⚽Expected Goals and Net Gains

‘Expected Goals: The Story of How Data Conquered Football and Changed the Game Forever’ by Rory Smith (2022)

‘Net Gains: Inside the Beautiful Game’s Analytics Revolution’ by Ryan O’Hanlon (2022)

Why hasn’t someone written ‘Moneyball’ for football? Given the greater availability of data and the rise in the use of analytics in the sport, it felt like only a matter of time before a Billy Beane like figure emerged to educate the masses in how the game is really played and what performance attributes are really valuable.

Football, however, seems to be much too secretive for any club to give away their secrets or to make the specifics of what they do quite as public as Beane was willing to do in Moneyball. But there are other reasons too for why the data revolution in the sport has been much slower – the greater complexity of football relative to baseball, the lack of data available to amateurs and the general public, and the low scoring nature of the game.

So in the absence of a Billy Beane (even though Beane himself is now heavily involved in football), football needed its own Michael Lewis to poke around and explain the rise in the use of data and analytics in football to the masses. Thankfully, it found two such authors resulting in these two excellent books published last year.

Both Expected Goals and Net Gains tell a broad story of a ground up evolution where smart people either with an interest in both data and football, or simply with an eye to making money, began to explore this space and make strides in both the collection and analyses of data on football matches. The influence of Moneyball on most people profiled in both books makes it clear that impact of Lewis’ book has spread far beyond baseball.

Expected Goals by Rory Smith has garnered significant praise in the UK and has been shortlisted for a bunch of football/sports book awards. Smith’s book is about people not data or formulae – he focuses on the analysts, academics and executives who have sought to expand the use of analytics in football rather than what the data says about the game. It’s a very England focused book largely looking at people who worked within English clubs but touching on the wider world occasionally. Smith tells the story of how data began to be collected in greater volume by Opta and Prozone and of some of the early pioneers who started to try and turn the data into useful insights.

A significant portion of Expected Goals focuses on Chris Anderson, a professor and the author of ‘The Numbers Game: Why Everything You Know About Soccer Is Wrong’ and his attempts, ultimately not quite successful, to find a backer to buy a football club and turn it over to him and a data led approach. The strength of the book is undoubtedly Smith’s exceptional ability to tell a story (for evidence just read this wonderful piece in the NYT) and keep a reader engaged. However, the Anderson subplot is given too much space. It’s an interesting story and, while relevant to the topic (especially on the sport’s resistance to change), it’s ultimately much more about the business of football than the revolutionary use of data.✓

Net Gains by Ryan O’Hanlon is by contrast a very American book. Not just written by an American and with an American audience in mind (hence the use of Americanisms like roster, cleats, overtime etc.), it also focuses much more on America/Americans – both through telling the story of some Americans involved in football analytics and in drawing parallels between football and other US sports. O’Hanlon applies a wider lens than Smith by also looking in more detail at broader off-pitch issues (like the research that shows higher wages = trophies) rather than a narrower focus on the collection and use of on-pitch data.

O’Hanlon also discusses in more detail what football data actually says and some of the specific uses of data – to improve set-pieces, to evaluate players who aren’t midfielders, to track physical performance and effort etc. O’Hanlon is also coming at the story partly as a former player and partly as a journalist – he is genuinely curious what the numbers say as well as in building the broader narrative.

My only criticism of Net Gains is the poor choice of narrator for the audiobook version – someone who doesn’t know how to pronounce familiar football names like Arsenal, UEFA or Jose Mourinho probably shouldn’t be reading a book full of familiar football names. I quit the audiobook and switched to hardcopy pretty quickly.

The evolution of one particular stat (expected goals or xG) is central to both books. It is undoubtedly the public face of the data revolution in football and it’s various origin stories and evolutions are covered well – Smith profiles the guy who designed the version seen on TV while O’Hanlon profiles the guy who coined the acronym xG.

For both books, the challenge remains the absence of a Billy Beane figure willing to spill the details. Hence both struggle with the question of which clubs can we identify as successfully using data and analytics to outperform expectations. O’Hanlon goes into more detail than Smith on the fascinating experiment at FC Midtjylland and Brentford with a particular focus on the little things they worked on like set pieces and use of data in recruitment. Other clubs seeking to use analytics more centrally are identified and discussed, like Fulham, but ultimately both books point to Liverpool as probably the best example of a top level team incorporating analytics into winning a big European league.

In reading both books, and despite the overly bold subtitle to Smith’s book, it becomes clear that the use of data and analytics in football has a long way to go. While its use in recruitment appears to now be standard practice, there remains plenty of scope for revolution in the use of data to amend how a team trains and plays. Where the other limits of this approach will ultimately lie remains to be seen as the complexity of football comes up against ever increasing brain and computer power.

Expected Goals and Net Gains complement each other very well. While there is inevitably some crossover, it is more limited than I would expect and only one particular analyst is heavily featured in both. In other areas where one book just touches on a topic the other dives deeper in a satisfying way. Highly recommend them both.