Finding Liverpool a Good Keeper

Let’s start with a statement that looks pretty obvious: Liverpool have an awful defense.

While it may look like their bad defense is down to the incompetence of Liverpool’s individual defenders, it’s largely down to the high-pressing system favored by Jurgen Klopp that’s gone awry this season. When the first line of press is broken (which has happened many times this season), the ball gets into dangerous areas because of the vacant spaces left behind. This leads to tricky situations for the Liverpool backline.

The signing of Virgil van Dijk for £75 million won’t change much. The Dutchman is good in the air and a fine passer, but better pressing from the Liverpool midfield and better tactical planning and training can do much more to improve Liverpool’s defensive record.

Now let’s move onto another statement: Simon Mignolet isn’t a great keeper – at all. Although Klopp’s system plays a small role in making him look worse than he actually is – more on that later – the Belgian makes mistakes and is just a bad shot-stopper. According to Raven Beale’s expected saves model, he ranked as the 3rd worst at saving shots on target last season. Liverpool also have Loris Karius, a young German keeper who’s friends with Justin Bieber and looks like a model,


…but he’s only a just-above-average shot-stopper and is mistake-prone. He would serve as a good backup, but he isn’t fit to be the first-choice keeper. However, it’s important to remember that he is 24 years old so has room to grow.

So, here’s an attempt at finding Liverpool a good keeper through data analysis. It’s a quick, three-step attempt. Let’s start with the most obvious attribute:

Step 1: Find me a guy who can stop shots!

The primary function of a keeper is to save shots, unless you’re Claudio Bravo. Remember I said that Klopp’s system has something to do with Liverpool’s keepers looking bad? Here’s why: we tend to rate keepers on the number of saves they make and the number of goals they concede in relation to that. When you concede a low amount of high-quality shots, you don’t have too many shots to save AND the probability of conceding goals from the shots you do face increases because they’re from good locations.

For a Liverpool keeper, the priority is stopping high-quality shots. So, I looked at the shots on target faced by keepers in the danger zone (see pic below, via…



…and looked at the percentage of DZ SOTs saved.  From the data, Mignolet is clearly not the man for the job. His DZ Save% is 34.6%, which is the 3rd-lowest in the Premier League for keepers who’ve played more than 500 minutes.

So, I calculated these numbers for keepers who played more than 500 minutes in Europe’s big-5 leagues this season, and filtered out all players whose DZ Sv% is below average. I ended up with a list of 66 players. The entire list is too large, so here are the top 20:

DZ Sv% top 20

With 66 players in a spreadsheet, I moved onto the next step.

Step 2: How about some keepers who are actually good with their feet?

A keeper who can take part in their team’s build-up (like Ederson of Manchester City) or play incisive long balls (like Everton’s Jordan Pickford) is definitely an asset.

For Liverpool, a keeper who can do both is ideal. Short passes from their keeper can help absorb pressure and help Liverpool control possession. Long passes to Liverpool’s Fab Four of Mohamed Salah, Sadio Mané, Philippe Coutinho, and Roberto Firmino can start counter-attacks or help the team get into dangerous areas.

To take both into account, I created a simple pass rating that rewards pass volume, pass variety and pass accuracy. Scroll to the bottom of this post for the method I used to calculate this. Once again, I filtered out everyone who was below average in the list of 66 keepers (I’d rather have false negatives than false positives), and ended up with 31 players. Here’s the top 20:

Weighted GK Pass Rating.png

And now to the last step.

Step 3: Find me keepers who can claim crosses!

This is the least important step, but a keeper who catches the majority of the crosses he tries to intercept would be great to have, especially if you have mistake-prone center-backs. I looked at the percentage of cross claims and punches that were caught and removed the players who were below the average in this 31-player list. I ended up with the following names (note that this is in descending order of DZ Sv%):

End of step 3.png

In the end

I removed everyone who would be either too expensive or too old (the maximum age I’ve set is 27), and this is the final shortlist:

Final shortlist.png

The next step from a scouting perspective would be rigorous video analysis of all these keepers, which I’m not very good at. From the data and my opinions after watching a bit of him, Roma’s Alisson is the best choice. Only 25 years old and presumably not too expensive, Liverpool should seriously consider buying him and keeping Karius as the second-choice keeper. Many others seem to rate him well too.

If Alisson isn’t available, Lazio’s Thomas Strakosha, Marco Sportiello (on loan at Fiorentina from Atalanta), or Mattia Perin of Genoa are decent alternatives.

So that’s it. While this was a little too simple, I hope I’ve done a half-decent job of statistically scouting goalkeepers. If you have any questions, feedback, or different (and better) ways to scout keepers, message me on Twitter, use the comments section, or send me an email.

Nerd corner

I gave a higher weight to completed long balls and a lower weight to completed short passes. This was because I wanted to reward keepers who could pass long and because short passes are played more than long balls.

As for inaccurate passes, I gave a higher negative weight to inaccurate short passes (because giving the ball away near your own goal almost always results in a goal for the opponent) and gave a lower negative weight to misplaced long balls (as you don’t end up in trouble for losing the ball far away from your goal and turnovers in the opposition half can lead to counter-pressing situations).

Yeah, I know, this isn’t very advanced. If you can think of any better ways to rate keeper passing, direct message me on Twitter or use the comments section below.

This article was written with the aid of StrataData, which is property of Stratagem Technologies. StrataData powers the StrataBet Sports Trading Platform, in addition to StrataBet Premium Recommendations.

Find me on Twitter: @thefutebolist








One thought on “Finding Liverpool a Good Keeper”

  1. Personally not knowing stats on GK’s I found this very useful information and your methodology seems to me to hold a lot of validity. Common sense with some statistics like this simplifies who to watch most. Price is obviously a huge factor as well! Thanks for the article.


Leave a Reply

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

You are commenting using your 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