How Two Statisticians Fixed Rain in Cricket

On 22 March 1992, South Africa needed 22 runs off 13 balls to beat England in the World Cup semi-final at Sydney. Then it rained. Under the run-rate rules in force, two overs were lost — and the target was simply recalculated to 22 runs off 1 ball. The crowd booed. South Africa lost. Cricket's interim rain rule had embarrassed itself on a global stage.

Frank Duckworth, a statistician with the British nuclear-power industry, and Tony Lewis, a mathematics lecturer at the University of the West of England, set out to fix it. Their core insight was that a chasing side has two scarce resources, not one: balls remaining and wickets in hand. Lose either and your scoring potential drops. Lose both and it collapses.

They built a table — published in 1998, adopted by the ICC the next year — of the percentage of resources a team has at any point in an innings, given the overs and wickets remaining. Cut a chase short by rain, and you compare the resources the chasing side actually had against the resources the first innings used. The revised target falls out of the ratio.

The method survived its first big embarrassment when high-scoring Twenty20 matches showed the original tables under-rewarding teams that started fast. Australian academic Steven Stern revised the resource table in 2014, and the rule picked up its third surname.

Nobody loves a math formula in the middle of a sport. The formula has the saving grace of being honest about what every cricket fan already knew: a wicket in hand is worth real runs, and rain is not a coin flip.