Is a Moneyball Approach to Soccer Pointless?

A good friend and regular reader of the blog sent me this story that appeared on Slate's website on Thursday.  In it, the author argues that fans' sabermetrics-style analysis is going to be limited by soccer's fluid and non-quantifiable nature as well as the big-business approach being taken to limit fans ability to access the statistics that are captured.  There are some fine points that I'd like to explore here.

First off, the article starts out with the well documented transatlantic gap for the love for sports statistics.  The author concludes the opening paragraphs of the article by recounting the demands of other commentators.

It's been argued, seriously, that all soccer needs to become a major sport in America is a better suite of stats. Give us the hard numbers, and we will give you Peoria. 

I would point out that I don't believe US soccer fans want any of the insta-stats available in other sports' televised coverage.  Frankly, they're often of dubious value in understanding the match at hand and are used simply to fill the breaks in play.  Soccer's penetration into US sports culture is held up by much bigger issues, ones I won't recount here but can be easily understood by reading chapters 2 and 3 of Gaming the World.  Slate's already imagined what such a stats-filled soccer match would look like (see video below), and let's all just agree that it would destroy what makes the game of soccer unique - uninterrupted, pure action.

It's that uninterupted chaotic action on the pitch that makes quantifying the game so difficult.  Sure, guys like me can track discrete statistics that represent snapshots in time - like halftime leads, the impact of the difference in the numbers of games played on playoffs, World Cup team and player ratings, and how transfer costs impact the English game - but being able to identify individual player statistics that can be used to predict success is much more difficult in a game where player contributions to team success are much more difficult to quantify. Such statistics are much more costly to come by, and thus teams are spending a good bit of money to develop their own models and keeping the data proprietary.  This should come as no surprise, as unlike US professional sports European soccer clubs are independent businesses that must justify their participation in top flight soccer by avoiding the drop or earning promotion from lower leagues.

The article does a good job pointing out that there are many routes to winning matches and trophies, so models developed by one club may not be as applicable to another club.

Teams operate under divergent tactical philosophies, player movement is interdependent, and a swarm of unanalyzable factors like "collective motivation" and "unity of vision" contribute something (but how much?) to every outcome. Stats can help soccer clubs, Gerrard says, with metrics designed around the clubs' specific needs.

I made as much of a point in the Liverpool case study of my post on squad transfer costs.

[Liverpool] also seem to have occupied an interesting position in the Big Four. Chelsea has spent absurd amounts of money to compensate for the manager carousel they’ve experienced. Manchester United has been able to combine both high expenditures and management stability to set the standard for championships in the Premier League. Arsenal has relied on the genius of Arsene Wenger to keep them competitive with a modest MSq£. Liverpool seems to have had the worst of both worlds – a high turnover in managers and a very modest MSq£ compared to the big spenders they were chasing.

To implement tactics a manager must have at least time, and preferably money, to build the team that he wants.  One can compensate for the time with a ton of money, but when a club gives neither it can be hard to assemble a quality side that implements a consistent strategy.  To compound matters, management and player turnover means that statistical analysis loses its efficacy as the few controlling factors in the model are constantly changing.

What this means is that bloggers and the readers of their blogs need to have realistic expectations as to potential content.  We'll keep working on trying to quantify the beautiful game - clearly the last decade or so of writing indicates we can continue to make improvements.  There are limits though - differences between the way soccer and baseball are played and financially controlled will lead to dramatically different access to data in the two sports.  I foresee an increased understanding into what enables a team, on a macro scale, to compete at the top levels of soccer.  Generating a unified theory of soccer success or player ratings is likely beyond the capabilities of the not-for-profit soccer community.

In the end, we all need to realize the limits of such models.  This may sound oversimplified, but my favorite quote regarding modeling complex systems is the one at the top of my blog.

Essentially, all models are wrong, but some are useful

The real question is how much more useful are the clubs' models than those being developed in the public domain? We may never know.

What is clear is that the author of the Slate article has an issue with applying statistical theory to the game of soccer.  After the chaos that was Liverpool's final day in the January transfer window, Brian Phillips provided the following commentary as part of a larger post:

Now, again, none of this means that Liverpool haven’t had a good day, or a good January, in terms of improving the squad. I’m reserving judgment until we know who’s actually in it, but at first glace I like Suárez and Carroll together, and as much as I love him, Torres isn’t worth £50 million soaking wet. (Well, maybe if you throw in the helicopter.) It just means that there are some interesting structural reasons why a regime of Puritan fractions has so far failed to overthrow the club’s usual sentimental chaos. Which, for lovers of sentimental chaos, may not be a bad thing after all

I think this is a bit too much wishful thinking on the part of Mr. Phillips - thinking that is interested in proving his point that cool-headed analysis can't prevail.  Apparently, he has not looked at Arsene Wenger and Arsenal, who are the best at beating the cash-rigged EPL system.  Additionally, it all seems like everyone wants to quote one aspect or another of Moneyball or Soccernomics, when neither book was written to address the crisis that Liverpool faced.  In fact, Liverpool probably did very well given the hand they were dealt:

The Moneyball theory is often misused and misquoted. Sabermetrics is bandied about regarding John W Henry as if everything from baseball will apply to football. He knows that’s not the case...

Liverpool have replaced a striker about to turn 27 with two who average out at 23. As I said in yesterday’s piece, the club have bought a strike-force. Torres missed a lot of games in the past two and a half seasons; and with just one of him, he was being paid while the team was shorn of his talents. Carroll may not be as good as Torres, but if he’s fit more often, and if his attitude is better, he’ll be of more use.

Responding to such crises, and then building a successful organization, can be done two ways.  The first is to pursue the Chelsea way, which is to just throw money at every problem even if it means you spend more money in one day than your entire annual loss the previous year.  The other is to build a system from the top down that defines the entire organization, one you're confident in when everyone else doubts you, one that allows the club to get more from less when it comes to budgets, transfer or otherwise.  Arsene Wenger and Arsenal realized they couldn't compete via the Chelsea method if they were going to establish their long-term financial solvency via the building of Emirates Stadium.  Thus, Arsene Wenger dismantled the Invincibles, focused on buying young talent inexpensively, and has now made Arsenal the envy of everyone in the league not named Manchester United or Chelsea by consistently finishing in the Top Four when they should be tenth.  They are the outlier, the anomaly, the guys who march to the beat of their own drummer no many how many times they hear about "not winning silverware the last six seasons."  They're playing the long-game, while Chelsea and Manchester City have management carousels constantly overpaying for players all the while playing in stadiums that house two-thirds the people that the Emirates does.

I sense Liverpool's ownership is attempting to follow the Arsenal model.  What some see as panic buys others see as crisis management meant to maintain the positive momentum that has only recently returned to the organization.  You can call it Moneyball.  You can call it Sabermetrics.  Whatever it is, it is a level-headed systematic approach to player valuation.  Such an approach is critical when faced with the chaos of this past weekend and building for the long-term.