Once again, Footiebusiness Contributor David Laidig back with a great look at an item of interest to soccer fans. Thanks to Dave for another great piece.
The use of analytics is rising in sports, and is a quiet revolution in soccer as well. Using analytics – or
quantitative based analysis – in soccer is nothing new. In fact, in the late 70’s, the manager for Dynamo
Kyiv would post targets of specific game behaviors (such as tackles, interceptions, headers on goal,
and types of passes) for his players, with the targets reflecting the style of play and tactics adopted for
that particular game.1 However, for the quantitative analyst – or Quant – obtaining this data remained
challenging and labor intensive. But in spite of this difficulty, efforts continue with the goal of applying
empirical analysis to tactics, strategies, revenue streams, and maximizing value under a salary cap
or other cost restrictions. For the fan, the last example is the first that comes to mind. Indeed, the
term Moneyball still remains a shortened definition for unearthing market inefficiencies in sports. An
attractive concept – even without Brad Pitt – for MLS teams with less purchasing power than teams in
established European leagues. But the Quants are gaining visibility and influence within the world of
soccer. And behind this trend are new technologies yielding new data and a growing class of soccer
analysts that bring their own perspective to the beautiful game.
Evidence of the Soccer Quants is undeniable. For example, MIT recently hosted the Sloan Sports
Analytics Conference where, for the second year in a row, soccer was the focus of its own session.2
There, interestingly, the biggest data-driven teams were listed as Chelsea, Everton, Fulham, and
Manchester City in England and Dortmund and Hamburg in the Bundesliga.3 And considering that
soccer is a sport that has embraced new media, it is heartening that several outlets provide a forum
for discussions of soccer analytics.4 Even twitter heralds the Quant trend as Gavin Fleig, Head of
Performance Analysis for Manchester City, recently offered to answer public questions via twitter for an
hour while on the train to a Man City fixture. And yes, I kicked myself for watching my kids instead of
my twitter feed.
Some have incorporated this trend towards analysis in innovative ways. For example, the EPL Index
allows fans, for a small subscription fee, to access a wide variety of game statistics.5 And not only is
access available, but fans are encouraged to post the results of their analyses on the site. The result
is a mix of organic social media with quantitative analysis. This not only strengthens public ties to the
league, but also advances the knowledge of the game. In the United States, such access to data is not
widely available. MLS does offer insightful chalkboards for its games, thanks to its partnership with Opta, but the aggregation of such data is time intensive and the information provided is more limited.
However, in spite of difficulties, the appetite for quantitative analysis continues unabated. A recent
example illustrates how grass roots pursuit of quantitative analysis sheds light on fundamental business
MLS has recently published a Performance Index, sponsored by the league sponsor Castrol, summarizing
the 2011 season. This Castrol Index assigns a number for each player; a number intended to represent
that individual’s contribution to the team’s wins. Benjamin Leinwand and Chris Anderson used the
Castrol Index and player salaries in their analysis suggesting that the MLS market undervalues defensive
play, and overvalues forwards.6 Using Leinwand and Anderson’s result, teams would be justified in
second-guessing a high-priced Forward contract, and may decide to emphasize Defenders.
However, the Castrol Index is highly correlated with minutes played, especially for the group of players
with less than 1800 minutes (20 full games). The number of minutes played contributes nearly 90% of
the variability to the Castrol Index for players with less than 1800 minutes. For the players with 1800 or
more minutes in 2011, their time was not related to the Castrol Index score – suggesting that the Index
is measuring something else besides playing time for this group.
Thus, by looking at the Castrol Index from another perspective (i.e., its relationship to minutes played),
one can better utilize the information by accounting for its limitations. For example, one might target
players that over perform based on minutes played (but still fall behind the league leaders that get
regular playing time). And an analysis of market efficiency using the Castrol Index should begin by
looking at players with significant playing time, or risk losing the value of an objective performance
measure.7 And with more questions than answers, active debates and the opportunities for teams to
gain advantage will only continue in the future.
As a lawyer who negotiates contracts on a regular basis, I understand that every negotiation is tied to an
underlying assessment of value. I recommend that front office types learn the language of the Quant in
order to better support their technical staff, and make their business more efficient.8 Brad Pitt may play
you in a movie.
1 Jonathon Wilson, Inverting the Pyramid, p. 244.
2For an insightful report of the conference from a soccer perspective, and an interview with Drew Carey, Zach
Slaton’s SSAC entries at http://www.abeautifulnumbersgame.com/ are a must read.
4A few of my frequently checked sites include:
http://www.futbolforgringos.com/ (for analysis of tactics with a quantitative flavor).
6See full article at http://www.soccerbythenumbers.com/2012/02/how-efficient-are-player-salaries-in.html
7And all of this was done by stat-loving fans looking to avoid real work.
8For reference, the mating call of the Quant is “who wants a Guinness”?