It’s holiday time. And manager sacking time. The period right before the January transfer window is when clubs in the bottom-half sack their managers. Although not too many clubs have fired their managers this season yet, Alan Pardew was given the pink slip on the 22nd of December by Crystal Palace.
As Soccernomics (Stefan Szymanski and Simon Kuper) and The Numbers Game (Chris Anderson and David Sally) have shown us, it really doesn’t make much sense to fire your manager. Bad results usually just regress to the mean, whether you sack your manager or not.
This, however, is soccer. Sense doesn’t matter. Anyway, can we use stats to say if Palace should have given Pardew the sack? I think I can do well enough…
Many managers, regardless of their skill, are fired because their style doesn’t go well with their ambitious bosses. Some like a possession-based, rapid-passing style, others like a high-pressing, dynamic team, and a few like fast-paced, counter-attacking football.
And what tactics did Pardew employ at Palace? Let’s first see whether Pardew was a Guardiola or a Mourinho. This scatter (created by Ed Shorthouse, like all the others in this article) shows us how many passes Premier League teams played and how much of the ball they kept.
Crystal Palace kept 49.1% of the ball, which is just below average. They played around 390 passes per 90 minutes. Their style wasn’t too counter-attacking, but they kept slightly less possession and played a few passes less than the average Premier League side.
Now let’s see how high and intense his Palace pressed. We can use a few metrics for this. First, I’m using pass prevention, which can be calculated as (100-Pass Completion% Against). Palace prevented 22.8% of opposition passes, which is the 4th highest score in the league. Only Manchester City, Tottenham and Liverpool have prevented more passes.
In Palace’s own defensive zone, they let 60.4 passes per 90 minutes. This is the 3rd highest in the league. Even Tottenham and Liverpool have conceded more passes. Based on pass prevention-based metrics, Palace are intense at pressing.
To sum up, Palace is a high-pressing team that doesn’t need the ball very much, like Diego Simeone’s Atlético Madrid. Did Crystal Palace’s board like that? Simeone has been criticized for his pragmatic tactics. Maybe Palace’s American owners wanted a different style of play.
Though style is a common reason for sacking a manager, the biggest reason is because their team was underperforming. Although the best performance indicator is Expected Goals, I don’t have xG numbers. However, I can use old stats.
The simplest, yet effective repeatable metric is Total Shots Ratio, or TSR. Crystal Palace’s TSR so far this season has been under par(dew). Their TSR is 47.6%, which means they have conceded more shots than they have produced. This scatter shows us how many shots Premier League clubs have produced and conceded.
As you can see here, Crystal Palace are just below average. Though they don’t belong in a Europa League spot, they shouldn’t be fighting relegation, either. Their shot ratio indicates they should be in a better position than they are in now.
Another metric, built on the same idea, is Total Shots on Target Ratio, or TSoTR. Crystal Palace are below average here, too, with a TSoTR of 45.5%. This scatter shows how many shots on target PL clubs have produced and conceded.
It’s the same here. Crystal Palace should be mid-table based on shot ratios. In fact, they should be in the top half of the table based on TSoTR. Palace certainly have no problem with shots.
It’s also important to create clear-cut chances. What Opta logs as a ‘big chance’ is a good indicator of that. While Palace has created 26 big chances, which is very good, they have conceded 29. Only Bournemouth, West Ham and Swansea have conceded more big chances. The following scatter plot displays big chances for and against in the Premier League.
Alan Pardew’s Palace have created lots of big chances, more than Everton and Tottenham. Their attacking play certainly isn’t a problem. However, they have real holes in defence. They have conceded far more big chances than average.
But what we need to know the most is if Palace was underperforming this season, and if their results were bound to regress to the mean. The best metric for measuring this is the controversial PDO. It is a simple sum of shots on target conversion and save percentage. PDO regresses to the mean of 100. If your PDO is significantly below 100, it means your results will improve soon. If it’s significantly above 100, it means your luck will run out soon. Most managers are sacked when their PDO numbers are below 100.
But Palace’s PDO is 99.5. That is very close to 100. Was Pardew unlucky? Palace’s PDO is only just lower than 100. Their bad results weren’t really down to bad luck. PDO may be an old and highly controversial and ambiguous metric, but it tells a story.
Ask anyone and they’ll tell you Crystal Palace are a horrible team. They kept losing. They lost to big sides, and got into goal festivals that didn’t end well for them with mediocre sides. The fact that results weren’t going to get a lot better meant there was only one possible solution – fire Pardew.
But were they actually doing that bad statistically? Their performances based on very reliable metrics are decent. Though Palace is 17th right now, they should really be 11th or 12th. So it is bad luck. If Palace stuck with Pardew and tried repairing their defense, Palace would have stayed afloat. And Pardew did quite well with the team last year, when they reached the final of the FA Cup.
While I don’t like the fact they fired Pardew, I do like their new gaffer – the analytics-reliant ‘Big Sam’ Allardyce who was recently fired as the manager of the English National Team after just 67 days in charge. Whatever Pardew and Palace do now, I hope they do well.
Anyway, this is ironic: https://www.youtube.com/watch?v=ten9Mp2wjhA
I cannot express how much help @EdShorthouse has provided me for this article. He stayed up till 4 am to create these (excellent) scatter plots. I also want to thank Benjamin Pugsley for his immensely useful website objective-football.blogspot.com. Every number on this article is from his site.