It has been quite the transformation for Bain from unwanted at now relegated Dundee to number one for the eight-time champions and on the verge of another Treble. The biggest compliment is that there isn’t even a debate anymore concerning the goalkeeping position. Bain in on the majority of fans’ team sheets.
Does the data support this?
Save Data
Clean sheet statistics are always welcome for ‘keepers but are not the most accurate reflection of individual performance, reflecting the collective. Similarlly, the oft quoted Save % and Number of Saves is more indicative of the relative strength of a team. Indeed, ‘keepers from teams lower in the league will generally perform better by those metrics.
Of course, this data has relevance in comparing different players from the same team as the level of opposition will be similar.
For the record Gordon has conceded 27 in 32 matches and Bain 11 in 30. Gordon played in all the European matches other than the Valencia ties, however.
Bain also makes more saves – 2 per 90m compared to 1.88 for Gordon. What is surprising is the gap in Save % – the % of all shots on target saved:
This does not tell you about the quality of the shots saved however. For that I measure the xG total of the shots saved per 90m:
Bain has a slight advantage in that his saves are expected to result in 0.05 of a goal per 90m than those saved by Gordon.
As mentioned above, Gordon has conceded 0.84 goals per game compared to only 0.37 per game with Bain in net. If we again consider the xG quality of all the shots on target in comparison:
What this shows is that Bain has conceded 0.37 goals per 90m but based on the xG quality of the shots on target, he would have been expected to concede 0.616.
Conversely, Gordon has conceded 0.84 goals per 90m but was only expected to concede 0.741.
Clear advantage Bain!
Distribution
Passing out from the back was always a feature Gordon tried hard to master but it never came naturally. The smaller, more agile Bain is comfortable with the modern requirement to pass from the back.
Bain completes 20.93 passes from open play per 90m compared to 18.47 but if we look at pass completion % there is quite a gulf:
8% is a hefty difference for passing data for players in the same position.
Bain also plays more penetrative forward passes. He completes 2.2 Pack Passes per 90m compared to 1.63 by Gordon.
Errors
Goalkeeping errors tend to be costly, as Bain found out against Heart of Midlothian last time out. But again, over the season, Gordon is averaging 0.25 Defensive Errors per 90m compared to Bain’s 0.16.
Conclusion
Across all the key goalkeeping metrics Bain demonstrates an irrefutable case for being Number One.
Not many would have predicted that at the start of this very odd season!
James E McLeod says
Love all your stuff, though I was a bit surprised (pleasantly, ‘cos I like Bain) about the difference in shot stopping, because I’d always thought that that was Gordon’s strength. But happy to learn otherwise ?
However got to say “deerie deerie me” about a couple of the charts… XG Saved and Pass Completion x-axes should start at 0. As they are, they totally exaggerate the gap… poor show.
But good work nevertheless.
Cheers
Jim
Martin says
What makes you think the axes should start at 0? I disagree. I think you’d need to know the league average and look at deviation from that, but I imagine the author has neither the time nor the inclination to calculate this!
celticbynumbers@btinternet.com says
Also it is primarily about communication and getting the point across to people many of whom won’t pore over every value.
James E McLeod says
But that is precisely the reason the axes should start at zero – because otherwise it gives a misleading impression of the size of the gap. The whole point of using a graph is to illustrate visually how two or more things compare; by omitting most of the bars it looks like Bain is twice as good as Gordon.
Martin says
I think this makes interesting reading for a couple of reasons: 1) I really felt that Gordon took Rodger’s desire for a ball-playing keeper to heart, and improved his distribution immeasurably over the last couple of years. Despite this, Bain is still demonstrably better. 2) I’m probably not alone at being totally shocked at how completely Bain has assumed the No1 spot. I keep expecting his bubble to burst, and like James above, always considered Gordon to be the better shot-stopper. To see evidence that Bain is not only better at distribution, but seems to be better at shot-stopping and almost every other measurable outcome, is a reminder of the strength of objective analysis!