During the last World Cup, three companies made predictions on the results of the final phase of fifteen matches. They were demonstrating the ability of their advanced technology to predict the outcome of football matches. Microsoft and Baidu correctly predicted all fifteen results while Google made only one mistake...
How were they able to make such accurate predictions? They crunched and analysed large numbers of historic results - what we call "big data" - and used that analysis to make their successful predictions.
It seems reasonable, therefore, to ask ourselves whether big data is changing the paradigm of the sports industry?
As is often the case, the answer can be found in the United States. In Chicago, for example, if you want to talk about yesterday's Steph Curry performance for the Golden State Warriors, people will talk about his performance stats, but in Europe it is very different. In Europe, we still predominantly talk about sports with our heart, in an emotional way, much more than with our brain. In Paris or Rome, fans will most probably tell you that sport is an art, not a science. Europe has started to follow the stats and data movement, but America is clearly showing the rest of us the way.
Here are a few fascinating examples:
- In the National Football League (NFL), players have sensors that provide impressive information to coaching staff, such as heart rate, lung capacity and body temperature;
- in Major League Baseball (MLB), their analysis tools use complex algorithms capable of tracking and analysing thousands of terabytes of data, to correlate events related to every movement on the field captured by camera;
- NBA basketball has installed motion tracking cameras in every arena to analyse player movements, break down shooting percentages, and even send information to the team trainers and doctors about players' endurance levels.
European football teams have started to use statistics much more over the last decade and the methods are becoming more and more sophisticated. The first club in the Premier League to do so were Bolton Wanderers and just a few years later almost every club in the top flight are using statistics to monitor player performance. Manchester City, for example, have a team of full-time statisticians that share information with the first team coaches.
A few years ago, Bayern Munich entered into a very interesting partnership with SAP, the multinational software corporation. The result was so effective on team performance that SAP also partnered with the German national team during the last World Cup. After winning the competition Joachim Löw said in several interviews that the partnership gave them a genuine competitive advantage.
It therefore seems that big data is genuinely changing sport, but it is also changing the games that are based on sports.
In our case it has enabled the creation of daily fantasy sports (DFS). The promise that our sector is making to customers is very simple - with fantasy sports, for the first time, you are able to prove to your friends that you know a sport better than they do. Therefore, the results have to be as close as possible to the reality of play on a football pitch.
If you have an encyclopaedic knowledge of football, you should reasonably expect to beat your friends that only have a social interest in football more times than they beat you. However, if the result of the game is random, your friends might beat you more frequently, which is obviously against the natural order of things. If fantasy sport is not a skill game then it loses much of its purpose and fun.
Claiming to be a skill game is very simple, anyone can do it, but actually becoming one is a much bigger challenge. At Oulala, we worked for six months with a team of statisticians to build a mathematical matrix that makes the results of our game as close as possible to reality. We then tested our scoring system for more than a year to prove that Oulala really is a skill game. We are constantly trying to improve it because this is the cornerstone of our game.
To give a real life example of how the use of data could evolve... if a team was to measure the serotonin level of the players, this could in turn give the coach an indication of the real time stress levels of each player live in any game. Before deciding which player will take a penalty, therefore, he could look at the data and make a decision by considering the fluctuating energy or stress levels of his players.
Some may ask whether this is a step too far, and there will clearly be people on both sides of that argument. There will undoubtedly be an increase in the number of tech individuals and organisations showing an interest in sports and looking to apply their skill set to this realm in new ways. Equally, there will be a number of traditional supporters who will speak with tears in their eyes about 'the good old days' when sport was not driven by numbers and data. However, we are confident about one thing, and that is that the desire to win will mean that very few professionals will want to be left behind by this revolution, quickly adopting any tactic that may help gain an advantage over their competitors.
Valery Bollier is CEO and co-founder of daily fantasy football game www.oulala.com