Joe Hart vs Claudio Bravo: Shot Stopping

Subjective criticism of Pep Guardiola’s tactics and team selection has been a constant theme this season in the English football media. One of the major issues has been his selection of Claudio Bravo as first choice keeper after he allowed Joe Hart to leave on loan.

After a loss to Leicester this weekend, Alan Shearer gave a subjective opinion:

“Yes [Hart] had not had a great Euros but this was still a top class goalkeeper who was one of the reasons Manchester City won two titles…To get [Man City] to play like Barcelona and Bayern they needed a keeper with a different skill set it seemed, so in came Claudio Bravo. He certainly does have a different skill set, but saving shots does not seem to be part of it. After the last two games there are now even bigger question marks over the signing of Bravo.”

Shearer says savings shots is not part of Bravo’s “skill set,” and implies Hart is top class.

Other writers have tried, and failed, to use statistics to prove their narrative that Hart is a better shot stopper than Bravo. For example, Alaistair Tweedale shows he either has no idea about sample sizes and/or shot quality, or just knows he can get away without using them:

The biggest worry of all has to be Bravo’s shot-stopping, though. While in Premier League games he has saved 73.7 percent of the shots on target he has faced – which puts him in mid-table among his peers – that rate drops to just 42.9 per cent in the Champions League. That is, he has made only three saves in three European games since moving to City. Hart has kept out 79.2 percent of the shots on target he has faced since moving to Italy.”

Those numbers inform nothing, unfortunately. For one, they don’t take into account location of shot. That is important – not all shots are equal. Also those percentages come from small sample sizes, especially the Champions league rate.

Paul Merson has added into the debate with his opinion of Bravo and Liverpool’s Karius in comparison to Hart:

“What chance do [England] have? It’s crazy, it’s an absolute joke. Our best goalkeeper has to play abroad to get a game. And we have two goalkeepers in the top four who can’t catch a cold.”

Is this narrative true? Who is the better shot stopper, Bravo or Hart?



I went back as far as I could go in my data, to the 2014-2015 season. I looked at 7 different types of save rates for each player across their matches in La Liga, the English Premier League, and Italian Serie A. I am defining a save rate as saves over shot on target including goals.



For all categories, Bravo has saved shots at a higher rate than Hart over the past 2.5 seasons. However, in only two categories was there a statistically significant difference in save rates between the two keepers.

Based on these results I would say there is some evidence that Bravo is a better shot stopper than Hart at high shots, and shots to his side. I would say there is no evidence that Hart is a better shot stopper in any of these categories.

There doesn’t seem to be evidence to the narrative that Hart is a better shot stopper than Bravo.

Future Extensions:

Given more data, I would go back further in time to increase sample size. Interestingly, Claudio Bravo and Joe Hart have both played in the in a Top 5 European since the 2006-2007 season. Although, Hart only had one start in 2006-2007, and Bravo had no starts in 2007-2008.

There is a downside to going back too far in the data, though. The further we go back in time, the more chance we have to find a statistically significant difference. However, the further we go back in time, the further away we are from answering the question, ‘Who is the better shot stopper today?” I think the 2.5 seasons of data I have used is reasonable, but I at least would like to run the numbers for further look backs.

Note – All images were sourced using Google Advances Image Search option with image rights set to ‘free to user or share.’  If Google’s classification was incorrect and you would like your image removed please contact me an I will do so immediately.