Some stats are simply codswallop...not these

GUEST COLUMNIST

- CHRIS HEWETT

Some years ago, a small group of Twickenham modernisers introduced a trendy statistical software package designed to identify the chief “influencers” in England games.

Invariably, this oval-ball algorithm ended up highlighting the wrong blokes. Which is what we’ve come to expect from algorithms.

Sports statisticians often make themselves out to be new Messiahs, but all too often they are Messiahs without a message. Or, even worse, Messiahs with a misleading message.

Far from being cutting-edge, it is an age-old story going all the way back to the end-of-season averages in county cricket, which told you all you needed to know about Geoff Boycott’s contribution to the Yorkshire cause except the context in which it was made.

Did he win games by crushing the spirits of opposition bowlers with his obduracy, or lose them by scoring at the pace of a superannuated snail? This was something the numbers never divulged.

World Cup rugby is especially prone to statistical interpretation, some of it utter codswallop. For instance, the pointy-heads never tire of insisting that defence is the most important indicator of success at global level: that the meanest-minded, least porous teams are inevitably the winners-in-waiting.