Rotating Keepers or Stick with One?

I’ve seen so many discussions out there about rotating keepers, and that was my initially my thought, as well, from the beginning. However, I started to dig deeper into the stats and landed on another opinion. Thought I should share it here!

Introduction

So, to better understand the stats, I have excluded new keepers (Allisson, Kepa, Leno) – you can of course choose one of them, but they are still unproven. The keepers I’ve been analyzing are David De Gea (6,0), Ederson (5,5), Lukasz Fabianski (4,5), Matthew Ryan (4,5), Jordan Pickford (5,0), Hugo Lloris (5,5), Jonas Løssl (4,5), Kasper Schmeichel (5,0), Ben Foster (4,5) and Asmir Begovic (4,5).

The numbers

Looking at the obvious stats from the previous season first,  David de Gea (6,0) was a superb keeper with his 172 points and 19 clean sheets, resulting in a 32,2 PPM (point pr million) and 4,7 PPG (point pr game)*.  However, Fabianski and Ryan delivered a better PPM but a lower PPG (Fabianski 34,9 | 4,6 | 157p and Ryan 32,4 | 146p). Ederson was the number four in PPM but not in PPG (30,8 | 4,5 |158). No other keeper was really performing that well, with the worst performers being  Begovic ( 24,9 | 2,9 | 112) and Lloris (27,6 | 4,0 | 144)

* I have adjusted PPM and PPG to actual games played, to avoid skewed stats. Most cases we know when a keeper isn’t starting (maybe with exception of DDG GW38 last year)

That part alone doesn’t indicate much more than you could choose to stick with De Gea or choose two weaker keepers and rotate. Digging deeper to understand the performance, we remove the points for only playing and try to understand what happens with PPM when gamepoints are excluded. Then we end up with a list like this:

DDG | 17,8 – Fabianski |18,0 – Ryan | 15,6 – Ederson | 16,9 – Lloris | 13,8 – Pickford | 13,8 – Løssl | 13,1 – Foster| 11,2 – Schmeichel | 13,2 and Begovic | 8,0.

de Gea, Fabianski and Ederson now indicate even better performance, while it gets harder to separate the other keepers in the 4,5-5,5 range, indicating that it might find better value in the 4,5 keepers than in the 5,0 / 5,5 range.

But let’s dig deeper. All keepers have their monster-games, where they outperform everyone and get the man of the match or save penalties points. That could happen to every keeper, but it’s quite random, and it might not happen, so what I did was that I removed the top three GW’s for all keepers, trying to even out the stats. Then we get the following results.

DDG| 13,6 PPM – Fabianski | 12,9 – Ryan | 10,2 – Ederson |10,2- Lloris |10,0 – Pickford | 8,4 – Løssl | 8,2 – Foster| 6,4 – Schmeichel | 6,2 and Begovic | 3,8.

This, for me, clearly indicates, with exception for de Gea, there is no proof you’ll get more value for money out of a premium keeper compared to a 4,5 keeper. Fabianski and Ryan both create more stable value than keepers like Ederson, Lloris or Pickford. In average the monster GW’s deliver 10p.

For a keeper, stability is key. I don’t want a keeper that delivers up and down, risking having to change him during the season. When looking at games where the keeper got more than two points, de Gea and Fabianski stand out even more, with 22 games. The other viable options landed between 15-19 games with 3+ points.

So – why not stick with two 4,5s and rotate them for gameweeks? Well, back to the numbers above. The monster games are much harder to predict, and without them, you’ll have much worse stats from the cheap keepers. Fabianski, for instance, had his top score against Tottenham away. Løssl scored 10 points against Man U away. Foster maxed out with 10 points playing against Liverpool away. Probably all three would be on your bench during those games.

Conclusion

DDG’s performance last year was extremely good, but he was also the most consistent keeper, and stats combined should have outperformed the others even more. Saving that 0.5 for going with any other premium keepers will statistically cost you between 20-30 points, and saving one million on two 4,5m might just cost you a fortune (or you’ll win the lottery and earn points). DDG is a great choice, especially since it’s the most secure way into the Man United defence. If you’re going for the cheaper keeper, choose one – and stick with him.

Fabianski was the obvious choice last year – will he perform this year as well?

Good luck!