The Runner Sports

Leave Rugby Number-Crunching To The Coaches – Just Sit Back And Enjoy The Magic

Late last year, British-based artificial intelligence firm ASI Data Science signed a deal with then-English Championship London Irish to help it find undiscovered rugby players.

It was one of those worthy rugby data stories that disappears into the ether almost as soon as it was announced. Maybe it shouldn’t have been, because it highlighted a data-driven trend that’s threatening to take the magic out of the game.

Upon entering the name of a player, the software matches statistics taken from sports-data company Opta to find other players who boast similar styles and records. According to the press release, the machine learning technique could remove biases that ‘cloud judgments’ of scouts and coaches when making new signings.

It was sold in those sections of the press that follow this sort of thing as the next-generation Moneyballing of rugby.

It all sounds very clever –and the sales pitch was compelling given the terrific what-happened-next story of the Moneyball film. And it completely misses the point. The direct replacement of players was never an issue. It was all about building a competitive team.

We all know the Moneyball story. Billy Beane, the general manager of Oakland A’s, was looking to build a team to challenge for the playoffs as cost-effectively as possible. So, he –and his partner in sporting statistical crime Peter Brand– crunched some numbers, worked out how many games they needed to win, and hired a team of misfit players whose averages fit their calculations. The rest has gone past history into the realms of folklore. After a slow start, this team of budget journeymen went on a record-breaking 20-game winning streak.

It’s one of the great sporting romances –and a recruitment model that has been adopted, adapted, refined, and improved ever since.

Here’s the trick of it. Beane rejected the idea that he had to find exact replacements for players he had lost to bigger clubs with deeper pockets. He –and Brand– realized that they could make up the numbers by combining players whose averages recreated the whole.

In rugby terms, he was not looking for the next Jonah Lomu to replace an irreplaceable original. How many times has the rugby world been ‘treated’ to the notion that this or that player is the new Jonah Lomu? How many players have been unfairly scarred by an unnecessary association that they cannot hope to match? At the end of the day, there was, still is, and will forever be only one Jonah. Others have scored more tries, and more regularly, than Lomu. But that doesn’t matter. He’s Jonah.

Pick your sport. In soccer, there is only one Pele, Cristiano Ronaldo, or Lionel Messi. In athletics, only one Usain Bolt or Michael Jordan. In boxing, only one Muhammad Ali.

It also gets to the nuts and bolts of rugby’s overwhelming 21st-century obsession with data. The focus these days is always on meters made, defenders beaten, tackle counts, lineouts and scrums won. It’s fallacious. Now, it is being used to publish ‘teams of the tournament’ that bear no relation to reality. Super Rugby’s recently published stats-assembled team of the season, for example, contains no players from Crusaders, the team that won the title.

British and Irish Lions’ coach Warren Gatland has an in-public hate-hate relationship with statistics. He has some sort of point. There can be no doubt data was used to help him select his squad. There can be no doubt he crunched some serious numbers ahead of and during the drawn tour of New Zealand in June. But, to him, they were not the be-all and end-all. They were a tool that helped him get the best out of his players. Nothing more.

And nor should they be the first, last, and always for anyone else. Yes, data gives an indication of player and team performances. The numbers can reveal a lot if you know what you’re looking for, but they are a long way from the whole story –what good is meters made if the one stat that actually matters, the final score, is against you?

Lomu wasn’t just the tries he scored, the defenders he beat or ran over, or the meters he made when half the opposition were hanging off him, desperately trying to stop him moving. He was magic. No amount of data can replicate that.

James Harrington
Follow Me

James Harrington

Cheese-eating, wine-drinking, France-living freelance sports journalist. Doting husband of one. Sickeningly proud father of three
James Harrington
Follow Me