This piece comes in two forms: the written piece here and the video below. If you feel less initiated into analytics, want me to expand on some points or want to watch me in my natural habitat, check it out:
Louis Van Gaal wasn’t always this conservative. Back in 1997-98 his Barcelona team won La Liga despite conceding 56, the 4th highest number of goals that season. In his one full season at Bayern Munich in 2009-10 they won the title with the most shots on target. The following year he coached them for the majority of a season in which they scored 81, the most any Bundesliga side had managed in 23 years and his two forays into the Dutch league at Ajax and AZ produced plenty of titles and goals.
His Manchester United side seem a little different. Shot monster Memphis is playing in a straitjacket, what remains of Wayne Rooney has scored only twice and renowned psychologist Paul Scholes noted that Martial “doesn’t look bothered”. O, expansive-football-playing-Louis, where art thou?
Let’s start with the good news – Van Gaal is sucking any and all signs of attacking life out of opposition teams. 8.9 shots against p90 is bested only by City this season, whilst 3.4 shots on target against p90 is just 0.4 behind Liverpool.
Any reasons to think they could significantly cool off as the team to concede the least in the league are speculative – sure, 78.7 is a very high save percentage but City, Chelsea and United have all roughly managed it at times during the enlightened era. A more interesting point is just how much time United have spent winning this season and how little they concede once they get there.
Despite seemingly never shooting, United are second to City by just a minute in the amount of time they’ve spent winning this season. In particular they are spending a huge amount of time in the +1 game state, where save percentages tend to be a bit higher on average. And what’s their save percentage in that state? 86.7%, vs 68% rest-of-the-league average. When United go 1 up, they have been incredibly hard to score against. Who knows if that will sustain – perhaps they’re a ridiculously good defensive side or perhaps they will succumb to some fundamental mean. Either way, they’re currently conceding a ridiculously low number of goals when they go ahead.
It’s not just shots and goals that Manchester United are avoiding conceding. They allow opponents the most passes in the opponent’s defensive third, but concede the least in both their own defensive third and the middle third.
While they are preventing the opposition from having any significant quantity of the ball at their end of the pitch, the possession they enjoy themselves is focused and prescriptive. That endless wide player > Carrick > other wide player passing pattern, as recently pointed out by your correspondent’s new mate Alan Shearer, makes their pass volume seem favourably high but, as Colin Trainor highlighted in this tweet, Manchester United’s game is characterised by perpetual rigidity:
Heatmap showing the location of MUFC's passes received this season. Did I hear someone say LvG likes a rigid system? pic.twitter.com/7VanuITUW9
— Colin Trainor (@colintrainor) November 27, 2015
Importantly, goals are a function of shots, not (directly) of passes. United are taking 10.3 shots per game (4th worst in the league), with just 3.8 being on target (=7th worst in the league) and while the endless passing is getting them into advanced positions, what they do with the ball when they get there has been less impressive. They’re still just 10th for goals scored in both Paul Riley and Michael Caley’s expected goals models. Their attack looks almost insultingly bad for a team challenging for top four:
It’s not just United’s attack which lacks historical precedent. No team in the Premier League’s enlightened era has had a shots for/against profile even remotely similar to theirs. Can they continue to score from a few shots and then ride their save percentage to narrow wins? Stats me is tempted to say no, but the truth is that we simply have scarcely seen such low-event football before; United at 19.2 combined shots per game are 1.6 lower than any team in the Europe’s top five leagues.
Joke all you like about Wayne Rooney, but it is the face of this entire Manchester United side that looks like a monster. LVG could well be running a freaky experiment based on an iffy hypothesis. I haven’t got the authority to say it’s necessarily bad, but I cannot wait for the results. Will United finally start shooting more? Can this new experiment maintain 37% conversion rates and top-end save percentages? Can van Gaal’s madness overcome our methods? Have we become too blanketed, cappuccino in hand, to realise “what a load of nonsense” analytics really is?
Given the obvious and varied depth of talent available I’m not necessarily sure there is an attacking problem at United but, assuming there is, where is it stemming from: the system or its parts? The parts, bar Juan Mata, have all had their shots restricted this season. I think Van Gaal’s tactical conservatism is killing their numbers. When a system is so evidently limiting, it does make you wonder what the point of paying top-dollar for star individuals was…
Rooney’s pattern jumps out and while he has played in varying positions, his goal, assist, shot, shot assist and key pass numbers have all been tanking since 2012, whilst the percentage of United’s league minutes he’s played has gone up every year over the same period. Already thirty, the ridiculous contract he’s on runs out in 2019. As each year passes and his contributions wane, it reminds us that this is a very clear example of how analytics can improve decision-making at the elite level. Purely from a common sense business perspective, something has to budge if United don’t want to pay Wayne’s corpse £300k p/w to jog about for the next 4 years. Taking a leaf from the Adebayor chapter of City’s book doesn’t look like the worst idea right now.
I find the interactions between ‘LVG-factor’, ‘upwards-sloping-age-curve’ and ‘better-league-factor’ intriguing in Depay and Martial’s shot numbers, but that’s for another day. What the above plots make clear to me is that there is something systematic moving those numbers down. It’s not simply position changes and it’s not just individuals dropping due to age or league effects.
Even under the pretty safe assumption that United get top 4 this season, is that enough? Will their international stars be happy giving up some of their best years to grind out 3rd? How long can the continental charm last if United don’t deliver a serious title challenge, something that surely can’t be possible from such a low base of shots?
In 2009 Louis Van Gaal reportedly lost up to a million pounds in the Bernie Madoff fraud, possibly forcing his big-money move to Bayern Munich a few months later. Hopefully his football can be lifted out of poverty with similar promptitude.