Cricket interruptus: fairness and incentive in interrupted cricket matches∗

We present a new adjustment rule for interrupted cricket matches that equalizes the probability of winning before and after the interruption. Our proposal differs from existing rules in the quantity preserved (the probability of winning), and also in the point at which it is measured (the time of interruption). We claim this is both fair and free of incentive effects. We give several examples of how our rule could have been applied in past matches, including some in which the ultimate result might have been different. The game of cricket is an immensely popular participatory and spectator sport in Britain and its former colonies in Africa (South Africa and Zimbabwe), Australasia (Australia and New Zealand), the Indian subcontinent (Bangladesh, India, Pakistan, Sri Lanka), and the West Indies. Traditionally, an international cricket match (“test match”) is played over five days and, more often than not, it ends in a draw. It is a game for afficionados. Recognising the potential popular appeal of a shorter game that provided a decisive result, the Australian media magnate Kerry Packer pioneered the introduction of a one-day game in the 1970s. Since then, the game has flourished commercially as a television sport, becoming the second most popular (after football) spectator sport. International teams travel the world, playing a mixture of one-day games and the traditional test matches. Annual one-day tournaments are held in Australia and Sharjah. Every four years, all major cricketing nations compete in the World Cup to produce an international champion. A one-day game typically lasts about seven hours, during which the two opposing teams each complete a single innings with a limit of 50 overs. Even over this short period, it is not unusual for a game to be interrupted by rain, especially in temperate countries such as England, Australia and New Zealand.1 Typically, the duration of the game must be reduced. Since the ∗We gratefully acknowledge the helpful comments of Sanghamitra Das and Paul Walker, and the research assistance of Gareth Stiven and Steen Videbeck. †Corresponding author: School of Economics and Finance, Victoria University of Wellington, Wellington, New Zealand. Email: [email protected] Phone: 64-4-4635763 Fax: 64-4-4955014 In the 1999 World Cup in England, eight of the 42 matches were affected by rain.