Jan Oblak vs. David de Gea vs. Manuel Neuer vs. Alisson – Which Goalkeeper Takes the (Analytics) Crown?

The Highs and Lows of Being a Goalkeeper

The goalkeeper’s fraternity lives by its own set of rules. This is the one position in football that has always been treated as a separate entity – shunned when a single mistake is made and praised to the heavens when a penalty is saved. Reminds me of the binary system – there is no middle ground. Either you’re the hero or the villain.

The scrutiny on a goalkeeper is unending and merciless. I remember growing up watching Oliver Kahn in German colours and thinking “this guy is unflappable”. Best in the world bar none, surely? And then the 2002 World Cup final happened when he let a simple shot slip and with that went Germany’s hopes of winning football’s greatest prize.

Kahn was a colossal keeper, make no mistake. But he was neither a goalkeeping god nor was he an overhyped has-been. He was, as a lot of my fellow analytics colleagues would say, a victim of the phrase “regression to the mean”.

I was recently watching the 2013 Champions League quarter-final between Bayern Munich and Manchester United (famous for Evra’s stunner and United’s quick collapse), and it got me thinking – who’s the best goalkeeper in the world right now? Is Manuel Neuer still as good as his fans claim? Are David de Gea’s standards slipping? Has Atletico Madrid’s Jan Oblak stolen the crown and made it his own? Is Alisson’s record-breaking streak enough to make him the best?

best_goalkeeper_analytics

I wanted to answer these questions through the lens of hardcore numbers. We can debate about each goalkeeper’s technique and try to find a winner objectively, sure. But I’m more of a numbers guy so I decided to take a swim in the analytics pool and figure out who reigns supreme in the upper echelons of the goalkeeping fraternity.

Note: I will use a few terms here you might not be familiar with if you’re new to football analytics. I recommend these articles to get started:

 

An Analytics Framework for Judging the Goalkeepers

So, what are the metrics we can look at to judge these four elite goalkeepers? The most obvious one that comes to mind is the number of clean sheets, right? But that’s a way too simplistic metric and it’s too random. A clean sheet is the outcome of a collective effort put in by the entire team, led by the team’s defence and the goalkeeper. A goalkeeper with a poor defence in front of him will always concede more. Does that make him a lesser keeper? Not really.

So, let’s see what we can work with:

  • Save %: Fairly straightforward but gives us a good estimate of a goalkeeper’s shot stopping prowess
  • Post-Shot Expected Goals – Goals Allowed (PsXG +/-): Sounds like a mouthful but this is a super helpful metric to gauge a keeper’s shot stopping ability. This basically tells us the expected goals based on the fact that the shot is on target, subtracted by the number of goals the keeper conceded. Here’s what you need to know – Positive numbers suggest an above average ability to stop shots
  • % Crosses Stopped: What’s the percentage of crosses the goalkeeper stopped? This helps us understand the aerial ability of the goalkeeper

Check out this article by Statsbomb that breaks down PSxG in detail. Suffice to say, it’s the best we have right now and is good enough to help us settle the debate!

A few things to note:

  • I have taken data starting from the 2018-19 season for each player’s domestic league. This is because that is what’s available out there. More advanced stats and numbers are kept hidden from the general public due to proprietary reasons. Also – 2 seasons is a good enough sample size to answer our questions
  • I have NOT included any penalty numbers here. I believe that’s a separate entity altogether and having a higher percentage of penalties saves doesn’t necessarily make you a better goalkeeper (or add anything to our debate here). I’m open to debate this anytime!
  • Yes, there are other things we can analyse, such as passes completed, defensive actions like sweeping, etc. But I feel those are largely dependent on the system the team is playing. It’s a bit unfair to judge the goalkeeper based on that. Again – open to hearing reasons!

Alright, with that out of the way, let’s dive in!

 

Save% – Which Goalkeeper Wins the Duel?

What does the eye test tell us? Which goalkeeper do you feel saves the most shots he faces? Let’s find out:

Save_percentage_best_goalkeeper_analytics

  • Alisson is so far ahead of the pack it’s not even close! With a save percentage of over 80% (that’s insane, by the way), he is the king among a pack of jokers
  • Interestingly, Manuel Neuer (in green) has significantly improved this season. He did suffer from a drop in form due to injuries last term but has seemingly gotten back to his old best
  • Jan Oblak (in red) has suffered the biggest dip in save percentage (82.1% to 73.3%). The eye test tells us he’s still an elite shot stopper so we’ll analyse his advanced numbers in a bit and see if they can tell us anything
  • David de Gea continues to disappoint. He was a bit of a downer last season (only 71.6% shots saved) and has declined a bit more this term (71.3%). No longer the best in this category

Winner: Alisson (Liverpool)

 

Post-Shot Expected Goals – Goals Allowed (PsXG +/-): What do the Advanced Stats Tell us?

This is where we separate the wheat from the chaff. The valuable from the worthless. What does the most advanced goalkeeping metric tell us about these four shot stoppers?

post_shot_expected_goals_best_goalkeeper_analytics

  • Wow, look at that precipitous drop for Jan Oblak. He went from +12.3 last season to -2.3 this season. The save percentage we saw earlier aligns with this alarming decline in Oblak’s underlying numbers. What this means is that Oblak was saving far more goals than he was expected to last season – and he is now conceding more goals than he is expected to. Huge, huge drop there
  • Alisson continues to impress massively in the advanced numbers as well. He was already a huge positive last season (+4) and has improved this season (+4.6)! There’s no stopping this Brazilian tornado
  • Neuer has clearly risen from last season (as we saw above). He is still in the negative (-0.2) but getting better. Here’s what this tells me – Neuer isn’t the best at shot stopping. These aren’t promising numbers for a goalkeeper who should be on top of his position
  • De Gea’s fall from grace continues this season. He’s dropped from a fairly decent +2.1 PSxG +/- to -0.1. He should watch out for the competition Dean Henderson will bring to the table next season for Manchester United

Winner: Alisson (Liverpool)

 

%Crosses Stopped – Which Goalkeeper Commands His Penalty Area Best?

Which goalkeeper commands his penalty area best when it comes to stifling crosses and the wing play of the opposition? This is an under-looked area when we’re judging goalkeepers so I wanted to analyse this as well. Yes, this isn’t exactly an advanced metric but it still gives us a bird’s eye view of how comfortable a goalkeeper is when up against the physicality of the opposition.

cross_stop_percentage_best_goalkeeper_analytics_1

  • It’s a clean sweep for Liverpool’s Alisson. He is the best cross-stopper in the world right now and this is another metric where he has improved from last season (+0.2% improvement)
  • Oh dear Jan Oblak. Another nose dive for the Slovenian goalkeeper. His cross stopping percentage has more than halved from 10.6% last season to 5.2% this term. What is going on with Oblak?
  • De Gea keeps up his consistently poor showing by dropping his numbers in this metric as well. He’s never been a great cross stopper before but his numbers are slightly alarming now. He’s by far the worst among the lot
  • Neuer is a close second to Alisson despite a very minimal drop in numbers. I wouldn’t be too worried if I was a Bayern Munich fan

Winner: Alisson (Liverpool)

 

Final Thoughts

Well, all the numbers, traditional as well as advanced, point towards Alisson taking the best goalkeeper crown. He has been immense behind an already impenetrable Liverpool defence and continues to inspire fear in opposition strikers. His all-round goalkeeping game doesn’t leave a lot to question and given that he’s only 27, he’s just starting to hit his prime.

4 thoughts on “Jan Oblak vs. David de Gea vs. Manuel Neuer vs. Alisson – Which Goalkeeper Takes the (Analytics) Crown?

  1. Akshar says:

    Hi Pranav!

    Your study seems interesting while you compare a few metrics to evaluate the best in the world considering the last two seasons.

    After reading the same, I thought of performing similar stats for Goalkeepers when they were in their best form, a cluster which talks about consistency and their milestones. I guess this would definitely explain who is best in the world when you compare the data when they were in the top form.

    1. pdar22 says:

      Hi Aksar,

      Thanks for reaching the article! How would you go about quantifying “best form”? Once you define that – you can proceed with comparing them. The problem I feel you’ll face here is that the advanced analytics numbers are not available for seasons before 2016/17 (not publicly anyway). That is a significant hurdle, unfortunately.

  2. Akshar says:

    One more thing, would you recommend the sources from which I could collect the data for this analysis?

    Also, if any suggestions from your end, do comment.

    1. pdar22 says:

      Honestly, there aren’t a lot of sources that give out this data to the public. I refer to 2 sites that I trust with the numbers: understat.com and fbref.com. The latter has recently added a lot of advanced stats so that’s worth checking out.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s