Predictive Analytics and the 2002 Football World Cup Finals
We’ve been saying data science has prevailed for long. To substantiate this fact, let’s time travel a bit to the year 2002 football world cup finals. Ronaldo scoring a goal for Brazil in the 2002 World Cup final gave joy to millions of supporters all over the world. However, there was one Irishman who was perhaps happier than even some of the most loyal Brazil fans.
Dr. Peter O’ Donoghue’s entire reputation depended on the final. He was perhaps the most anxious of the 2.5 billion worldwide TV audience. He was a sports studies lecturer in the University of Ulster and built a sophisticated computer model to predict the outcome of the final before the tournament even started.
The model predicted that Brazil, universally regarded as the greatest side in the world, would lift the World Cup that year and that their final opponents would be Italy, who knew a thing or two about winning tournaments.
Dr O’ Donoghue spoke to the BBC before the tournament began. He said, ” This is number crunching versus subjective judgement. The pundits have predicted Argentina to beat Italy in the final. But I’m confident. My model takes into account FIFA rankings, distance travelled and rest between games. I’ve run the simulation two thousand times and it’s definitely Brazil to beat Italy.”
He was so confident that for the first time in his life, he entered a betting shop – £20 on the exact outcome. Italy couldn’t progress beyond the round of 16. Despite the £20 loss and jibes from fellow academics, Dr O’ Donoghue was relieved to see Brazil cruise to the final.
Warwick University’s Glover Automated Results Indicator had predicted an Argentina triumph by some margin. After they failed to get past the group stage, their spokesperson said, “We’re far from perfect. In reality, the point of football is for us to marvel at the underlying chaos and unpredictability of the game.”
He basically meant that his sophisticated model had failed badly.
The same could have happened to Dr O’ Donoghue. But Ronaldo’s goals saved him from the embarrassment. Though it took a couple of days for the media to make headlines that read ‘ Computer predicts World Cup winner,’ the old mathematician beamed with pride when they did.
Machine Learning. Artificial Intelligence. Data Science.
Ponder over these as much as you want but it will be very difficult to beat Dr O’Donoghue’s work in 2002.
Or was it just the luck of the Irish?