This match could not have gone worse for Arsenal and more to the plan for Jose Mourinho.
I’ve spent a day away from this and I am still very frustrated by this performance and unfortunately, I don’t think that there are any quick fixes that are on the horizon to make things better at Arsenal.
Sp*rs 2-0 Arsenal: By the graphs
Sp*rs 2-0 Arsenal: By the numbers
656 – Minutes in the Premier League since Arsenal scored a goal from open play.
46 – Shots in the Premier League singe Arsenal scored a goal from open play. Good for 6.3 open play shots per 90, which would rank 17th in the League only ahead of Burnley, West Brom and Newstacle.
4.3 – Expected goals from the shots from open play, since Arsenal last scored a goal from open play. Good for 0.6 open play xG per 90, which would rank 19th in the Premier League ahead of only West Brom.
5 – Open play Big Chances since Arsenal last scored a goal from open play. Good for 0.7 open play big chances per 90, which would rank last in the Premier League.
It has been a long time since Arsenal last scored a goal from open play in the Premier League, with the last goal was from Nicolas Pepe against Sheffield United in the 64th minute. A quick aside, remember how excited it felt to score 2 goals in less than 5 minutes? It felt like Mikel Arteta had figured something out in the attack, with the Saka/Auba/Pepe front 3, if we had only known things were going to fall apart after that.
For those wondering the Gabriel goal against Wolves is still classified as “from a corner” because the play never reset but also shows that not everything is cut and dry with statistics. If you want to count that, that’s fine but that one goal really doesn’t change the inept attack that we have seen from Arsenal over the last 7+ matches at creating from open play.
That brings us to the crosses.
44 – Crosses attempted against Spurs
9 – Crosses completed, good for a 20% cross completion percentage below the Premier League Average of 22% but better than Arsenal’s previous completion rate of 17%.
32 – Crosses attempted from open play
7 – Open play crosses completed, of which 3 became shots (1%), which had an xG value of 0.14
81% – The number of crosses that were “chipped” or hit into the air from open play. In general, these are lower quality chances with an average xG of 0.1 vs 0.15 for low crosses. I kind of joked during the match that Arsenal should have put one of their centerbacks upfront because of this, in hindsight, it probably wouldn’t have been a bad idea because with Alexandre Lacazette (career Aerial duel win percentage of 35%) and Pierre-Emerick Aubameyang (career Aerial duel win percentage of 33%) as the guys being aimed for wasn’t likely to end well.
178 – Crosses that Arsenal would have needed to attempt from open play to tally 1 expected goal, given the same conversion rate into shots and a pretty generous average xG of 0.06 (the actual average xG of Arsenal’s shots assisted by crosses was 0.047).
22.5 – The probability of creating a “Clear Cut Chance” with an xG greater than 0.4 from the crosses that Arsenal attempted
When you look at the possession numbers that Arsenal had in this match, you might come away thinking that Arsenal put that team under a lot of pressure. Arsenal had 52 touches within 25 yards of their goal, they put up an impressive “non-shot xG” (a way of measuring how dangerous possession a team has) that was over 2 but in reality, watching this match you could see that it was mostly comfortable to defend and especially the final balls that Arsenal tried.
Most of that comes to the crosses that Arsenal attempted in this match. The headline number of 44 seems really high (because it is) but even worse is that three-quarters of that came from open play attempts. For these crosses, Arsenal had held the ball for an average of 20 or more seconds before they attempted the cross, in the second half that average creeps up even higher to almost 25 seconds of possession before attempting a cross.
One of the things that can make crossing more effective is if the cross comes into a defense that isn’t set, this was not the case for the majority of the crosses that Arsenal attempted in this match.
It is easy for me to sit here writing about why this is a low reward strategy without offering any alternatives to creating things. Unfortunately looking at the players that can play in the Premier League in the squad, there isn’t a lot in the way of players on this team that would even be willing to attempt risky passes into box or make runs from deep. I think that a lot of this ultimately falls on the way that this squad has been constructed and that fixing it is not going to be easy.
The Aubameyang worry Level rises
2 – Shots in this match, 0 on target. This was the 8th time this season that he has had 2 or less shots in a match.
19 – Shots total that he has in the Primer League, this ranks tied for 31st.
0 – Goals in this match, the 10th time this season that he has failed to score a goal (non-penalty).
0.09 – Non-Penalty goals per 90 in the Premier League, this ranks 111th.
0.2 – Expected goals in this match. This was the 7th match that he has had 0.2 expected goals of less in a match. (0.2 xG per 90 for context would rank at around the 10th percentile for a striker).
1.8 – Non-Penalty expected goals in the Premier League, this ranks tied for 40th.
0.17 – Non-Penalty xG per 90 in the Premier League, this ranks 56th.
0 – Key passes in this match. This was the 4th time this season that he did not create a shot for a teamate.
1.3 – Expected goals assisted in the Premier League, this ranks tied for 38th.
It might be a bit simplistic to only look at things like shots and shot assists but for a player making big money as a forward, that is really what you are looking for from that type of player.
After the match against Wolves I had written that I was somewhat hopeful that his performance might have meant that there were “green shoots” for him. Unfortunately that did not carry over into this performance and my worry level is rising. I was cautiously optimistic that extending Aubameyang’s contract wouldn’t be a bad idea but now I think that I was too optimistic in that his dip in his underlying numbers was a blip and not the first red flags of a decline. That was probably a mistake and one where because I so like watching the player play that I overlooked the evidence that this was a very real possibility.
It is possible that this is still just a slump and that he will come out of it looking something close to his former self but the longer he goes without showing that the less likely it becomes.
P.S. thanks to everyone who participated in the approval rating survey, the results are here but I think that people would probably be a lot less approving if they were asked today.
Sources: Opta via WhoScored and my own database. StatsBomb via FBRef.