It has taken 32 years for South Africa to vanquish the demons of a rain-affected semi-final defeat to England and finally secure a spot in the decider at a men’s World Cup. Six more doses of semi-final suffering followed that initial heartbreak at the 1992 ODI tournament , when heavy showers and a quickly discredited rain rule ended South Africa’s hopes of a triumphant return to the global stage.
Chasing England’s target of 253, South Africa needed 22 runs from 13 balls when rain interrupted play at the SCG. After the skies cleared and play was set to resume, England’s two lowest-scoring overs were wiped from their total and the recalculations reduced South Africa’s target to 21… from one delivery.
Rain might well come into play again when South Africa face India in the 2024 T20 World Cup final at Kensington Oval in Bridgetown, with wet weather forecast across Barbados on Saturday. But even if the match is reduced from 20 overs each, the impact on the result should be a far cry from the outrage 32-years ago when South Africa’s World Cup hoodoo began.
The DLS method now used in limited-over cricket matches is the legacy of English statisticians Frank Duckworth, who died on 21 June aged 84, and Tony Lewis, who passed away in 2020 aged 78. It might not be a simple system for the ordinary fan – and seemingly most commentators – to comprehend, but the current custodian Steven Stern is adamant that the method is the most statistically fair.
“Frank’s perspective on this as a cricket fan is always interesting to me, because I’m a mathematician by trade, a data analyst,” Stern, a professor of data science at Bond University, tells Guardian Australia. “He was quite quick in keeping me grounded, saying, ‘Simple is better unless we need to go complicated, as long as a simple approach gives us something that’s justifiable’. He wanted the method to be understandable to your rank and file fan.”
The original ‘DL’ method created by Duckworth and Lewis was born out of the 1992 debacle – which the former dubbed “a fair result in foul weather” – and officially adopted by the International Cricket Council in 1999. It remains the basis for the system that rain-affected limited-over matches are now decided under, although there have been several tweaks over the years.
Stern put the S into ‘DLS’ when his work was incorporated into a significant update released in 2014 and, following the retirement of Duckworth and Lewis the same year, he has held the reins ever since. The Australian statistician is currently working on an annual review, but does not expect the first significant change in a decade to be needed this time around.
“I do a reanalysis every year using a four-year window,” Stern says. “So every July 1st, I use the previous four years from that point, so there’s a new analysis but with three years of overlap with the previous analysis.
after newsletter promotion
“I look at all of the DLS matches that actually occurred in the past four years and say, ‘With this new method that I’ve just developed with the current analysis, what targets would it have set?’. If those targets only changed by one or two runs, well that’s within the run of play, we’re not going to make a change.
“But as soon as we start seeing it’s out by three or four runs, maybe the patterns have changed enough. Then we have a discussion with the ICC about whether that is big enough for us to make a change.”
The last overhaul in 2014 was sparked by “the T20 revolution”. “Batters realised in a very short span of time what was really possible in those last 10 overs, or that you can score 80 runs in the last five if you really went for it,” Stern adds. “That caused a fundamental shift in the way that the method worked, when the scores got really, really big.”
Stern is now paying closer attention to the “relative impact of partnerships”, as the leading sides are more likely to save their big-hitting weapons for further down the order, and the depth at emerging nations continues to grow. But while Stern and the ICC are open to having separate DLS methods for men’s and women’s matches, if the data calls for it, there is no evidence of that being needed at this stage.
“The method still works perfectly for women’s cricket despite the fact that they score fewer runs,” Stern says. “The reason is that the method only works on percentages. So even though women score on average about 50 runs fewer, when they score those runs during the innings, the percentage that’s at the end, versus the percentage in the middle and at the beginning, is almost identical to the men.”
The method became a key part of the drama in 2024, amid the chaotic finale to the T20 World Cup group stage when Afghanistan all-rounder Gulbadin Naib feigned injury as DLS calculations were being made, rather than the method directly influencing the result. This is much as Duckworth could surely have hoped for.
“Frank Duckworth is, at least from a professional perspective, the most generous human being I have ever known in my life,” Stern says. “Whenever I do the yearly calculations, I will think about him. And fortunately, his name will be remembered for as long as we play cricket.”
