IBM TryTracker Brings Big Data To Rugby
IBM agrees five year deal as official analytics provider for the RFU
IBM has agreed a strategic five year Big Data agreement with the Rugby Football Union (RFU) to become the organisation’s official analytics partner.
The company hopes to build on technology used at Grand Slam tennis tournaments like Wimbledon, the Australian Open and US Open and believes that most sports have data that has not been captured or contextualised.
Big Data in sport
As part of the deal, IBM will provide the analytics for IBM TryTracker and aims to use its data to assist the RFU in understanding rugby fans and players.
Speaking at a demonstration of TryTracker, Matin Jouzdani, Big Data Lead UK at IBM said that the company believes that it is part of wider trend of a smarter planet where data can transform every walk of life.
“We’ve very enthusiastically embraced sport as a metaphor of that,” he told TechWeekEurope.
Jouzdani said that different sports had a different level of statistical discourse, but that there was still the potential to offer a new kind of fan experience.
IBM Try Tracker
TryTracker aggregates a lot of data under three headings, which IBM refers to as “keys to the game”, momentum and “key players”. These are shown alongside more traditional components like live text commentary and TV-style statistics.
Three ‘keys to the game’, such as winning a certain number of lineouts or aachieving a successful goal kick percentage of more than 74 percent, will be released prior to each of England’s fixtures at this year’s Six Nations Championship. These are drawn from a large historical database of matches involving the two teams and if completed, will improve either side’s chances of winning.
For example, England achieved all three keys released by IBM prior to their 38-13 victory over Scotland on 2 February, while Scotland failed to complete any of theirs.
The second component of TryTracker is momentum, which is a visualisastion of the ebb and flow of the game created from actual match events such as tackles and metres gained and weights them accordingly.
On 2 February, the momentum never once swung in Scotland’s favour, something which made Michael Nicholson, sports analytics lead at IBM, a happy analyst, but a depressed Scot.
Lies, damned lies and statistics
After the match, three players who had the most influence on the match are named, with vastly different criteria for forwards and backs. Curiously, England fly-half Owen Farrell was not among England’s three ‘key influencers’ despite kicking four penalties and three conversions.
This, Nicholson explained, is because goal kicking is excluded as kickers would always be disproportionately favoured and that you could argue was the result of other players’ work. He said that kicking could eventually be factored in and that it would continue to tweak and experiment with the process of selection. He added that it was possible to name the three worst players, but IBM has elected not to do so.
IBM hopes that the data will be accessible but still useful for experts and that there was no jargon or anything mathematical involved
“We’re not trying to be elite analysts,” said Jouzdani, who added that the aim was to make the data available for everyone while still being useful for experts. IBM TryTracker is available on PC, smartphones and tablets throughout the tournament.
Are you a tech Olympian? Find out with our sporting IT quiz!
