Pep Guardiola’s first team selection caused more than a few raised eyebrows, with the former Barcelona and Bayern manager dropping Yaya Toure, playing Alexander Kolorov at centre back and opting for Willy Caballero over Joe Hart in goal. Guardiola’s main reason for leaving out Joe Hart was his passing ability, so how does Hart compare to other keepers?
First let’s compare Joe Hart’s passing statistics with those of Guardiola’s previous keepers Victor Valdes and Manuel Neuer. Figure 1 shows the average pass completion for each keeper over the past seven seasons. From this we can see Hart has had the lowest passing success of any of the three keepers. The impact of Guardiola's playing style can be seen in the graph from the passing successes of Valdes and Neuer. Valdes' passing success dropped from around 80% from 2009-12 (under Guardiola), to just above 60% for the following two seasons (post Guardiola). Neuer's passing success jumped from 70 and 75% prior to Guardiola manging Bayern to 88, 86 and 80% under Guardiola's management. So there's an argument that we would expect Hart's passing success to improve to from its baseline of around 50% under the Spainiard.
Perhaps the reason for this can be explained by team's playing a greater proportion of shorter passes, which are more likely to reach there target, under Guardiola. We see thaat Neuer played an average of over 20 short passes per game each season under Guardiola compared to 9.2 and 13 per games in the two previous seasons at Bayern (see Figure 2).
In terms of long balls there is no real pattern in terms of the number of accurate and inaccurate long balls under Guardiola. It is interesting that Hart has played a higher number of inaccurate long balls than the other two keepers but this could be a result of the characterisitcs of the different leagues.
So how does Hart compare to the other Premier League keepers? Last season Hart's passing success was above the average (50.04%) but was the 9th highest in the league (for keepers who had played ten or more games), see Figure 3.
Also compared to other Premier League keepers Hart played one of the lowest numbers of long balls (both accurate and inaccurate), Figure 4. In fact Hart played the lowest number of average passes per game (21.3) in the league last season. This may have been a tactical ploy on City's behalf, perhaps they knew Hart's ability on the ball was lacking or may be they had more control of games so needed to utilise Hart less.
From the above analysis it is clear that Hart's passing game could do with some work compared to both Guardiola's previous keepers and also other keepers in the Premier League. Encouraging shorter passing and getting Hart on the ball more would likely see his passing average improve. It's likely City will build from the back, especially with John Stones' ability on the ball so don't be too surprised if we see this happen.
Of course it's also possible that Guardiola simply dropped Hart to try and get the best out of him after a disappointing Euro 2016, Sunderland at home was never likely to be the most testing game for Caballero and on top of that Caballero looked far from convincing with his kicking - I would not be at all surprised if we see Hart back in goal much sooner than some expect.
Data courtesy of whoscored.com