Predictive analytics at World Cup

By Roopen Roy Jul 15 2014

When economists and investment bankers come out of their zone of comfort and predict the outcome of events, they usually commit blunders for many a time, they know nothing about say football or elections. But that does not stop them from predicting. They claim to use predictive analytics and stochastic models. They are often wrong but seldom in doubt. One wonders if they are emulating what Winston Churchill famously said about politicians: “A politician must be able to predict what will happen tomorrow, next week, next month and next year and then explain why it did not happen.”

Goldman Sachs crunched data on 14,000 past matches and produced a 67-page report and predicted that Brazil would win this year’s football world cup. In their own words, this was the methodology used: “The predictions for each match are based on a regression analysis that uses the entire history of mandatory international football matches — that is, no friendlies — since 1960. This gives us about 14,000 observations to estimate the coefficients of our model. The dependent variable in the regression analysis is the number of goals scored by each side in each match. Following the literature on modelling football matches, we assume that the number of goals scored by a particular side in a particular match follows a Poisson distribution.”

Excuse me? Why would you predict something by looking at the rear view mirror? The longer back you go, higher would be the chances that you will pick up the wrong trend or pattern. By crunching mobile phone sales data in the 1990s, could you have predicted the success of Apple in mobile telephony? Never. Often a competitor emerges out of the blue and not from the known set of competitors. Sometimes, there is a re-arrangement of the pecking order.

Goldman Sachs has also published a paper titled “When will China and India play in the World Cup Final”. Here is their prediction, “There is still a lot to be done to lift Indian football to international standards. We think that even a 20-year horizon is probably too short for India to make it to the World Cup final.” With the track record of Goldman Sachs in predicting football match outcomes, I am beginning to see glimmers of hope for Indian football. And guess what? I may have a more than even chance of seeing it in 2034!

Technology companies are not far behind in predicting or taking credit for successful outcomes either. SAP, for instance, has partnered with the German national football team. They have developed a tool called Match Insights which analyses video data from on-field cameras capable of capturing thousands of data points per second, including player position and speed. That data then goes into a SAP database that runs analytics, allows coaches to target performance metrics for specific players and gives them feedback. Though SAP has not predicted the outcome of football matches, it claims that Big Data analytics is a decisive weapon in the hands of the German team. Clearly, success has many fathers.

Predictive analytics, however, is not voodoo mathematics or a branch of astrology or palmistry. It deploys statistical techniques from modelling, machine learning and data mining that analyse present and historical facts to make predictions about future, or otherwise unknown, events. Black swans are, by definition, impossible to anticipate. But predictive analytics and scenario planning do help companies and nations to be more prepared for volatility, surprises and grey swans.

Yet, nothing succeeds like success. Paul, the octopus, would have been long forgotten but for its large number of correct predictions. On the other hand, while Apple’s Siri has been giving ambivalent answers, Microsoft’s digital assistant Cortana has accurately predicted outcomes in 15 straight games including the World Cup final which was decided in extra time in a sensational goal scored by a substitute player Mario Goetze. That is a mind-numbing and awesome track record. I do not know how Cortana managed to do it, but it admits to have also crowd-sourced the betting trend insights. Here is what it says about its methodology: “Our models evaluate the strength of each team through a variety of factors such as previous win/loss/tie record in qualification matches and other international competitions and margin of victory in these contests, adjusted for location since home field advantage is a known bias. Further adjustments are made related to other factors which give one team advantages over another, such as home field (for Brazil) or proximity (South American teams), playing surface (hybrid grass), game-time weather conditions, and other such factors. In addition, data obtained from prediction markets allows us to tune the win/lose/tie probabilities due to the ‘wisdom of the crowds’ phenomenon captured by the people wagering on the outcomes.”

Thus, even in business, big data analytics coupled with crowd-sourcing of gamblers’ wisdom could be a new formula of success for the future!

At the end of the day, Cortana made exact predictions while Goldman Sachs and others got it wrong. It reminds me of what Abraham Lincoln said, “I do the very best I know how; the very best I can, and mean to keep doing so until the end. If the end brings me out all right, what is said against me won’t amount to anything. If the end brings me out wrong, ten angels swearing I was right would make no difference.”

(The writer is managing director of Deloitte Consulting, India. These are his personal views)