Some stats-based predictions for the coming season – and why Dortmund may well be a law unto themselves.
We know that both for teams and players, significantly over- or under-achieving in terms of expected goals and expected points tends to be unsustainable in the longer term, despite the considerable turnover of playing and coaching staff that occurs at many clubs each year.
So by looking at last season’s outliers in these metrics, we can make some predictions for the coming campaign.
For example, at the end of the 2017-18 season, the two teams with the biggest over-performances in expected points – in other words, the clubs that had picked up far more points than their displays would have earned them on average – were Bayern and Schalke.
Those sides had over-achieved by 10 and 11 points respectively. As a result, we expected them to struggle to maintain these returns during the 2018-19 campaign, which we all know is precisely how things panned out.
Over-achieving in this manner is often wrongly and unfairly equated with teams and players simply being lucky, which in turns leads to a great many unnecessary arguments on social media. It can be a sign of good fortune, of course, but it might also be due to other factors – unsustainable brilliance, for instance.
Take Hertha’s opening goal last Friday – Dodi Lukebakio’s shot seemed, at best, to be on course to hit the post but was then inadvertently deflected into the net by team-mate Vedad Ibisevic, who was clearly trying to get out of the way of the ball.
REUPLOAD. Bramka Dodiego Lukebakio na 1:0 w meczu 1. kolejki Bundesligi pomiędzy FC Bayernem Monachium a Herthą BSC./Dodi Lukebakio's 1: 0 goal in the match of the 1st Bundesliga match between FC Bayern Munich and Hertha BSC.#FCBBSC pic.twitter.com/zHbvkdC5W5
— Gole z meczów piłkarskich (@Golezmeczow) August 20, 2019
That’s undeniably a moment of good luck, albeit one that brought a great deal of pleasure to many Bundesliga fans.
But now take Max Arnold’s strike for Wolfsburg against Cologne. There’s no disputing that it’s a thing of beauty, and there’s also no doubt that the midfielder intended to put the ball where he did.
At the same time, however, we know that his chance had an expected goals (xG) value of just 0.013. This means that by comparing it to a large number of very similar past opportunities, we can see that on average such chances are converted at a rate of about 13 in every 100.
In addition, we know that Arnold took 43 shots from outside the box last season in the Bundesliga – but scored from just one of them.
So while it would seem inaccurate to label Arnold’s goal lucky, we can still state that such brilliance is unsustainable in the longer term, both for Arnold and any other player.
Another obvious explanation for players and teams over- or under-achieving in these metrics is that they may be doing things well – or badly – that the models just don’t pick up on. And this is worth bearing in mind when considering Dortmund’s numbers in some of the charts below.
We’ll start with what’s probably the most basic of these metrics: expected points.
The first thing to clarify is the aforementioned caveat about Lucien Favre. As Statsbomb’s Ted Knutson put it last season:
“Lucien Favre has a long history of high performance in the table with his teams spoofing how good xG models expect them to be.
“He’s done this consistently enough at Gladbach, then Nice, and now Dortmund that I believe his style of play basically exists in all the holes of naive xG models.”
Like Favre, Julian Nagelsmann also seems able to spoof the models – but unlike the Dortmund head coach, he doesn’t appear to be able to do so in a consistent manner.
Hoffenheim had a very strange last couple of seasons under Nagelsmann in terms of their underlying numbers. In 2017-18 they finished third above two teams – Dortmund and Leverkusen – who respectively had 13 and 12 more expected points than Hoffenheim did.
But last season they ended up ninth despite having the same expected points total as the teams in second and third – Dortmund and Leipzig.
Moving on, there are some obvious takeaways from the above chart. If anyone other than Favre were in charge at Dortmund, then we’d expect the club’s fortunes to nosedive this term. But given that he’s still – thankfully – in situ, that’s far less certain.
It also seems probable that Bremen will have a tougher time of things, although as I like both Werder and Florian Kohfeldt a lot, I’m hoping that the latter has some of Favre’s voodoo in him.
And while you’d think that ushering Bruno Labbadia towards the exit door would give any club a lift, Wolfsburg are another side that look likely to suffer a downturn.
At the other end of the scale, either Nuernberg were bad in ways that this – and every other model I’ve seen – missed, or they suffered some terrible luck.
Similarly, we shouldn’t be at all surprised if both Schalke and Hoffenheim do considerably better this season.
The conclusions are much the same in terms of expected goals, with one notable exception. While Leverkusen ended up on almost exactly the number of points they were expected to, they over-achieved conspicuously in xG, scoring eight more goals than their chances would typically have led to.
As with Bremen, I’m hoping that this prediction doesn’t come true, because Leverkusen were consistently a pleasure to watch after Peter Bosz took over.
Expected goals against
The good news for Leverkusen, statistically at least, is that they also under-achieved markedly in terms of expected goals against (xGA), conceding six more times than they would have on average.
Stuttgart may have been a little unfortunate in this regard, but they were horrible to watch, so it’s unlikely too many people outside of Baden-Wuerttemberg will care.
The stand-outs here are Gladbach and Leipzig, whose defences seem to have been greatly flattered by their final goals-conceded totals last term.
With regards to Gladbach, as perhaps the world’s only Dieter Hecking apologist among the Bundesliga’s many English-speaking fans, I’m not going to add to the chorus of disapproval that seems to follow him wherever he goes. The reasons why Hecking is perhaps not deserving of such relentless criticism are best left to a separate article.
Similarly, while I suspect that Leipzig’s numbers may well be down to the cynical and safety-first brand of football that they’ve often adopted since they arrived in the Bundesliga – something that highlight shows tend to obscure but that the few people who actually watch the full 90 minutes of Leipzig’s games surely can’t help but observe – that too is a tirade for another day.
In the next piece, we’ll look at individual players. And in the meantime, you can read Mathew Burt’s excellent round-up of the latest Bundesliga transfer rumours here.
Latest posts by Warren Pegg (see all)
- Matchday 3 pass-maps for BVB, Bayern, Werder & Gladbach - September 4, 2019
- BuLiStats: MD02 pass-maps and touch-plots - August 28, 2019
- BuLiStats: Predictions for the 2019-20 season (Part 1 – Teams) - August 22, 2019