April 26, 2016

Does the decision to toss or not affect the pitch?

The further away a match travels from the coin spin, the harder it becomes to find solid evidence of advantage

Jonathan Trott is one of four double-centurions in the first month of the Championship played under the new toss rules © Getty Images

In a game as finely calibrated as cricket, small changes in method can bring about overwhelming changes in outcome. Here is the opening paragraph from chapter 13 of Simon Hughes' memoir of the graft and grit in county cricket, A Lot Of Hard Yakka:

"It was the pong of perspiration that dominated most changing areas in 1990, because bowlers toiled all day without getting anybody out. 'You're not going to like the balls we're using this season,' the umpire Alan Jones remarked one morning pre-season, and he was right. They were even more orange-like than we'd feared. The leather was dull and unpolishable, the seam imperceptible. It was like bowling with a large red mothball. Their arrival coincided with new, TCCB-defined straw-coloured pitches and the hottest spring since 1929. It was, in short, the prologue to the year of the bat."

As Hughes went on, ruefully, to point out, there were two triple-hundreds and a 291 by the first week of May, and at the end of a summer of carnage, ten players had made more than 2000 first-class runs, plus, astonishingly, 428 individual centuries and 32 doubles. All of this because of a slightly different cricket ball.

The tweak to County Championship cricket in 2016 is not to the ball but to the toss. It is, in fine style, a wonderfully baroque adjustment, entered into in the hope that a connected issue may be resolved by it. The central concern, set out well by the ECB's lead spin coach Peter Such, is a valid one: "I feel this is a way for the game to balance itself, because the statistics show us it hasn't been balanced in the County Championship in recent seasons - and specifically, that it has been balanced against spin bowling, and fast bowling, and has therefore not been preparing batsmen or bowlers for the challenges of international cricket."

It's a quote that presumes the point of Championship cricket is to serve the England team, but he who pays the piper calls the tune, and it's true that a drastic disconnect between first-class and Test cricket is in no one's interest. In other words, when the deadliest bowler in the land is Jesse Ryder, something's gotta give.

This is an argument about causality. Will the automatic option of a visiting captain being able to bowl first force the production of tracks that will make genuine pace and spin the most effective weapons? Or, put differently, is the outcome of the toss connected enough to the state of the wickets produced?

Late last year I had a conversation with Nathan Leamon, England's performance analyst. He is fascinated by the role of the toss as a system, and by what winning or losing it means, statistically. And while it's sometimes possible to assess its value early in a match, its overall effect on results is surprisingly resistant to interpretation. The further away a match travels from the coin spin, the harder it becomes to find solid evidence of advantage.

Of 11 completed Championship games, there have been six uncontested tosses. These have resulted in five draws and a loss. Of the five completed games that have gone to the toss, there has been one win, two draws and two losses for the captains calling successfully

Leamon visited the idea of granting the toss to an away captain in terms of Test cricket in a piece for the Nightwatchman, identifying three reasons, "which should make us question whether it would achieve its aims". The first is that it assumes groundsmen can reliably prepare pitches to order; the second, the law of unintended consequences - home teams might try to produce pitches even more favourable to their strengths; and finally that winning the toss may not be an advantage at all: "This [system] assumes that captains are able to use the toss to their advantage. This is not in any way proven. In fact evidence suggests this just isn't the case."

In Test cricket, the advantage of batting first - which, in the era of uncovered pitches was, as Leamon notes, "robustly successful" - has eroded. In 700 Test matches played since January 2000, the side batting first has won 36% and the team bowling first 39%. Within those results are so many variables of pitches, home bias, relative team strength and so on, that the toss is a hard advantage for the winning captain to successfully exploit - if indeed it's an "advantage" at all.

But that is Test cricket, played across the globe. County cricket enjoys a more localised environment in terms of conditions, and at a guess, the adjustment to the toss is aimed mostly at first-day pitches overly helpful to medium-pace seam bowling. Causality can be more closely linked.

We are at season one, round three of Championship games, which any statistician would suggest is too small a sample size to show any change. Yet, as Leamon points out, humans aren't designed to store vast amounts of data: "the brain encodes less than ten per cent of what we experience. The rest it simply makes up… faced with a huge number of random or near random events (a cricket match for instance), our brains pattern-spot, even when there is no pattern."

This is part of what the psychologist Daniel Kahneman calls "the availability heuristic". We make our decisions based on the most readily available or obvious pieces of information.

The spring of 2016 in England has been one of scudding clouds, chill winds and cold nights. The trees were late in bud, and the jobbing cricketer remains encased in several sweaters. I have laid eyes on two April pitches, one in Hove and the other at The Oval, and both sang out from the rich, wintry green of the outfields, dun in colour, perhaps even edging towards tan.

Of 11 completed Championship games across two divisions (there has been one washout in Worcester), there have been six uncontested tosses (i.e. the visiting side have elected to field). These have resulted in five draws and a loss. Of the five completed games that have gone to the toss, there has been one win, two draws and two losses for the captains calling successfully. Of the seven games currently in progress, only two tosses have been uncontested.

There have been four double-centuries (Bairstow, Robson and Trott in Division One; Duckett in Division Two) plus 20 other centuries. The leading wicket-takers are all brisk: Jake Ball (12 wickets), Keith Barker (12) and Neil Wagner (11) in Division One; and in Division Two, Graham Napier (14), his Essex team-mate Jamie Porter (14), and the veteran Aussie Clint McKay (10). Jeetan Patel (seven) and Gloucester's Jack Taylor (seven) are the first spinners to surface.

May is yet to dawn, the sun yet to beat down, but it would appear, to our feeble brains able only to process this superficial information, that the idea has had some effect.

Jon Hotten blogs here. @theoldbatsman