Claudio Bravo: Man City's goalkeeper analysed by Opta scientists
For many, Claudio Bravo's first five months in English football were neatly summed up by the caller to the BBC's 606 phone-in last weekend who dubbed the Manchester City goalkeeper a 'false no. 1'.
The 33-year-old Chile international arrived with a reputation for fancy footwork but is now, to some, thought of as being a shot-stopper who does not stop shots - something Everton took full advantage of when they scored with every effort they had on target in Sunday's 4-0 win.
City's defence admittedly did not do much to help him at Goodison Park, but Bravo has now conceded a total of 23 goals from the 57 shots on target he has faced in the Premier League this season.
"It looks like people are playing City and thinking if they hit the target, they will score," MOTD2 pundit Phil Neville said in his analysis after watching Bravo's latest less than convincing performance.
"It is not even as if they are all going into the corners of the net - he is being beaten in the central areas of his goal too."
Do the stats back up that belief Bravo should have done better? According to Tom Worville, a data scientist at football analysts Opta, the answer is a resounding yes.
Using an historical database built using information such as the build-up, distance, angle and placement of more than 250,000 shots, Opta can evaluate 'Expected Goals' - the quality of any chance a player has, and how likely it is to be saved.
Or, in Bravo's case, whether it should have been...
How bad is Bravo then? 'The figures are pretty damning'
Based on the historic quality of those 57 shots on target against City, only 16 should have beaten Bravo, not 23.
The £15.4m summer signing from Barcelona has conceded 6.7 goals more than he should have done, putting him in the bottom five of the 23 Premier League keepers to have faced more than 100 shots in 2016-17.
|PL keepers to have faced more than 100 shots in 2016-17 - the worst five|
|Rank||Keeper/club||Expected goals conceded||Actual goals conceded||Additional goals conceded|
|19/23||Fraser Forster (Southampton)||19.7||26||6.3|
|20/23||Claudio Bravo (Man City)||16.3||23||6.7|
|21/23||Lukasz Fabianski (Swansea)||42.2||49||6.8|
|22/23||Ron-Robert Zieler (Leicester)||7.7||16||8.3|
|23/23||David Marshall (Hull)||26.3||37||9.7|
It gets worse for Bravo when Opta focus on his current eight-game run that has seen him concede 14 goals from 22 shots on target.
"Of those 22 shots, we get an 'Expected Goals' figure of 7.33 - so he has conceded nearly double than expected in that time," added Worville. "It's a huge deficit, and the figures are pretty damning.
"I'm hesitant to jump to the conclusion that Bravo is a poor goalkeeper - we know that he isn't - but he's having a terrible time at the moment."
Most keepers are having a better time right now, including all of Bravo's counterparts at other top-six clubs.
It is a 12-goal swing from Bravo's figures to those of Burnley's Tom Heaton, the Premier League's top performer by Opta's calculations.
Heaton has saved almost five 'expected' goals this season - enough to give him the edge over the likes of Tottenham's Hugo Lloris and Arsenal's Petr Cech, who average around three apiece.
David de Gea has 'saved' one goal for Manchester United while Liverpool's Simon Mignolet has conceded one more than expected. All of them are doing much better than Bravo, though.
|PL keepers to have faced more than 100 shots in 2016-17 - the best five|
|Rank||Keeper/club||Expected goals conceded||Actual goals conceded||Additional goals prevented|
|1/23||Tom Heaton (Burnley)||31.9||27||4.9|
|2/23||Eldin Jakupovic (Hull)||11.2||8||3.2|
|3/23||Hugo Lloris (Tottenham)||16.1||13||3.1|
|4/23||Petr Cech (Arsenal)||24.6||22||2.6|
|5/23||Kasper Schmeichel (Leicester)||20||18||2.0|
Would Joe Hart be doing any better?
Few City fans thought that allowing Joe Hart to leave for a season in Serie A was a good idea back in August. It is possible only Pep Guardiola thinks that is the case now.
Over in Italy, Hart has had a positive effect to the Torino backline, despite conceding the same amount of goals and playing the same number of minutes as Bravo at City.
|Bravo v Hart|
|Claudio Bravo||2016-17||Joe Hart|
|57||On target shots faced||81|
|16.3||Expected goals conceded||24.5|
|23||Actual goals conceded||23|
|6.7 conceded||Additional goals prevented/conceded||1.5 prevented|
Bravo has delivered in one area for City, however, and it is the main reason they signed him in the first place.
His passing accuracy in his 18 Premier League games so far is 71.5%, ranked second only to Lloris. That might provide some solace to Guardiola, although even he would probably prefer it if his keeper started making some saves.
"Considering Joe Hart has conceded about one goal fewer than expected, it's easy to jump to the conclusion that with him between the sticks, City may have conceded fewer this season," Worville explained.
"Hart's playing behind a different defence though, which definitely has an impact on the quality of shots that he's facing.
"Additionally there's a lot more to goalkeeping than making stops, with the distribution side of things being one of the main reasons Bravo was bought back in the summer.
"Having Hart in the team may have lead to fewer goals conceded, but at the same time it may have impacted the style of play Pep wanted to implement."