One of the world's leading rugby teams is about to learn whether the sport, once described as "a ruffian's game played by gentlemen," can benefit from the genteel field of science.
The Leicester Tigers are the most successful English rugby union club of the past 25 years, winning the Premiership title a record nine times and twice capturing the European championship. Later this month, they'll play in their eighth consecutive league final. In an effort to maintain that impressive success, the club is working with IBM to use statistical analysis software in assessing the likelihood of injury.
In the past, Big Blue has used its SPSS Modeler predictive analytics software to help clients like Coinstar find the most profitable location for its coin-conversion machines and show the Memphis Police Department where to deploy officers to reduce crime. It believes the same analytical approach can help the Tigers optimize training and keep the best players on the field.
Although the company's system can gather thousands of variables, only a surprisingly small number really matter, said Erick Brethenoux, IBM's director of Business Analytics and Decision Management Strategy.
"We've found that on average, maybe 15 to 20 variables are more than enough — not a thousand — to be predictive," Brethenoux said. "When one of those factors gets out of line, we can reduce training in some instances, or prescribe other types of training for a period of time."
Players wear equipment during training that captures statistics like heart rate as well as the forces absorbed by rugby's frequent jarring hits. Data are uploaded to the team's database, where IBM's modeling software looks for patterns. Additional info is drawn from player questionnaires that could provide insight into their stress levels and injury history.
Leicester used the program for a trial period this season, and the team's sports scientist, Andy Shelton, said the goal is to monitor the entire squad an entire campaign, gathering minute-by-minute data.
"We're at the very early stages," Shelton said. "We've run some models so far on stats that we've collected over the past five years and although they need refining, it's looking very promising in terms of injury prediction."
As usual, the devil is in the details. Or, in this case, the data. The success of the program depends not on the analysis, but on the reliability of the data being gathered, said Will Carroll, who writes the Under the Knife column about sports injuries for SI.com.
"How do you measure fatigue if you can't directly measure it?," Carroll said. "Data analysis is only as good as the data."
As professional rugby has grown in popularity around the globe in recent years, the influence of money has grown as well. With that maturation comes economic constraints for some teams, even successful ones like the Tigers. Although they dominated the British leagues during the past decade, Leicester has met strong challenges from some of Europe's top clubs. One reason for that is English teams have a salary cap of 4.5 million British pounds (about $7.2 million), while the top French clubs top out at about 7.2 million pounds ($11.5 million). If that sounds like a situation ripe for the Moneyball approach popularized by baseball's Oakland A's, you'd be right. Shelton said teams with big payrolls can easily sub in a top international player when one of their own stars goes down, but Leicester doesn't have that luxury.
"In one game, it might not make a difference, but over the course of a season and in the big games toward the end of the season, that's when it makes a big difference," Shelton said.
In addition to keeping their best players healthy, Shelton said IBM's system could also be used to develop young players from their sports academy into long-term productive pros.
"All of this is just a tool to aid the coaches and help the players along," Shelton said. "None of this will ever replace coaching or talent. If we can keep our front-line players fit for longer and if we can give a little bit of help to the coaches to select academy players or pick up on ones they may not have noticed, then it can only be beneficial."
