Big Data & Analytics At The World Cup

The recent Brazil World Cup was the most data heavy yet


People have called it the best World Cup, ever. What it wasn’t was predictable - many had Brazil as the tournament favourites and they capitulated 7-1 to eventual winners, Germany. Spain and England were knocked out after two games and Costa Rica got to the quarter-finals. For many, the results we saw at the World Cup were proof that football is the antithesis of sports analytics.

This notion was encapsulated by Nate Silver, former New York Times and now ESPN analyst, who rose to fame during the 2012 presidential election. The power of prediction seemed to flow through his veins - there was a time when he just didn’t get anything wrong. Their Soccer Power Index (SPI), which he swore by, gave Germany a 35% chance of victory and a paltry 0.022% percent probability of them scoring seven or more goals (about 1 in 4,500).

There is nothing remarkable about a 65% favourite losing - upsets happen and data shouldn't be judged on the accuracy of every single result, it is after all a tool for prediction, not a way of guaranteeing the outcome of a game. The issue was that the result was so far off - based on the ELO rating shift it was the most unexpected scoreline of all time. It leads me to ask whether Big Data should be ostracised from football?

One team that would 100% be against excluding Big Data and analytics is Germany. Many often refer to the crowd as the 12th man, but for the German’s the 12th man man was data. Their pre-match preparation was transformed by analytics - they analysed player performance and measured key indicators like the number of touches, movement speeds and average possession time. In 2010, they set an objective to reduce their average possession time - with Big Data playing a pivotal role. In 2010 their average possession time was 3.4 seconds and in 2014 it was down to 1.1 seconds. For many, this was the essential cog that allowed them to win the World Cup and why teams like Brazil and Spain, who were playing noticeably slower games, failed to live up to expectations.

The analysts might have got it wrong with their pre-World Cup predictions, but to go as far as to say that Big Data is unwelcome in the football word is a dangerous precedent for professionals looking to get the best out of their players. If analytics can work for the German’s, there is no reason why it can’t work for everyone else - and after that who and what will be the 12th man?


Read next:

Top 5 Uses Of Big Data Outside Business