Arsenal and luck
3 min read

Arsenal and luck

Arsenal and luck

The above tweet is going viral and while it is true I think it does miss some context and that is what I am going to look at today. My xG for these matches are slightly different with 7.1 xG for Arsenal and 4.5 xG against.

First let's look at the probability of Arsenal's actual goal scored and conceded based on the individual shot quality.

Arsenal are on the wrong side of the distribution for both.

Giving up 8 goals is a little more normal, with this expected to happen about 3% of the time that a team would concede these types of chances. This is 2.7 standard deviations from the mean.

Arsenal scoring just 2 goals from these shots is less likely to happen. Given this shot quality and volume only scoring twice is expected to happen just 1% of the time. This is 2.95 standard deviations from the mean.

So yeah that's not good.

Next let's look a little deeper at how Arsenal got to this point based on their game state.

Even Game State

With the score tied Arsenal have taken 18 shots, created 2.18 xG, and scored 1 goal. Arsenal have allowed 16 shots, 1.64 xG against, and 4 goals.

Broken out by match:

Liverpool: 7 shots, 0.6 xG | 4 shots, 0.2 xGA

Aston Villa: 4 shots, 0.7 xG | 0 shots, 0 xGA

Crystal Palace: 0 shots, 0 xG | 2 shots, 0.2 xGA

Brighton: 3 shots, 0.5 xG | 3 shots, 0.4 xGA

Southampton: 4 shots, 0.4 xG | 7 shots, 0.8 xGA

Things with the score tied are just okay. Arsenal are doing better than their opponents but not exactly tearing up trees. We are also seeing perhaps Aaron Ramsdale go through a cold patch with the xG based on shot placement for the shots on target coming in at 2.4 while he has given up 4 goals. Arsenal have also seen the opposing goalkeepers do well, with Arsenal getting 38% of their shots on target (average is around 30%) and producing a shot placement xG of 2.0

Losing Game State

With Arsenal losing this where they have created the vast majority of their xG and shots for this period. They have taken 50 shots, created 4.54 xG, and scored 1 goal. Arsenal have allowed 16 shots, 2.58 xG against, and 4 goals (including one penalty).

Broken out by match:

Liverpool: 2 shots, 0.1 xG | 5 shots, 1.1 xGA

Aston Villa: never losing

Crystal Palace: 12 shots, 0.9 xG | 4 shots, 1.1 xGA (including Penalty)

Brighton: 17 shots, 2.5 xG | 5 shots, 0.3 xGA

Southampton: 19 shots, 1.0 xG | 2 shots, 0.1 xGA

Over the last 3 matches this is where Arsenal have really accumulated their stats. They have taken 48 shots while losing, with half of those shots coming in the final 15 minutes of the matches (75 minutes or later). These are mostly desperation shots against defenses that are sitting deep looking to not give up space.

So yeah Arsenal have taken a lot of shots but they have only created 3 "good" chances and even these might be artifacts of my xG model not realizing that some of shots close to goal were actually tough.

Arsenal have been "unlucky" but they have also not played all that well. This is a team at the moment that feels like it needs to score the first goal to have any chance to win a match.