The eye test tells any Arsenal fan that goalkeeper Manuela Zinsberger has improved in every season that she has been with the club and that there has been an even bigger jump in output during this campaign. I decided to peruse the numbers (all via FBRef and WSL only) to see whether they backed up the eye test and found that they pretty much do.
First of all, the more routine stats are in Zinsberger’s favour, her goals against per 90 are down compared to her first two seasons even though the number of shots on target she is facing per 90 minutes has not notably moved, as the table below shows.
|Season||Goals against per 90||Shots on target faced per 90|
Zinsberger’s save percentage has also steadily climbed season on season and in this campaign, it has taken another steep climb. In fact, Manu has the best save percentage in the WSL this season (that ought to be taken in the context of Elle Roebuck’s injury, she usually tops this metric but hasn’t been able to play yet this season). Zinsberger has also kept seven clean sheets which, again, is the best in the WSL with Ann Katrin Berger in second place on five.
Of course, this level of data never tells the whole story, because conceding fewer goals can be just as much about improved defending as improved goalkeeping. If the shots on target are from less opportune positions due to better defensive positional play, then only partial credit is due to the goalkeeper. Though having the best shot save percentage in the division does suggest that the goalkeeper can take a good chunk of the plaudits for that data point. Bad goalkeepers don’t tend to do that.
A data point that can really help us to judge the quality of a goalkeeper’s contribution is ‘post shot XG’. This measures how likely a goalkeeper is to save a shot- think of it like XG in reverse, XG measures how likely the attacker is to score from a certain position based on historical data. Post shot XG measures how likely a goalkeeper is to save a shot based on the position of the attempt.
Now in isolation, post shot XG only tells us about the quality of the shots the team is facing and there is little a goalkeeper can do to influence that, that’s down to the defence. Arsenal’s post shot XG when Zinsberger has been in goal has not moved a lot during her Arsenal stint, measuring at 0.27 per 90 in her first season (so on average, Arsenal were conceding 0.27 XG per shot on target a game), then to 0.26 in her second campaign and to 0.28 this season.
🏆 The @BarclaysFAWSL Save(s) of the Season 😱
— Arsenal Women (@ArsenalWFC) May 26, 2021
To really gauge the quality of the goalkeeping, you subtract goals allowed from post shot XG and this calculates how many goals a goalkeeper is saving you over the season (or losing you!) based on the quality of the opponent’s shots. We have already established that Arsenal are giving up roughly the same quality of shots on target, so it’s easier to track whether there has been an improvement and, well, look at the table below. Any plus number shows you a goalkeeper is adding value.
|Season||Post shot XG – goals allowed||Post shot XG – goals allowed per 90|
In her first season, the data shows that Zinsberger allowed a goal and a half more than she ought to have over the whole year. Last season she saved Arsenal a touch under a goal. This season, that has shot up to 2.4 already at the halfway point of the season. If her current form holds, she is on course to save Arsenal five goals more than she would be expected to and if Arsenal win the WSL and the trend in the data continues, the Austrian will be in with a shout for player of the year.
Zinsi has always been known for her passing and her risk taking as a sweeper keeper. In her three seasons at Arsenal, she has completed 100% of her short passes (between 5-10 yards) and only misplaced 1% of her medium passes (between 15-30 yards), which explains why Joe Montemurro signed her with his emphasis on playing out from the back. Her possession data shows some changes in what she is being asked to do with the ball at her feet.
Under Montemurro, Arsenal had a “go backwards to go forwards” style, which was about trying to tempt the opposition out of their defensive block and up the pitch. Zinsberger operated almost like a third centre-half in these phases of play, offering herself as an outlet so the team could build and start again. This still happens under Jonas Eidevall, but not as often as the emphasis is on getting the ball forward more quickly and then counterpressing high up the pitch.
|Season||Passes attempted per 90||Passes completed per 90|
Her completion percentage is stable but we can see she is less involved from a possession standpoint, which is not a problem. However, with a counterpressing style comes a higher defensive line and that means a goalkeeper is called into action to leave their area and sweep up and we can see the high defensive line has had an impact in this area when we look at the defensive actions outside the penalty area.
|Season||Defensive actions outside the penalty area per 90|
Though she is less involved in building play (though still pretty heavily involved in that respect) she is more involved in the defensive aspect with a high line. However, one curious area of the data I have not been able to parse is that she is attempting to catch crosses at a much lower rate compared to previous seasons, as seen below.
|Season||Crosses into Arsenal area per 90||Crossed stopped by goalkeeper per 90||Percentage of crosses stopped per 90|
This could be an improvement area but the data is so stark that I think it can only be instruction from goalkeeping coach Sebastien Barton- because the real downward trend in the data here begins with his appointment. I imagine Zinsberger is being told to stay on her line more for crosses, maybe because the coaching staff trust the likes of Jen Beattie and Simone Boye to head them clear.
However, another key piece of data possibly goes some way to explaining her improvement (which I now consider largely backed by the other supplementary data!) Zinsberger has played 88.9% of the WSL minutes available compared to 59.1% last season. Montemurro had a policy of goalkeeper rotation, a calculation made based on the strength of his goalkeepers versus the game plan for each opponent. Eidevall prefers to have a distinct number 1 and Zinsberger has undoubtedly benefitted from being a more secure first choice between the sticks.