How much is a howler worth?
Three days in a row brought three notable showings of defensive mistakes. On Sunday, Liverpool’s early demolition of Manchester City involved several very key errors from City defenders; Monday saw Scott Dann gift a winning goal to Jermain Defoe; and on Tuesday evening Bayer Leverkusen keeper Bernd Leno let a fairly mediocre shot from a BATE Borisov striker squirm between his legs. With goals such a rarity in football, relative to other sports, mistakes that do lead to opposition goals are very costly.
This is compounded by the fact that, by and large, it seems to be quite difficult to judge who are genuinely good defenders and goalkeepers. Not only have various members in the analytics community tried and struggled, but there are lists of failed transfers which suggests that those employed to do this very job aren’t quite certain (Mignolet and Lovren to Liverpool being just the first examples that spring to mind, followed by Manchester United’s goalkeeping crisis between the years of Schmeichel and Van der Sar).
Heurelho Gomes may well be the face of this conundrum, at least on the goalkeeping side, known for a degree of good goalkeeping but also huge, costly howlers. The errors that lead directly to goals are the most obvious, but to get a more accurate idea of his ability/worth one would have to subtract these from all of the chances that he saved (and then compare this to other goalkeepers to see whether his net cost/benefit is more or less than theirs). The fact that goals and even shots are so rare, relative to other sports with more advanced statistical knowledge such as basketball and ice hockey, means that the samples are small and it’s therefore hard to know whether the performances of keepers are repeatable.
With defenders, the picture gets muddier. The job of a defender (purely in a defensive mode) is to prevent chances, but there are a variety of different ways in which this can be done, and chances that the team concedes while they’re on the pitch are not always their fault. Their mistakes can often be more slight and subtle, and might not even lead to an opening for the opposition, let alone a shot or even a goal.
For example, a centre-back might not be aware of a striker making a good run behind them, and a decent pass to them would leave them through on goal. However, no pass comes, so no chance, shot, or goal is made, but the defender has still made a mistake (of sorts). Perhaps ‘mistake’ is too strong a word. ‘Unwise decision’ is probably better, though a perhaps a little lengthy if talking on Twitter. If a Dejan Lovren or a Phil Jones makes a ‘howler’, just how costly is it? How many more mistakes, or unwise decisions, are they making over the course of a match? How much good work are they doing and how does this balance compare to other players in their position? Why are defenders who make semi-regular howlers playing for the Top 4 and not the Bottom 3?
Fortunately, there has recently been some good work done in this area by Thom Lawrence (http://deepxg.com/2015/11/22/defence-territory-and-control/, @deepxg on Twitter);
Sam Gregory (http://analyticsfc.co.uk/2015/11/24/evaluating-defending-without-actions/, @GregorydSam on Twitter);
and Tom Worville (https://worvilleanalysis.wordpress.com/2015/09/13/visualizing-and-measuring-defensive-contribution/ and http://analyticsfc.co.uk/2015/11/20/expected-goal-relif-adding-context/, @Worville on Twitter) which I would highly recommend taking a look at.
We’re a long way from knowing just how much a howler – and those who make them – are worth, but we’re getting there.
Eastbridge Soccer Betting Broker, winners welcomed!