There is talk of a black ops stats department at the ECB, a group of well-funded geeks running offbeat numbers for Andy Flower, who is a notable convert to the ways of Moneyball ["It really opened my eyes to a different way of looking at stats... I don't think we've tapped the potential of how stats can drive our strategy... [The stats department] are doing some really interesting work. Some of what we think we know we don't want others to know"].
In this sharp-edged numerical wonderland, the simple batting average is now a lumpen tool, lacking the meaning that analysts like Flower need. Last week came The Bradman Class: An Evaluation of Batsmen For Test Matches 1877-2006, a further attempt at producing a figure that gives a value to the quality of performance locked away within the broader mean of the average. Andy Bull examines it nicely here.
What they're really in search of, but might never find, is the Ian Bell Number. Bell has the kind of Test average that denotes a fine player, a mid-40s figure that sets him alongside Strauss, Collingwood, Cook and below obviously better men like Pietersen, Sehwag, Dravid, Jayawardene.
Yet it's a figure that also puts him level with, for example, Graham Gooch and Gordon Greenidge, and above Mike Atherton and Alec Stewart. What's needed, think statisticians, is a degree of difficulty figure that sets a player like Bell in his proper context [the upside for Bell is that his number would be climbing, given his recent toughening up] .
Moneyball is a terrific book, but one of its outcomes has been the assumption that there is a statistical measure for everything [not, incidentally, a claim made by the book itself]. It also assumes that statistical data is predictive as well as reflective. Where humans are concerned, though, there is always that element of unknowability, and within that lies the true beauty of the game.
The case for Matt Renshaw
1 week ago