A challenger to the Duckworth-Lewis method?

The Duckworth-Lewis system for calculating target scores in rain-interrupted one-day internationals has come in for plenty of debate, but it is clearly the best of all methods that have been used so far. The question is, is there a better system which can challenge Duckworth and Lewis?

V Jayadevan, an engineer from Kerala, has devised a method which has the backing of the Indian board and will be discussed by the ICC's Cricket Committee during their two-day meet in Dubai on May 13 and 14. Like the D-L method, Jayadevan's system (the VJD method) also comes from an analysis of numerous one-day matches, and predicting scores and targets on the basis of scoring patterns recorded from earlier games. The table below lists the targets calculated by both methods for some real and hypothetical situations.

The D-L system currently in use is their new Professional system, an improvement over the earlier D-L method, and one which requires a computer to work out the calculations. (In the earlier D-L process, manual calculations were possible.) The new method is a definite improvement over the earlier one - in the 2003 World Cup final, India, requiring 360 for victory, would have clinched the trophy with a total of 158 for 3 after 25 overs under the earlier system; with the new version, the target would have been a more realistic 176, a better output than Jayadevan's 167.

However, a couple of flaws remain even in the improved D-L rule. The method tends to put far too much emphasis on keeping wickets in hand. As example 1 in the table shows, a score of 115 without loss is enough to win a match when chasing 301, though the required rate of 7.44 is a huge ask over a period of 25 overs, even with all wickets in tact. Jayadevan's target of 121 is only slightly better.

Second, in matches where the teams batting first have their innings curtailed, the D-L system tends to prop their totals up far too much (see eg. 2 and 3). That's partly because the method doesn't take into account the fact that in a truncated game, the team chasing has fewer overs to bat with the field restrictions. In fact, the D-L method doesn't factor in the field restrictions in ODIs that often leads to a deluge of runs in the early stages.

Moreover, the D-L system, on occasions, tends to favour the team which is batting when the interruption occurs. Sample the World Cup match between South Africa and New Zealand: chasing 307, New Zealand were 182 for 1 in 30.3 overs when rain intervened and reduced the New Zealand innings to 39 overs. When the interruption occurred, the requirement was 124 from 19.3 overs - a rate of 6.4 an over. According to the D-L method then in use, the target came down to just 226 from 39 - that's a mere 44 more from 8.3 overs, at just five an over. The new D-L version puts the target at 230, still well below the required rate before the rain delay, but Jayadevan's target is 240 - that's 58 more from 8.3 overs, an asking rate of nearly seven, which seems fair considering the reduction in overs.

It needs to be pointed out, though, that in most cases the results of both methods are fairly similar. The earlier version of D-L had some glaring flaws, but the newer version has eliminated most of them. However, Jayadevan's method is one that merits closer scrutiny. The ICC may still want to continue with the D-L method for the near future, but it might be a good idea for them to monitor Jayadevan's system and study the results it delivers in different match scenarios over a period of time.