
Lacazette gets a lot of criticism for his goal drought this season but look at how many touches he’s taking in the center circle. That’s wild, especially compared to 2018/19:
I will leave the conclusions to you all and instead I will close by just saying this: I don’t dislike any of these players and I have a lot of faith in Mikel Arteta to get us back to where we need to be. This article shouldn’t be used to score points with people on the internet, but rather to see where Arsenal are, in terms of what is probably the worst Arsenal team that most of you all have ever seen.
This is what Arteta has to fix. This is the wreckage of the last two years.
Up the Arsenal.
And by “up” I mean return Arsenal to one of the best attacking teams in the League like they were under Arsene Wenger.