How bad do Alexis Sánchez’s giveaways hurt Arsenal?
After an Arsenal match, especially one where Arsenal lose and Alexis Sánchez fails to score (and even then it might not get him off the hook), you will find people bemoaning his style of play.
He will be called selfish, wasteful and all sorts of other things, as fans and pundits complain about the number of times that he loses the ball with a bad touch, how often he is dispossessed trying to dribble, and the number of times he fails to complete passes.
I have always accepted that this is his style of play, and that it’s something that you have to live with to be able to enjoy the world class output that he can produce. I certainly get frustrated when his chipped through balls don’t come off, or when he tries to take on two to three defenders and loses the ball, but I believe that the times his actions do work outweigh the times that they do not.
After seeing so much frustration with him losing the ball, I started to question if maybe my hypothesis was wrong, so I have set out to try to answer the question “how badly does Alexis hurt the team with his turnovers?”
Creating the test
This next section shows the goal added/subtracted values for completing passes, dribbles and turning over the ball on the different areas of the pitch.
The top numbers are for distance from the opponent’s endline, with smaller numbers closer. The numbers on the left show the distances from sideline to sideline with 50 the middle of the field. 0 the right (facing the attacking direction) and 100 the left.
To determine the values for every area of the pitch, I have re-purposed the work I had done previously for determining value added passing.
There were modifications because dribbles and turnovers can happen in just one area and can also not involve moving the ball a significant distance, typically the case with passing. They also happen less often, but when they do the single value of the action is much higher than a typical pass.
The vaules reflect what you would expect: it is more valuable to complete a dribble close to your own goal and in the more central part of the pitch. The opposite is, also somewhat intuitively, true for losing the ball with turnovers in the defensive third, a much a higher negative value than in the final third.
To measure overall value added and lost, the total of all actions are summed.
For passing, the short version is that the starting value of the pass is subtracted from the end value with a bonus of 0.003 added for maintaining possession. If a pass is not completed, the value of where the opponent takes over is subtracted with an additional 0.015 subtracted for losing possession.
For a completed dribble, the player is credited for the value of that zone plus 0.005 for completing the dribble. For a possession loss it is the same, the negative value is taken away from the player.
Looking at Alexis
What I found looking through the data is that it looks like Alexis turns the ball over a lot - so far this season, he has lost the ball to a bad touch 38 times and he has been dispossessed 27 times. He has also turned the ball over via bad passes an additional 160 times.
Losing the ball 225 times sounds like a lot (and is frustrating to watch) but this is where context can be helpful; that only ranks him 17th highest in the Premier League, and 4th among players who play in attacking positions. Kevin De Bruyne actually leads the Premier League in total turnovers, which sounds crazy, but is less so when you also see that he leads the league in total touches.
If we want to attempt to control for players with more possession, we can look at times turning over per time on the ball. In this metric, Alexis turns the ball over 31% of the time. This ranks 63rd in the Premier League, and 39th among attacking players.
His turnover percentage compared to other forwards and attacking midfielders is a bit higher than the league average of 27.7% , but just slightly, ranking in the 63rd percentile in this measure.
Now that we have established that Alexis turns the ball over often (but not out of line with other attacking players) the next step is to try to determine a cost benefit analysis of the positives he brings to the Arsenal attack and take out the negatives that his lose possession style takes away.
First up passing:
While some may get frustrated with his repeated attempts to try hard passes, there can be little doubt that when they come off they are incredibly valuable for Arsenal. The above chart shows the start point for each of his positive passes played this season, with the size and color representing how much value they added to the Arsenal attack.
What is apparent right away is that his value comes from completing passes in the final third. Overall in value added passing he is 3rd among Arsenal players with a value of 4.2 and 14th overall in the Premier League. The overall positive value created (taking out negative passes which we will get to in a second) he has created 5.9 in positive passing value.
He has completed 36 passes with a value of at least 0.04 (the cutoff I use as a high value added pass) which ranks 7th in the Premier League and is second behind Mesut Özil with 38 for Arsenal. Controlling for minutes played, he completes 3.7 high value passes per 90 minutes.
Looking at the other side of the ledger, the vast majority of his misplayed passes - 93%, or 149/160 - happen in the opponents half of the field. Of these a further 82% happen in the final third (122 of 149). Overall his bad passes have averaged 76.5 yards away from the goal that Arsenal are defending.
What this means is that while giving the ball away is never great, the further you do it from the goal that you are defending the less bad each turnover is. This season Alexis has just one “horrible” pass (PPVA less than -0.04), which is tied for least among Arsenal players with at least 300 minutes played.
Looking at “bad” passes (PPVA less than -0.015 but greater than -0.04), he has only 7 of these this season, which just is behind Danny Welbeck (2) and Alex Iwobi (5).
While Alexis adds quite a bit of value with his passing, he is also one of the better dribblers in the Premier League.
This season he ranks third in dribble value added, even though he only ranks 16th in total successful dribbles with 38. His dribble value added of 4.45 ranks only behind Eden Hazard and Andros Townsend, and both have nearly double the amount of completed dribbles at 70 this season.
Looking at dribbles with a high value (dribble value > 0.1) he is second behind Hazard (30) with 15 this season, which represent 40% of his total dribbles this season. This shows that he attempts high reward dribbles close to goal.
Alexis has been dispossessed 27 times this season, and these have happened an average of 72.5 yards away from the goal Arsenal are defending. While the majority of them have been in the attacking half, seven have happened in his own half, which is tied for 7th most in the Premier League, with three of these happening in the defensive third.
Overall his dispossessions have cost Arsenal -0.98 in value, which is 10th worst in the Premier League.
Alexis has had 38 bad touches, and these have happened on average 71.4 yards away from the goal Arsenal are defending. Eight of these bad touches have happened in the defensive half, with one really hurtful bad touch that happened on the edge of Arsenal’s 18 yard box against Tottenham that was thankfully hastily cleared away by Shkodran Mustafi.
Overall his bad touches have cost Arsenal -1.11 in value, which is 8th worst in the Premier League. His overall negative value taken away from Arsenal is 4.01 which is 16th worst in the Premier League.
Conclusions: Alexis adds a ton of value
Overall, Alexis has added 10.4 xG in positive value through passing and dribbling. He has added an additional 5.73 in xG for the shots that he has taken this season. His overall positive value added to Arsenal this season is 16.1, which is 4th best in the league, on a per 90 basis he has produced 1.63 in positive value which is 2nd best in the league behind only Eden Hazard.
Controlling for the amount of touches each player gets, he ranks 10th in positive value per 100 times on the ball with 2.2 positive value added.
Looking at the net value (positives minus negatives) Sanchez has produced 6.4 overall xG value added through passing and dribbling. Including his shot values, his overall net value added to Arsenal this season is 12.1 which is 4th best in the league.
Controlling for the amount of touches each player gets, he ranks 9th in net value added per 100 times on the ball with 1.7 net value added.
What all of this means is that while Alexis’ style of play that tries high reward plays that have a lower success rate than safer options, he makes enough of them come off to be a very valuable player.
This post originally ran on The Short Fuse