## Saturday, September 19, 2009

### Duckworth-Lewis method – how statistics and cricket combine

Duckworth-Lewis method – how statistics and cricket combine

The Duckworth-Lewis method (D/L method) is a mathematical way to calculate the target score for the team batting second in any form of limited overs cricket match. It is usually used when the match has been interrupted by weather. It is generally accepted to be a fair and accurate method of setting a target score or at least it is the most fair and accurate method to have been developed thus far. However as it is a predictive statistical method and as such tries to create what would have happened had the game continued uninterrupted it obviously will create controversy for supporters. The D/L method was devised by two English statisticians, Frank Duckworth and Tony Lewis.

The D/L method is relatively simple to apply, at least when compared to several other statistical methods but requires a published reference table and mathematical calculations. As with most statistical derivations, however, the D/L method can produce results that may seem to call into question the accuracy of the derived answer as opposed to the perception of what the answer should be. As such the announcement of the derived target score can provoke a good deal of second-guessing and discussion amongst the crowd at the cricket ground.

The published table that the D/L method uses for its derivations is regularly updated, as in recent years one-day teams were piling up significantly higher scores than in previous years, as such the continuity of the chart would be affected.

After the embarrassing loss in the first ODI between West Indies and England almost all of the blame has fallen squarely upon the coach John Dyson, who to be fair has accepted the criticism after a truly shocking calculation error that left the West Indies short of 1 run to tie the match. Even more perplexing was the fact that everyone seemed to know that accepting the light at that time would be, in fact, accepting a loss. All the forms of media , including the television, Cricinfo live match commentary and radio coverage had continual statistical updates of the required total needed on the D/L method.

Surely it at least one form of the above media would have been available to the West Indian dressing room and even if it were not, the official match statistician would have had the requirements , although I do not know if he is allowed to communicate with the players. Even if he is not, the website; http://www.duckworth-lewis.com has a calculator available on it and can be used with relative ease due to its layout.

I have never John Dyson make a claim that he is a statistician or mathematician. The question that would then arise is why was he then left with the responsibility of making those calculations? Should the coach even be reading complicated D/L tables, it certainly doesn’t seem as a necessary coaching skill at any level of the game.

So West Indies have at least not equaled the low point that South Africa set when sending themselves out of the 2002-3 World Cup, which they were hosting, when they too misread the Duckworth-Lewis rules. In that match they thought they didn’t need an extra run to take the win which would have kept them alive in the competition and did not attempt to score off of the last ball. This is unlikely to be of any comfort to the fans , especially as the game was poised for an exciting finish with 27 runs needed off of 22 balls with only 3 West Indian wickets in hand.