Thanks to the digital revolution (of which I understand little), sport today is awash with what is now called data. Coaches and selectors, journalists and those who watch sport on television have more information available than ever before. This matters a lot to selectors and coaches – often nowadays the same thing. In Rugby at least, the days of the semi-detached panel of selectors are gone.
The ECB (England and Wales Cricket Board) has just appointed a new chief selector, Ed Smith, who played county cricket for Middlesex and Kent, had a brief Test career and, since retirement, has been a successful journalist and author, writing on a wide range of subjects. He once wrote that “the most underrated force on the cricket field is selection. Once the players are on the pitch, they are essentially on their own. All the more reason to get the right ones there in the first place.” Well, yes, we can agree about that; it’s obviously true. But on what basis do you make your pick?
Ed Smith, a highly intelligent man, is, I would guess, sceptical, even dismissive, of what you might called the “hunch and a prayer” method. Reviewing Michael Lewis’s book Moneyball in The Spectator, he found it “an example of how the scientific method could revolutionise unreformed but apparently sophisticated businesses.” Collect and analyse the data and draw the right conclusions.
The crude data – statistics – may be interesting but misleading. Information may flash up on your screen – for example, “metres made”, “offloads made”. Interesting, of course, but how much does this tell you? A full-back may make lots of metres if the opposition kick deep and don’t chase their kick, so these are easy metres, often of no great significance. As for an offload, was it well-timed and effective, creating an opportunity for the recipient, or was it made clumsily or under pressure, putting the recipient in deeper trouble?
How do you choose between two goal-kickers? The percentage of successful kicks is obviously a guide, but a very crude one. It takes no account of the difficulty of the kicks, of the spot from which they were taken, of pitch and weather conditions, or, indeed, of the ambition of the kicker. A goal-kicker who declines kicks from a position near the limits of his range will probably show a better success rate than one who attempts such kicks. What does this tell you about the two? It certainly doesn’t determine which is the better kicker. Nor, of course, does the success/failure percentage say anything about the significance or importance of the kicks taken. A goal kick when you are 20 points up in the last quarter of a match is surely less demanding that one taken from the same position when you are two points down and time is almost up.
Over the years, cricket has been more in love with statistics than any other sport, though golf these days must run it close and others are catching up. Yet, there is perhaps no sport in which, on account of its peculiar combination of individual and team contests, where figures – unanalysed data – may be more deceptive. Cricket followers have always recognised the difference between hard runs made against good bowling in difficult conditions, and soft runs made against a mediocre attack on an easy-paced pitch. An innings of 40 such hard runs won’t do as much for the player’s average as one of 150 easy runs, but may, nevertheless, be a better guide to quality. Certainly a good selector will know which innings to value.
Character is something that may not be revealed by the data. Nobody would regard Paul Collingwood as one of the most talented English batsmen of the last 20 years, but he was more valuable than many who scored more runs, because he had the ability – character, resolution – to be at his best when it mattered most. Likewise, for a couple of seasons now, lots of English rugby journalists have been saying Dylan Hartley’s time should be up, because he’s not the best hooker available. Eddie Jones has stuck by his captain, and not simply from obstinacy. Evidently, he finds that Hartley contributes something to his team that the data may not reveal.
In the old days, before computers, the way of judging a player’s performance was to watch him throughout the match, while paying little heed to the overall game. Now the data give coaches and selectors a lot of obviously very useful information which, properly analysed, will inform their judgments. Yet I suspect that it’s a timid coach who relies most on data, and that judgment of character and simply a hunch may still be decisive. Anyway, how long will it be before hackers intervene and coaches, confronted with lots of fake data, have to use their eyes and trust to their experience when they come to making a judgment?