Talk:Duckworth-Lewis method

From Wikipedia, the free encyclopedia

cricket ball
Cricket Portal
Click here for information about how the WikiProject assesses notability
Duckworth-Lewis method is part of WikiProject Cricket which aims to expand and organise information better in articles related to the sport of cricket. Please participate by visiting the project page for more details.
??? This article has not yet received a rating on the quality scale.
??? This article has not yet received a rating on the importance scale.

Contents

[edit] "Criticism" in the opening paragraph

Unless there is some specific criticism of the _current_ D/L method, better yet with a link to a public statement of such, occasional cricket-fan "uncomfortableness" with D/L is covered in detail in the body of the article. Note that "generally accepted" includes "there are a couple of commentators who have raised largely insubstantial and uninformed criticisms" --Shannonr 22:40, 30 November 2005 (UTC)

[edit] Tony Lewis

The article states clearly that Tony Lewis is a statistician, and the link points to Tony Lewis (mathematician). Thus I do not think it is worthwhile cluttering an already dense first paragraph with the information that this isn't an English cricketer from the 70s. Shannonr 01:32, 8 September 2005 (UTC)

Not sure I'd agree with this. i'd always assumed that The Lewis was tony the cricketer and commentator- it required a tour throught the Tony Lewis disambiguation pages to disavow me of this notion. Epeeist smudge 12:52, 12 December 2006 (UTC)
Yes, that's what disambiguation pages are for. Mentioning in this article -- which is not about Tony Lewis, but about the D/L Method -- that the co-creator of this method is *not* someone he might be mistaken for by English cricket tragics of a certain age (and I use that term with the greatest of respect, I am myself a cricket tragic) really seems, to me, a waste of space. Let's stick to explaining the Method itself. Shannonr 12:44, 27 December 2006 (UTC)
Everyone always assumed he was The Tony Lewis. -- Unsigned comment
See above. The disambiguation page covers this. --Shannonr 07:09, 9 August 2007 (UTC)

[edit] Search for non-copyright scoreboard image

IMO, what would really be cool here is if someone could post a shot of a scoreboard, somewhere, with the target score displayed... Anyone? --Shannonr 02:48, 26 April 2006 (UTC)

Image:Scoreboard showing DL target.jpg This picture is quite blurry, but it does show the Duckworth/Lewis target score (even though people are standing in front of 'D/Lewis'). Will it suffice? Deaþe gecweald 12:41, 5 November 2006 (UTC)

Nice photo! But you're right -- people are standing right in front of the area we're interested in for this article! Shannonr 15:38, 6 November 2006 (UTC)

[edit] Entry heading

I'm a big cricket fan and I've almost exclusively seen this term written Duckworth/Lewis, not Duckworth-Lewis. The abbreviation is D/L not D-L, and if the hyphen were to be used surely it should be the middle hyphen –, not the Unicode -. I propose the title be changed accordingly (there is already a redirect from Duckworth/Lewis method), with this entry being redirected to it. Kris 20:40, 11 February 2007 (UTC)

I agree — The preceding unsigned comment was added by 89.168.18.255 (talk • contribs) .
Disagree. The ICC, Cricinfo, and Wisden use both Duckworth/Lewis and Duckworth-Lewis. Major media also seems to use both interchangeably. As "slashes" are sometimes problematic, I feel that the redirect here from Duckworth/Lewis should suffice. Shannonr 13:22, 8 April 2007 (UTC)

[edit] Australia-UAE game

I'm removing the bit about the Australia - UAE game added by 144.131.7.245 because: a) there is no record of this game in the Cricinfo match archive and b) Most (all?) of the 1992 season was pre-D/L and I assume from the example given (if that game did indeed take place) that the match actually represents a good example of the problems with pre-D/L methods of target calculation. I'm also removing the 'criticism' which mentions radio commentators as there are no sources.—The preceding unsigned comment was added by 89.168.18.255 (talk) 21:20, 11 February 2007.

[edit] World Cup 2007 Pakistan/Ireland match

There's an interesting question over at Talk:2007 Cricket World Cup. Pakistan scored 132. Ireland were set 128 under the D-L. They scored a six to win when on 127, so ended up on 133. The 2007 Cricket World Cup article currently states that they won under the D-L method. But they won regardless right? So is it appropriate to say that they won, or do we have to qualify it with "under the D-L method"? Hesperian 05:44, 19 March 2007 (UTC)

I think you'd still say they won under D/L, because effectively the target of 132 was irrelevant and shouldn't be considered. It certainly wasn't a consideration to Ireland when they were batting. They were trying to obtain a target of 128 and they got it, end of story. Good point, though. Kris 08:19, 19 March 2007 (UTC)

[edit] Statistical questions on D/L

The method's use of statistics raises certain questions....

1. Does historical records/statistics used by D/L include the matches that were decided by previous net run rate or other methods? If yes, they do not accurately represent the proper use of resources (overs or wickets). That would create incorrect statistics.

2. How are the matches decided by D/L method iteself represented in future D/L statistics? Will it not skew the statistics in their own favour (because tragets are updated using D/L method now and have the capability of changing the "normal use of resource" behaviour in terms of overs remaining).

Ajhau 03:41, 22 March 2007 (UTC)Ajit

They are not included. Infact looking at the DL statistics they should be able to refine it. People are already pointing out that achieving +8 or above run rate becomes unrealistic as it would most likly require a boundry which isn't always obtainable depending on pitch conditions. For example if a team scores 330 in 50 overs, and the next side only has 20 overs, the target will be set at about 200 runs, requiring a run rate of 10 per over. Yet the results show that much higher percentage of teams can make 330 in 50 overs, and very little percentage of teams has ever successfull made 200 runs in 20 overs. —The preceding unsigned comment was added by 155.144.251.120 (talk) 04:32, 28 March 2007 (UTC).
Previous D/L matches are not included, and this is by design. A statistical method to determine what would have happen had the game gone on "normally" is what the D/L promises -- adding results determined by it would make it less fair. Also, in response to the "back of an envelope" example -- that is exactly the sort of calculation that the D/L does extremely well. The chasing team would not be set 200 in such an example. Firstly, let's agree that any team that makes 330 has done a great job batting, and the target set for the chasing team's 20 overs had better be high. Second, let's agree that the second team _knows_ they've only got 20 overs. With all ten wickets in hand, D/L would actually set the winning target at 194.37, or 195 to win. I cannot see how this is not completely fair, and more to the point, better than any previous method. --Shannonr 03:41, 7 April 2007 (UTC)

Can someone give the formula used by the D/L method in calculating the score required by the team batting 2nd in a one-day game? I mean, how do people go about calculating the score required? I've heard it's an extremely complicated method. So can someone please give the formula or the steps involved in calculating the score?

You can find a simplified table online by searching for "duckworth lewis tables" on your search engine of choice. The full Simplified Table can be purchased, as can the full Professional version software package. Links in the article will take you there. --Shannonr 03:41, 7 April 2007 (UTC)

[edit] ICC World Cup 2007 final

Anyone want to add anything about the Australia/Sri Lanka final? Especially as it was fairly controversial.. I'm not that expert at cricket so I didnt do it myself Feudonym 00:52, 30 April 2007 (UTC)

Not sure that the umpire confusion in that game helps to explain Duckworth-Lewis! And the explanation given on the Cricket World Cup 2007 page is probably enough. Least said, soonest mended, and all that. Although, if someone were to find a clever (and brief!) way to use that match as an example of "D/L gone wrong" I wouldn't object. Shannonr 07:37, 30 July 2007 (UTC) Update: I've made a comment about this in the "Criticism" section, as on further research I think the controversy Feudonym talks about is real and widespread. Shannonr 07:15, 31 July 2007 (UTC)

[edit] Major Re-org

I've made a major reorganization of the page, attempting to achieve three things:

  • The use of the given example to cement the concept of "if the game had continued" removing the need for further explanation of this in the Theory section
  • A more natural "flow" -- the "upside down triangle" of information importance
  • The removal of unreferenced criticisms

I've also taken the opportunity to remove a couple of redundant qualifications, and removed (yet again!) the horrible "analogy" that keeps creeping back into the Theory section.

Please feel free to comment/criticize! Shannonr 07:15, 31 July 2007 (UTC)

These changes are barely literate. And do not improve the article. Revert. Stop the vandalism and stop putting spam back!62.64.201.211 13:48, 1 August 2007 (UTC)
Your comments are odd to say the least, as much of the version you claim to prefer was written by me! There is no "vandalism" -- except your continual reverting to a version with spelling errors, grammar errors, and unreferenced statments -- just a sincere desire to improve the readability of this article. Thanks for your comments, but I feel the changes I've made are quite supportable. Lastly, a link to the publishers of the Duckworth/Lewis Table is hardly "spam". Shannonr 13:53, 1 August 2007 (UTC)
"statments" - is that how much care you take in your editing or is it a spelling mistake? If you limited yourself to "spelling errors, grammar errors, and unreferenced statments(sic)" that would be fine, but the article now has less readability and less structural logic. But let's leave it - rotten and disfigured as it is. I distain to involve myself in an edit war 62.64.208.148 23:26, 2 August 2007 (UTC)
One thing you said previously made me look again at Acumen Books. They no longer appear to have an exclusive on the D/L guide, so I'm removing the link. As an aside, you may want to check the spelling of "disdain". Cheers. Shannonr 08:06, 6 August 2007 (UTC)
You don't seem prepared to receive comment or criticism. You seem to have decided to take over the article and have yet to improve it. The way you write is dismal. 62.64.207.121 09:13, 7 August 2007 (UTC)
Another odd comment, as I think two good changes came about because I was prepared to wade through your personal attacks to find ideas for improving the article. Oh, and please do not add comments and try to pretend they are from other users -- it's obvious from the history who made the change. I still don't understand your hostility, and I certainly haven't "taken over" anything, but that's also obvious from the history. Many people have worked together to make this article. --Shannonr 16:41, 8 August 2007 (UTC)

[edit] Another D-L criticism?

D-L does not take into account the strength of the batsmen who are not yet out. For example, if a team's final three batsmen happen to be very capable, a team losing a rained-out match according to the D-L might have more than likely won the match. Not sure whether this is criticism or an oddity, because perhaps a team's winning or losing a game should not be judged on the strength of batsmen to follow, but rather the performance of those who have batted in the match. However, one thing is for certain: as a predictor of who WOULD HAVE won the match had it not rained, the D-L method can be vastly improved by looking at the strength of batsmen not yet out. Perhaps it would be useful to note that the D-L method cannot be used to predict who would have won a match had it not rained out, but rather who performed "better" during the play that actually took place. I have never heard anyone make this distinction, but i think it is a very important distinction to make, and one that might even affect the future development of D-L. -- Barryvz (talk) 01:39, 6 January 2008 (UTC)

You make an interesting point, and I thank you for raising it here. You're right that the D/L method doesn't specifically try to assess the "strength" of the batsmen remaining, but it does do it indirectly through the averages of all team scores over ODI history which are at its heart. There is, IMO, at least one strong reason not to try and factor "strengths" in: it would be exploitable. As soon as the second team came out to bat, they would send in numbers 3 and 4, and "save" their "openers" (ie. their strongest batsmen) to lower the target score (eg. "Our two strongest bats haven't batted yet, they'd easily get those runs if the match went on...") But even assuming you could change the rules to prevent such a strategy (or similar) D/L is not designed to operate as a handicapping system (which factoring in the "strengths" of the batting in the two sides would make it into). The target score merely represents what the all-time average team would have got (with whatever variables of wickets and balls remaining). In that way it's fair for an above average team (who merely have to play to their standard to hit a low target) and a below average team (who have to play above their standard to even hit a "normal" target). -- Shannonr (talk) 06:55, 12 January 2008 (UTC)
In heavily truncated innings, D/L does replace the score that a team with a great depth of batsmen would probably have achieved with a more 'average' score based on performances by teams of all strengths (as it tends to 'average up' the likely final score of teams with just a few good batsmen when the innings ends very early). (A truncated innings also means that a strong bowling team did not get the chance to show its depth of bowling.) Has anyone done an analysis of % wins between teams when D/L is used against % wins when matches run their course? It might be a measure of how much D/L makes the performances of teams more 'average' (downwards and upwards). Are the averages different for test sides playing ODI and county sides? Arguably the figures used at each level of the game should be taken from matches at the same level of the game so that the D/L averages are as close as possible to the averages of the teams playing. The method is likely to be more unfair where the difference between the strength of the teams and/or the 'source' of the averages is greatest. In the end there is no perfect system that can replace a match played to the full number of overs, it is about balancing historic results (at whatever level of specificity) with what happened (as far as it happened) on the day - as well as taking into account ease of use of the system. There are lots of minor unfairnesses in sport, weather, injury, luck. The best teams show themselves as such over a period of time, when minor unfairnesses iron themselves out. Jagdfeld (talk) 13:33, 13 January 2008 (UTC)