Chelsea: Comparing Edouard Mendy to Premier League rivals

Chelsea’s new keeper Edouard Mendy has already impressed Blues fans, but how does he stack up compared to the rest of the Premier League’s keepers?

How is a keeper’s ability assessed? There are a number of factors that go into evaluating how good a goalkeeper is. In the modern era, the role of the goalkeeper has evolved from pure shot-stopping one to a range of different roles. As teams increasingly look to play out from the back and employ a high defensive line, it is important for a goalkeeper to be good with his feet.

Context is everything in football and how “good” a goalkeeper is a function of the team’s preferred playstyle and how the manager sees the role of his goalkeeper, not only in terms of shot-stopping, but also in dominating the box and being part of build-up play.

In this article, data from last season is used to assess the quality of Chelsea’s new goalkeeper Edouard Mendy. Three criteria will be used to create a ranking system: shot-stopping ability, behaviour in the boxed based on claim activity, and distribution.

Shot stopping

Ted Knutson of Statsbomb devised two metrics to assess shot stopping ability mainly, the goals saved above average (GSSA) and the adjusted save percentage (aSV%).

GSSA is the difference between the goals conceded and the xG conceded (based off the quality of shots the keeper had to face). Mendy conceded 15 goals last season to his xG conceded of 16.37, giving him as GSSA of 1.37. Kepa Arrizabalaga, on the other hand, conceded 37 goals compared to his xG conceded of 27.8 giving him a GSSA of -9.16.

The aSV% is then calculated by dividing the GSSA over the shots faced on target. In Mendy’s case that was two percent last season. Arrizabalaga’s aSV% was -10 percent.

Ranking Mendy in the Premier League

In order to rank where Mendy would stand in the league for shot-stopping ability, these metrics were calculated for all the Premier League goalkeepers last season.

The table above shows the goals, xG goals against and shots against faced by the goalkeepers in the Premier League as well as Mendy’s statistics for his season in France. The adjusted save percentage and GSAA were calculated for all the goalkeepers (see table below).

Based on adjusted save percentage, Mendy would have ranked eighth in the Premier League last season. Arrizabalaga was last based on these metrics, suggesting Mendy would be a significant upgrade.

Of course, this is not adjusted for factors such as difference is league competitiveness as the Premier League is arguably a more intense league to be a goalkeeper in.  However, the difference in quality between the goalkeepers based on these metrics is stark. This suggests that even if Mendy has a dip in performance as a result of playing in a better league or taking time to adjust, he would likely still perform better than Arrizabalaga.

*Note the x-axis is inverted with highest goals conceded closer to the left and highest goals conceded closer to the right 

The graph above is a visual representation of adjusted save percentage to goals for all the goalkeepers in the Premier League, some of the world’s best keepers in Europe, and Mendy.

Claims activity in the box

Another important thing to evaluate is the goalkeeper’s profile in the box in terms of coming off his line to claim crosses and how often the keeper punches the ball versus claims it.

Coming off the line to claim the crosses is measured by a metric called exits per 90. It tells us how many times on average per 90 minutes the keeper exits/leaves his line to go for the ball. This is plotted on the X-axis above. A high exit per 90 gives some indication of how dominant the keeper is from set-pieces and crosses.

On the Y-axis, we have the claims/punches ratio (i.e. for every one punch how many times does he claim the ball). Ideally, a better keeper would have a higher claims/punches ratio as that ensures more safety.

Mendy comes off his line a fair bit over 2.2 times per 90, one of the highest compared to the Premier League and the world’s best keepers. Of course, this could be due to the fact that he played in a team that was midtable in France and thus faced a lot of situations where he had to come off his line to claim/punch balls into his box from crosses and set pieces.

Distribution

There a lot of ways to measure the passing ability for a goalkeeper. To keep it simple, we can measure pass accuracy percentage to the average pass length.

Arrizabalaga outperforms Mendy in terms of pass accuracy percentage, although his average pass length is lower than Mendy’s. So, if there’s anything Chelsea fans will miss from Arrizabalaga, it is his outstanding pass accuracy.

Conclusion

There are many more ways to assess a goalkeeper and as mentioned in the introduction, context is crucial. A Pep Guardiola team requires a different goalkeeper to Burnley for instance. Attributes are valued differently based on playing system and managerial preference and expectations.

In Chelsea’s case, given Frank Lampard’s high pressing system, a goalkeeper who can come out of the box and play a sweeper-keeper role could be important. Further, given Lampard’s preference for playing out from the back someone who is good with their feet is also crucial.

Arrizabalaga seems to be good with his feet even though he makes a few silly errors (for example versus Liverpool). However, a goalkeeper’s primary job when all said and done is to stop the ball from going into the back of the net.  This is where Arrizabalaga has been rather underwhelming.

Mendy may or may not be the right man long term in terms of distribution and performing other roles. However, in the short term, he looks to be the right man to solve two major problems the Blue’s faced last season: shot-stopping and being a dominant aerial presence in the box.

All data is from Wyscout. The model is based off this Statsbomb article by Ted Knutson. All graphs are creations of Siddhanth Khanvilkar.