The nerd's case for exciting football

Welcome to Get Goalside!. It’s good to see you.

Times continue to be hard - consider giving to your local food bank if you are able. If you need to, you can find a foodbank in the UK here.


The people who plug numbers into computers and point to spreadsheets like E.T. points to home are much maligned. Short corners, an overreliance on retaining possession, buying Roberto Firmino: the list of things the nerds can be blamed for is ever-added to.

In particular, there is (or, judging from other sports, soon will be) fears that analytics homogifies styles of play into one ‘Right Way to Play’. There are many features of the game that get given approval or disapproval by numerical experimentation: cut-backs - good; crossing - bad (revisionist takes are available); through-balls - good; long shots - bad.

However, football is a fluid enough sport that I think you could make an argument for playing an exciting style of football which is backed up by ‘analytics’.

First of all, take the Red Bull clubs (other brands of energy drink are available, and encouraged). Although their high-pressing, quick-transition approach to the game didn’t stem directly from number-crunching, it’s one that analytics bods like because it’s so clearly effective.

You don’t need me to tell you that losing the ball close to your own goal is bad, but the statistical work has, indeed, been done on it. If we look at these numbers from American Soccer Analysis, winning the ball just inside the final third could be four times as valuable as winning it just inside your own half. [Full piece here. Team attacking the goal at the top in the left-hand pitch, and defending the goal at the bottom in the right-hand pitch. Methods of similar metrics vary, image used as illustration of the idea]


And as for attacking quickly, as the sugar-water-sellers endorse, well expected goals models have long used ‘speed of attack’ as a proxy for an absence of defenders around a shooter. A shot from 18 yards generally has several opponents in the way, except when it comes at the end of a counter when it might be close to a 1-on-1. Taking advantage of your opponents being out of their shape is undeniably good.

Not only could you find some stats to back up this style of play, but analytics could help keep the fun going for longer. Players only have limited amounts of energy, and knowing the most effective ways to press means that this precious resource can be used when and where it counts.

The first football club to incorporate body pose software — where you give the computer the video and it can tell you what way all the players are facing — could have a major advantage here. Suddenly you can have the direction that every player faces when they receive every pass. Who are the players who receive on the turn, and who are the ones who are closed off?

Not only can that latter group be pressed, but if there are players in the first group who receive on the turn in predictable ways, then they could be vulnerable to having their pockets picked. I pity the central midfielders who are playing the game when this first comes in. There’ll be no escape.

Enjoying this? Share it around:


Now, we need to talk about passing. This is an area where analytics — the use of data — seems to have been mixed with the analytical mindsets of certain schools of coaches.

There’s definitely a case to be made for retaining possession as a method of ‘defending’ (in fact, several moons ago I wrote a piece about ‘sterile’ possession being Pep Guardiola’s version of parking the bus). Having possession of the ball is also, generally speaking, a requirement of creating chances.

However, football’s a sport where possession switches hands (or, feet) around 100 times a match, and only three goals are scored, on average, per match. Is it really so bad to lose the ball? Elsewhere, Kees van Hemmen has argued that the good teams should look to create open games to give them a better probabalistic chance of winning.

A part of that could be through another undeniably fun facet of the game: dribbling. Coincidentally, the rise of what I shall call the ‘Wyscout generation’ of stats-curious online people has also seemed to lead to a rise in interest for dribbling. Dribbling is fun, and checking which players dribble the most is fun.

But dribbling is also a perfectly good way of getting the ball up-field, which is valuable. Not only that, but the type of ball-carrying that central defenders can do when they’re given time and space can disrupt the opposition’s defensive shape.

This latter thing is another of the things that ‘analytics’ (or: the use of data) can quantify, with tracking data and the expertise to use it becoming increasingly available. In 2019, Mladen Sormaz and Dan Nichol gave a presentation at the Opta Pro conference where they looked at the damage to team shape that off-ball runs made; presumably the same thing could be applied to on-ball carries causing an opponent’s midfield structure to collapse.

Feasibly, you could construct an analytics-powered game model of high-press, high-transition, high-risk, high-fun football.

Although, that said…


Subscribe now

Shout-out corner

Forgive me for shouting out my own thing. I’ve launched a brand new newsletter, going out each Friday, where I write specifically about the football on-the-pitch.

Get Goalside! will continue into what it’s kind of morphed into: writing about the data/tech/theory of football and where the future of football might be heading.

Mark’s Notebook will be just football. The first edition was about how Manchester City’s counterpress worked in their 4-2-3-1 against Wolves (this was, you’ll note, two days before it became such a big deal in their defeat to Leicester).

If that sounds like your thing, feel free to subscribe and/or share the news. There’s a link to the newsletter’s site here. There’ll occasionally be short pieces that go up on that site, like this about Wolves being Wolves-ed by West Ham, but the newsletter itself will be one full article every Friday.