Why bad balls get wickets

This is an extract from The Best: How Elite Athletes are Made (Nicholas Brealey Publishing) by Mark Williams and Tim Wigmore, a book that looks at how champions develop the skills that allow them to perform remarkable feats under extreme pressure

How do you enfeeble one of the greatest cricket batsmen of all time, and make them resemble an incompetent club player on a park green?

AB de Villiers - who reached number one in the world rankings in both Test and one-day international cricket - had made 46 runs in a Test match against Bangladesh in 2008. Then, Mohammad Ashraful - an occasional spin bowler who was far off the standard of the other bowlers - came on.

Ashraful's very first ball pitched so short that it bounced twice before reaching de Villiers - the sort of delivery that gave copious time for even a novice player to hit. Only, de Villiers hung back in the crease, bewildered to be facing a delivery of such ineptitude. De Villiers tried to thrash the ball through the leg side, but - his characteristic timing completely awry - he got a top edge, which looped up on the off side to Ashraful, who gleefully took the catch. For several seconds de Villiers remained in his crease, apparently thinking that the delivery was a no-ball - as balls which bounce more than twice are, but those which bounced only twice, as Ashraful's did, were not at the time.

"I was hoping for someone to call it a no-ball," de Villiers said after play. "It is the first time in my career that I have been dismissed in such a fashion."

In cricket, deliveries that would be considered bad balls at any level of the game - those that are very short, like Ashraful's to de Villiers, misdirected and wide, or full tosses that do not bounce before reaching the batter - get a curiously high number of wickets at the top level of the sport. It isn't that elite batters cannot play such deliveries - these are exactly the sort of balls they have been thrashing away from their childhoods. Batters like de Villiers can struggle against such deliveries because they aren't expecting them, and the balls are totally incongruous given the high standard of the match.

Information like the placement of fielders, the previous deliveries from a particular bowler and the bowler's run-up prepares batters for what delivery will come next. But, in top-level international cricket, the cues derived from this information almost never point to batsmen receiving wide balls that bounce in an easy way, at what is known as "half volley length". Paradoxically, this is what can make such deliveries lethal.

Players are more likely to make mistakes when the delivery does not align with the cues available before, and as, it is delivered. This tendency can make them vulnerable to deliveries that do not align with their expectations - sometimes because they are brilliantly disguised and executed, and sometimes because they are simply egregious balls.

A study in 2019 showed that elite batsmen make more errors when the ball they face does not match the contextual information that precedes the delivery, including when the ball bowled is conventionally thought of as very bad. While elite batters are far better than lesser players at reacting to deliveries when the contextual information aligns with the ball, the errors made by elite batters increase markedly when the contextual information does not align with what happens. Elite batters seem to actually play such balls worse than sub-elite batters, presumably because they are less used to them in the standard they play. So when the best bowlers are bowling to the best batters, just occasionally deliberately delivering a bad ball can be a smart ploy.

"Skilled cricket batters can start to anticipate the possible line and length of the next delivery based on the score, field setting, state of the game, and previous deliveries from that bowler - not as a 'pre-meditation' but a narrowing of possible outcomes," explained Oliver Runswick, a scientist at King's College London and one of the authors of a study on why bad balls get wickets. "While this offers a significant advantage in playing most deliveries, it can have its downsides. When less probable deliveries come, the batter is more likely to be dismissed."

These findings were confirmed by research into full tosses - balls that don't bounce before reaching the batter, widely considered very poor deliveries - by Liam Sanders from Loughborough University. Analysing 70,000 balls by spin bowlers in Test matches from 2006 to 2015, Sanders found that, remarkably, full tosses got wickets more regularly than any other length of delivery. All the balls that would have bounced under 1 metre before the stumps - and so reached the batter before they had bounced - got a wicket every 27 balls, compared to every 61.2 balls for deliveries that bounced 4-5 metres before the stumps. The conventional wisdom about the worth of these two types of deliveries is not wrong - on average, full tosses conceded almost eight runs per six-ball over, while balls that bounced 4-5 metres before the stumps conceded just 2.2 runs every six balls.

But it is the very rareness of full tosses in Test cricket that can make them potent. In their desperation to hit full tosses some batters can be over-eager and make seemingly inexplicable mistakes, Sanders said. "It seems there is a significant negative impact on anticipation and shot execution in skilled batsman when pattern recognition was broken - a lack of agreement between contextual information, such as the previous deliveries received, and the outcome of the event that followed."

This is why, paradoxically, the worst deliveries in cricket can be effective at taking wickets. Even when their opponents don't mean it, elite athletes can be conned.

Mark Williams is a scientist, author and educator, and an authority on expertise and its acquisition in sport. Tim Wigmore is the co-author of Cricket 2.0: Inside the T20 Revolution