Who's the GK that Manchester United should target?

De Gea this season so far

De Gea is notoriously not a great passer, but Ten Hag adapted things at United to make him fit and do well. But some still question De Gea's position, and sooner or later United should think of a succession plan. On Twitter, you will find some saying this summer is the time, and they often talk about Diogo Costa but does United need to push this button?

De Gea this season so far in the Premier League has a plus/minus for PSxG - Goals Conceded per 90 of -0.15 which is not optimal since you would have someone with a positive metric. Notoriously he’s not someone who is great at attacking crosses and he’s not a sweeper keeper like Neuer, or - to make a name that Ten Hag knows well - like Pasveer. De Gea's save % in Premier League this season is 67.3% not great not so bad. Statistically, he’s not being elite right now.

The Spaniard’s passing style is not excellent tho:

And if we compare it with Pasveer’s from last season we see that even though the average height at which passes end is similar, the distribution is not. The Dutch goalkeeper going less and less long and much nearer the half line, covering the width of the pitch inside his half.

But does United need to invest in this role right now?

Manchester United team is doing fairly well this season, but we should say that Liverpool, Chelsea, and Tottenham forms so far helped. Also, United has other positions to fill this summer, while the owner situation doesn’t give any sort of certainty. Most of all, United need to invest in the most expensive role of them all this summer, the striker.

For all of these reasons, I would say United should either wait another season to go for a really good goalkeeper, get creative with payment and recruitment structures for one or more of the roles they would go for, or find a cheap good goalkeeper.

Easier said than done, I would say. But I’ve tried to scout a good goalkeeper for United and found five profiles that would be good, at different price points and maybe United could end up doing good business in case they would target a goalkeeper.

Starting point

For the database I’ve in place, which includes other than the big five leagues also Eredivisie, Brazilian Serie A, MLS, Mexican League, and Liga Portugal, I’ve filtered out all the GKs that played less than 1000 minutes this season or last. Then I had to find the teams with a similar play style to Manchester United this season and Ajax last season.

Since we’re talking about Manchester United I’ve decided to further filter things down and keep only 17 teams:

Since we want to replace a GK that statistically is not doing very well I’ve further filtered the list with only GKs born after 1993 and that have a PSxG - Goals conceded per 90 that is at least zero, which brought to my attention this wall of numbers:

I’ll save it for you, this shortlist is composed of 23 players that answer all my conditions, but we want to further speed up our scouting process.

Two ways to further filter things

The first shorter list that I made is the one that pops up from keeping only the players that ranked higher than the tenth position in the big list for defensive actions outside the box per 90 possession adjusted, save %, and PSxG - Goals conceded per 90.

I’ve got a list that includes five GKs:

• Mike Maignan 21/22

• Diogo Costa 22/23

• Kepa 22/23

• Brice Samba 22/23

• Gregor Kobel 22/23

The second shorter list would include only players that have a higher value than the median of the big list for:

• PSxG - Goal concede per 90

• Save %

• Defensive actions outside the box per 90 padj

• Short passes %

• Short passes per 90 padj

This second list only includes two players, Kobel and Kepa for the 22/23 season.

Going into details

All the metrics have been processed to make things even, so like I’ve pointed out in the list of metrics right above, some are possession adjusted and on a per 90 bases when needed. Keep that in mind because I won’t say it every single time. Also, I’ll refer to PSxG - Goals conceded as +/- for speed and the ranks refer to the list of 23 which is also highlighted above.

Diogo Costa

Let’s start with the Portuguese, he’s one of the most talked about on Twitter by United’s fans and in the list is the only one that appears two times with Kobel.

In the list, he ranks sixth for Save % and third for +/-, nonetheless should be noted that he’s thirteenth for non-penalty PSxG per Shot on Target. This metric is the average value of probability that a Shot on Target the GK face has to become a goal, the higher the value the higher the probability, so the difficulty for the GK. Since this value is related to players that play against him and defenders’ ability to limit shooters, would be interesting to see how Diogo Costa's performance would change in a different league and team.

He stops about 8% of the crosses he faces, and he’s limited in sweeping ability, which also appears from his seventh rank in defensive actions outside the box, which are made with the fourteenth score for distance from the goal, meaning that he doesn’t go far from it when needed. To end it all he’s ninth for % of short passes.

Looking at his map we can see he focuses his passes on the center backs and passes a lot inside the penalty box, all of this and the fact that he won’t be cheap make me think that United should really weigh their priority and scout him properly before making a move for him.

Brice Samba

Brice Samba he’s doing wonders at Lens. He scores third for save % and third for non-penalty PSxG per Shot on Target faced, which results in him doing worse than Diogo Costa, but he’s sixth. He’s only tenth for defensive actions outside the box, but he goes far from the goal with the fifth rank for distance. He engages 7.3% of the crosses he faces and ranks seventh for short passes %.

His pass map is similar to Pasveer’s the thing is Samba is currently playing in Ligue 1 after helping Nottingham Forest secure a Premier League spot.

Again he’s doing wonders but United should ask if he’s one season wonder or not, even though Ligue 1 is a good benchmark.

Mike Maignan

Mike Maignan doesn’t need any sort of presentation. In 21/22 he ranks first for the average difficulty of saves but still first for save %, obviously this average difficulty appears in his +/- rank, eight. Ninth for defensive actions outside the box and sixth for distance which is fairly far from the goal.

The rank for short passes % is fourteenth which means that he goes long fairly often for a player the UMAP algorithm classifies as short playing GK.

The pass map shows he likes to find players outside the box for all the width, he doesn’t go till the halfway line like Pasveer but we can see he finds some pockets of space even further.

Maignan has two problems. This season his data are probably worse because of Milan's bad form, while still being pretty expensive since he’s France’s and Milan’s Starting GK, he won already trophies but most of all both last and this season he suffered some important injuries that kept him out for some month.

He would probably be the safest bet in my opinion if physical concerns are cleared but they have to be, and money needs to be right.

Gregor Kobel

The German GK is the other one that appears for both last two seasons and he’s UCL experienced. He ranks tenth for the average difficulty of save while doing seventh for save % and fifth for +/- which means he’s a good GK. He ranks fifth for actions outside the box and second for distance from the goal, only one player goes further from the goal than him in the list.

He ranks eight for short passes % and his pass heat map feels good, with a good spread and distance from the box:

Kobel, he’s the only one with Kepa surviving the filtering for median values, so he’s doing well this season. Like probably everyone else on the list he would be expensive and Borussia Dortmund is not someone that needs to sell or do it without getting what they want.

So now let’s say it - the hot take -, Manchester United should go for Kepa. Yes, Kepa.

Kepa Arrizabalaga

Okay, okay, let me explain. Let’s start with the bad things about this one. Kepa started doing well only this season after years on the bench, and his pass heat map is not the best one if United wants something different from De Gea in possession. He also engages only 4.3% of crosses.

He likes to go short with passes, indeed he ranks fifth in the list of 23 for short passes %. But. He is first for defensive actions outside the box and third for average distance, he goes out and he’s not scared too far. One problem would be that the average difficulty of shots he faces ranks him eighth, he has an easier life than Samba and Maignan, but not than Kobel and Costa.

Keep that in mind because he ranks fourth for save % and first for +/-, he’s doing well. So, why Chelsea should sell him? And why United? Well, Chelsea seems in the market for a GK as far as reports from the Athletic, and Kepa never played until Potter came.

Last summer he was in talks for a loan to Napoli. I see the chance for United to get him on a cheap transfer, maybe even on a loan with the chance to make it permanent.

Conclusion

If United decide that right now they have to plan for the future of their GK position I would think of Maignan, Kepa, or Kobel. The wise thing to do in my opinion would be to get Kepa and have him compete with De Gea next season, while in this Summer going for other positions, because I don’t think United should be aggressive for GK for now, unless they could get a good one for cheap more or less.

If he doesn’t work, United hopefully can end a loan and spend on another GK. But all of this would mean Kepa and Chelsea agree on the conditions.

United has some work to do on the GK position.