What scores produce the most competitive ODI cricket?
|May 27||Public post|
My favourite kind of ODI is the one where the team batting first makes a score between 220 and 260, and the team chasing loses an early wicket or two. My memory might lead me to a biased sample, but this kind of game usually leads to hard-fought contests. There will be a few tight overs here, a clump of wickets there, dogged batting by the chasing middle order, and the game going “deep” with the result not being known until the 48th or 49th over.
If we go by what we hear on air, though, commentators don’t like such games. For them, a “good pitch” is one that produces lots of runs. And forgetting that bowling is also cricket, analysts mistake the number of runs scored with the quality of a game.
Now, there is nothing inherently bad about high-scoring games. In T20s, we are used to teams scoring at eight or nine runs per over, and those games are definitely not uninteresting. The real problem with high-scoring games is that the impact of a wicket in such games can be large, and this can result in largely one-sided games.
So if you were a curator in charge of preparing a pitch for a ODI, and you had perfect control on a narrow run range that the team batting first would score, what score should you pick? Do you bow to the commentators and go for a 300 wicket or hark back to the glory days of the 90s and make a 225 pitch? At what run scoring range does the game have the largest chance to be interesting?
For starters, if we look at all ODIs ever played, we find that when the team batting first scores between 225 and 250, the chance of winning is very close to 50%.
As we saw in the first edition of this newsletter, ODI scores have been increasingly steadily over time, with the “second big bull run” taking place between 2010 and 2015. So we might think that the score range that results in 50% victory should also have gone up. However, we find that even in the period after the 2015 World Cup, a score of 250-274 has over a 50% chance of victory. While scores have been going up on average, the scores needed to win a game haven’t really been going up by that much (notwithstanding the recent England-Pakistan series).
Another way to look at whether a game is interesting is to look at the ultimate margin of victory. And we find again that the 225-250 range produces the closest games. For games where the chasing team wins, we extrapolate the runs scored (linearly) to see how many runs the team chasing would’ve scored in 300 balls, and then take the difference in scores as the margin (shortened games have been left out of the analysis).
Historically we find that scores between 200 and 275 produce the closest games. However, if we look at the period after the 2015 World Cup, we find that scores of 200-225 aren’t close any more, with the smallest margins of victory being in the 225-300 run range (interestingly, in this period, 325-350 games have also been competitive). Beyond 350, though, the game becomes one-sided.
Getting more technical
Writing for Mint in the wake of the ICC World T20 in 2016, I had come up with a measure of interestingness of a game. This measure is dependent on a graphic that I’ve developed to visualise limited overs games. At the essence of it, I calculate the likelihood of the team batting first winning after every ball (using an offshoot of the WASP algorithm), and then join up the dots to give the “story of the game”. I had made a short video explaining the graphic a few months ago.
So in order to measure the interestingness of the game, we look at the change in probabilities after each ball, square them and sum them across the period of the game (the Mint article gives a full explanation of why this is a good measure of interestingness). And for each game, we get a single number that indicates how interesting it is (the higher the better - since it indicates more “swings” in the game).
The result is the same - looking at all-time data, games where the team batting first scored between 200 and 250 are the most interesting. While games in this range remain interesting even after the 2015 World Cup, we find that games in the 250-300 range are on average more interesting, with interestingness sharply dropping off after 300.
So what does this mean if you are a curator for the forthcoming World Cup? Based on all this data, preparing a pitch that produces between 250 and 275 runs on average should ensure you an interesting game by all measures. Below 225 or over 300 will result in a sharp drop in interestingness by one measure or another, and that should be avoided.
Let’s see how the curators actually do!