As such, you must also consider how the composition of pass type (long/short) varies between goalies. Only 6.2% of Adam Federici's total passes attempted were short while, on the other end of the spectrum, 45.5% of Michael Vorm's passes were short. Below is a table illustrating the different passing style of each EPL goalie. This is sorted by overall passing %.
A couple things immediately jump out. One is the direct relationship between the % of short passes and the short passing%. This is interesting, though somewhat intuitive: the fewer short passes a goalie attempts the lower their completion percent; they are likely under defensive duress, which is the same reason why the majority of their passes are long balls.
I ran a regression between a player's overall passing completion % and the % of their passes that were long/short. The idea is simple: the more long passes you attempt relative to short passes the lower your overall passing %, and vice versa. The relationship was quite strong for only 20 observations.
|Adjusted R Square||0.701239|
With regression in hand, the next step is figuring out which goalies over and under-performed their expected overall passing completion percentage with a crude predicted pass completion % model. Here are the EPL goalies ranked from best to worst passer. Begovic is the best over-performer and Julio Cesar is the biggest under-performer.