Soccernomics certainly sounds interesting enough, but I wasn't terribly impressed. The authors raise some good points, and puncture some long standing common wisdom type myths about sports in general and soccer in particular. But there were quite a few times where I, as a mathematically literate, highly educated soccer fan who has some feel for the more recent history of game sat up and said 'well yes, but you're ignoring...'
As an illustrative example, in one section, the authors are considering attendance numbers for different European leagues. They point out the relatively high numbers for certain leagues, then point to low numbers in Italy in the period they're considering- pointing out that in 2007-2008, average Serie A attendance was only 23,180. Even Juve's average attendance was only 20,930! Well, yes, that sounds dire- until you recall that Juve were playing in Turin's Stadio Olimpico while building their new stadium, and the Olimpico at that time only had a capacity of 25,500. Moreover, look at the
capacity of the stadiums of
2007-2008 Serie A1- average capacity was 41,200. But even this average is skewed upward by the Olimpico in Rome (Lazio and Roma), the San Siro in Milan (AC Milan and Inter)
2, and the San Paolo in Napoli
3, all of which are significantly larger than other
Serie A stadiums4. Three teams played in stadiums with capacity below the average Serie A
attendance, while another two had capacity roughly equal to the average Serie A attendance- so for a quarter of the teams in Serie A that season, achieving the average attendance would have meant a capacity or overcapacity crowd. Finally, the statistics fail to account for any events that might affect attendance, such as disciplinary stadium bans (and there were several) or the Sandri death. Clearly simple average attendance numbers do not present a meaningful comparison.
In other words, while the data the authors present is technically accurate, their interpretation and presentation is not. A more accurate comparison of attendance numbers across leagues would compare an average of attendance vs. capacity- and using that metric, half the league averaged two thirds of its official capacity or better. Juve, the supposed example of poor Italian attendance, led the table with an average attendance of 81% of its official capacity!
It would be one thing if this were the only such instance, but as I said, there were several. (I wish I had kept a list.) It has been said that there are lies, damn lies, and statistics, and no one knows how to produce misleading statistics and present them convincingly quite like economists. The charitable interpretation is that the authors do not mean to mislead, but have simply committed the error of not examining the context of their data. The book is certainly food for thought (something the game could do with a little more of) but reader, beware- if something the authors say sounds a bit odd, you should probably do your own research.
1. Juventus is incorrectly listed as still playing at the Stadio delle Alpi.
2. Given that most people understand 'average' to mean the mid-point, it may be more practical to take the median capacity (32295) rather than the average to account for the skewing effect of the San Siro and Olimpico appearing twice in the data set. In fact, only seven teams- the 5 names above plus Fiorentina and Udinese- played in stadiums that had a capacity equal or greater than the Serie A average.
3. I can't speak for the San Siro or Olimpico, but I know that the San Paolo does not routinely open all sections for league games- there are sections opened only for the highest intensity Serie A ties, or European/international matches, so actual ticket capacity for most league matches is lower than the official stadium capacity. Again, this skews the disparity between capacity and attendance numbers.
4. Note the link is for all football stadiums in Italy, not just the ones that were Serie A in 2007-2008.