This is horsesh*t!

[thevortex]

Duckworth puts up a fairly good defence of the system here

Hmm..I would not call that a 'good' defense. He simply defends his system saying that he has tested the system out in 80 odd T20 games. That is just the same as how Richardson trots out those numbers about umpires and their efficiencies vis a vis Hawkeye. :)

[SJS50]

Hmm..I would not call that a 'good' defense. He simply defends his system saying that he has tested the system out in 80 odd T20 games. That is just the same as how Richardson trots out those numbers about umpires and their efficiencies vis a vis Hawkeye. :)

Well his point is that the D?L system comes in after rain has interrupted play. In the case of England this happened after 2.2 overs had been bowled by them to the West Indies. With West Indies at 31 in 2.2 overs, their position looks much stronger than it would have had it been 15 for no loss in 2.2 overs. What West Indies had to score after that in almost four overs, was just about 8 runs an over but that was because they had scored at 14 runs per over already. I did not know that since I did not watch the West Indies bat at all. I thought this was the target given just as the Windies innings started. That looked pretty bad if it had been the case.

[The Outsider]

Duckworth puts up a fairly good defence of the system here

It's a pretty lame defense in fact. Particularly, the statement :

"No I don't, quite frankly," he told Cricinfo's Switch Hit podcast. "Remember that there have been a total of about 70 matches decided by Duckworth-Lewis since Twenty20 was invented in 2002, and there's only been two instances where any dissent has been expressed, and both of those were by Paul Collingwood and the England team, as a result of failing to win against West Indies.

The Duckworth-Lewis system, despite it's appearance of being some esoteric black box is nothing but a statistical modeling algorithm and like all statistical modeling algorithms it will have a range of applicability. It is not a universal law of Physics or a theorem of Math which has general applicability. It was designed for giving reasonable targets in the event of the curtailment of a 50 over match into a minimum of 20 overs, and it does work quite well there. It is now being used in situations pertaining the curtailment of a 20 over match into a minimum of 5 overs. There is absolutely no reason to expect that the results would spill over. To use the success of the system in ODIs and simply transfer it to T20s is nuts. When Duckworth says "70 matches" one has to assume he is talking about majority ODIs, for there have not been anywhere close to 70 T20s decided by the system. Contrary to what Duckworth is saying, every single time there is a curtailment of a T20 due to rain and a revised target is set up, something seems missing. One such instance I had followed and pointed out was : http://www.indiancricketfans.com/showthread.php?t=104594 The biggest problem with the system which came out in the England-West Indies match and also the England-Ireland match was very low targets while chasing. And this is not something which Duckworth can simply say, "Oh it's fair". In the event of possibility of rain, every captain wants to field and every follower of the game knows that the targets going to be set will be ridiculously low. The game is being played on the field and everyone on the field knows where the advantage lies. This cannot be said of the system in ODIs. To me, without getting into the gory details, is proof enough that something is wrong with the system as it gives a clear arbitrage opportunity.

[flamy]

brah ... you haz been PWNED by Tontee-Tontee :hehe:

I want moar IPL :woot:

[thevortex]

Well his point is that the D?L system comes in after rain has interrupted play. In the case of England this happened after 2.2 overs had been bowled by them to the West Indies. With West Indies at 31 in 2.2 overs, their position looks much stronger than it would have had it been 15 for no loss in 2.2 overs. What West Indies had to score after that in almost four overs, was just about 8 runs an over but that was because they had scored at 14 runs per over already. I did not know that since I did not watch the West Indies bat at all. I thought this was the target given just as the Windies innings started. That looked pretty bad if it had been the case.

:) Well, SJS, his system is supposed to come in after rain has stopped play, na? So no prizes for guessing that for Mr. Duckworth. What is his argument against claims and calls for the system to be tweaked? Just a claim that he is sure that his system is fine and that he tunes it every 5 years or so based on match results? Is that not lame? In that particular match do you not think that the team batting first is absolutely and completely screwed (pardon the french)? This is what is Collingwood's point. In any case, feel free to disregard my views as I absolutely abhor these mathematical systems which profess to have the gall to actually predict which team would win based on some probability theories :). Over and above abhorring I have zero respect for these systems. The game is far more unpredictable than these gentlemen give credit to. These systems are the ultimate nod to TV broadcast-driven sport today. People want a result no matter what. Even if there is none. So Messrs Duckworth and Lewis manufacture a result. Plain and simple.

[The Outsider]

Nice article on the limitations of applying a model developed for a 50 over match to a 20 over match on the lines of post #43, but explained in much more detail : http://www.cricinfo.com/magazine/content/story/459431.html