What's new

Should cricket get rid of this farcical Duckworth-Lewis method?

Canford Cliffs

Canford Cliffs

Local Club Star
Joined
Jul 22, 2018
Runs
1,803
So DLS is a mathematical formula designed by 2 british statistians but I am pretty sure most pundits don't understand how it works. This method has screwed many teams over the years with the most infamous screwjob being against SA in the 1992 WC.

Even if we analyze today's T20 game, Australia scored at a run rate of 6.95 for 19 overs and then it rained. How come the req run rate assigned to chasing team is 11.5 per over? What kind of calculation is this?

We have plenty of intelligent modern mathematicians. Its time to put this colonial age old DL method in the bin and adopt a new one.
 
How can ICC have such rules, which may cause India to lose?

We should boycott WC'19. How dare ICC make India lose?

We are kings and, we should define rules, not ICC.

BCCI hai toh ICC hai, BCCI nahi toh kuch bhi nahi.
 
Canford Cliffs said:
So DLS is a mathematical formula designed by 2 british statistians but I am pretty sure most pundits don't understand how it works. This method has screwed many teams over the years with the most infamous screwjob being against SA in the 1992 WC.

Even if we analyze today's T20 game, Australia scored at a run rate of 6.95 for 19 overs and then it rained. How come the req run rate assigned to chasing team is 11.5 per over? What kind of calculation is this?

We have plenty of intelligent modern mathematicians. Its time to put this colonial age old DL method in the bin and adopt a new one.
Click to expand...

Because if it was a 5 over game at the start it can be reasonably assumed Australia's RR would not have been 6.95 as they would have batted at a much higher rate as they would have known its only 5 overs, not 20 where they could have taken more time to accelerate. DRS takes that into account.
 
Kohli said:
How can ICC have such rules, which may cause India to lose?

We should boycott WC'19. How dare ICC make India lose?

We are kings and, we should define rules, not ICC.

BCCI hai toh ICC hai, BCCI nahi toh kuch bhi nahi.
Click to expand...

You can try these sarcastic dramas with some other posters....not me bruv. Lets stick on the topic.

DLS ruling is same for all team, so its not only India or BCCI. Explain me how 6.95 run rate in 1st innings equate to req rate of 11.5 in 2nd? What kind of calculation is that?

There is no question this farcical rule needs to be changed with a better one.
 
As soon as India starts playing abroad, excuses threads from [MENTION=147347]Canford Cliffs[/MENTION] starts to pop up. :inti
 
Canford Cliffs said:
You can try these sarcastic dramas with some other posters....not me bruv. Lets stick on the topic.

DLS ruling is same for all team, so its not only India or BCCI. Explain me how 6.95 run rate in 1st innings equate to req rate of 11.5 in 2nd? What kind of calculation is that?

There is no question this farcical rule needs to be changed with a better one.
Click to expand...

Like what. No point moaning without suggesting credible alternatives.
 
GLORY OF '92 said:
Like what. No point moaning without suggesting credible alternatives.
Click to expand...

So are you saying DLS is all fair and square? I am no mathematician...so cant suggest an alternatives. But surely there are many IT guys who can come up with better formula.

Remember that SA game in 92 WC? SA were on course for a victory with 22 runs needed in 13 balls. It rained then for 12 mins and the equation was 22 runs in 1 ball :)))

I am amazed in this era, this rule still exists. Its one of the many things that is wrong with cricket compared to other sports.
 
It's absolutely fine. Complex to understand but its a model which impacts all teams same way. And is a much better model than anything else we have had for rain effected matches
 
Canford Cliffs said:
So are you saying DLS is all fair and square? I am no mathematician...so cant suggest an alternatives. But surely there are many IT guys who can come up with better formula.

Remember that SA game in 92 WC? SA were on course for a victory with 22 runs needed in 13 balls. It rained then for 12 mins and the equation was 22 runs in 1 ball :)))

I am amazed in this era, this rule still exists. Its one of the many things that is wrong with cricket compared to other sports.
Click to expand...
I am pretty sure that was the previous method and post 92 world cup DL was introduced someone correct me if I am wrong.
 
DL might be farce but it's the best alternative right now

If someone can come up with a better one we can see
 
DL/S is fine.

The only place it falls back a bit is in T20s.

Because DL/S is based on historical data and T20 team scores fluctuate so much, that the margin of error is more while using DLS in T20s as compared to ODIs.

But that's something we have to live with. We cannot predict the future. So the best we can do is something like DL/S.

Nothing as such wrong with how DL/S works.
 
Let’s remove DWL
Let’s Remove DRS
Let’s Remove away tours for India
 
Canford Cliffs said:
So DLS is a mathematical formula designed by 2 british statistians but I am pretty sure most pundits don't understand how it works. This method has screwed many teams over the years with the most infamous screwjob being against SA in the 1992 WC.

Even if we analyze today's T20 game, Australia scored at a run rate of 6.95 for 19 overs and then it rained. How come the req run rate assigned to chasing team is 11.5 per over? What kind of calculation is this?

We have plenty of intelligent modern mathematicians. Its time to put this colonial age old DL method in the bin and adopt a new one.
Click to expand...

You are incorrect in saying the DLS method screwed South Africa over in 1992. It was the old "rain rule". The DLS was derived after that fiasco as an alternative.
 
DL is fine. I am pretty sure the RRR wouldn't be 11.5 if it was a 5-over game in today's match. And also certain that the 1992 SA match wasn't based on DL. It was only introduced after that farce. So don't make these kinds of posts without getting facts right.
 
Salma_T said:
I am pretty sure that was the previous method and post 92 world cup DL was introduced someone correct me if I am wrong.
Click to expand...

spot on, it was actually this very game which led to the introduction of DLS as the system or method used before that was completely in favour of the fielding team.
 
Last edited:
OP - this is an embarrassing thread.

First your facts about SA game in WC'19 are plain wrong.

Secondly, you know that any team with 10 wickets in hand in 5 overs would bat differently.

I remember all kinds of issues in 80's, 90's - because of all other systems that were tried. Duckworth Lewis has been a great success.

You need to understand how DLS works before making a thread.
 
The problem with the DLS method is that

1) It values all the batsmen at the same level. Its gives same value to both the opener and the tail ender.

2) In the 1st innings of an ODI, if some overs were cut short by rain and the batting team has many wickets left at that point, the target to chase increases dramatically. DLS method says if the batting side knew it was, say, a 40 over game, with the remaining wickets they would've scored that many runs. Unused sources, it says.

But at the same time it fails to consider that the bowling team knew it was a 40 over game, it would've used main bowlers inside the 40 overs and maybe the batting team wouldn't got that many runs or taken more wickets. DLS method doesn't use the same unused source principle with the bowlers.
 
par said:
The problem with the DLS method is that

1) It values all the batsmen at the same level. Its gives same value to both the opener and the tail ender.

2) In the 1st innings of an ODI, if some overs were cut short by rain and the batting team has many wickets left at that point, the target to chase increases dramatically. DLS method says if the batting side knew it was, say, a 40 over game, with the remaining wickets they would've scored that many runs. Unused sources, it says.

But at the same time it fails to consider that the bowling team knew it was a 40 over game, it would've used main bowlers inside the 40 overs and maybe the batting team wouldn't got that many runs or taken more wickets. DLS method doesn't use the same unused source principle with the bowlers.
Click to expand...

DLS assumes that your best batsmen are at the top of the order and vice versa. And that is how it is supposed to be
 
OP has skipped maths I think in high school .DLS takes into account a lot of factors which are abstract but matter eg momentum,advantage of knowing the target vs team which bats first at normal rate etc.U have to offset the losses the team batting first undergoes.
As a simple maths explanation probability of an event is very different than its conditional probability
 
the_outsider said:
DL/S is fine.

The only place it falls back a bit is in T20s.

Because DL/S is based on historical data and T20 team scores fluctuate so much, that the margin of error is more while using DLS in T20s as compared to ODIs.

But that's something we have to live with. We cannot predict the future. So the best we can do is something like DL/S.

Nothing as such wrong with how DL/S works.
Click to expand...

close thread. Factual answer that answers everything
 
DLS is fine. The best option we have right now. If you don't want to abandon the game
 
So cringeworthy to see how Indians want to twist tried and tested methods just because of a definitive loss against a C grade Aussie side.. pathetic
 
Good lord, this malady of dumbness is infectious. Poster claiming he's not a mathematician and then proceeding to rant nonsensically about he doesn't understand statistics so the rule must be wrong. Classic for the age of populism.
 
Shandarchowka said:
So cringeworthy to see how Indians want to twist tried and tested methods just because of a definitive loss against a C grade Aussie side.. pathetic
Click to expand...

hold on the horses. even indians saying it is ok except one or two. Don't label us all same.
 
Executioner said:
DLS assumes that your best batsmen are at the top of the order and vice versa. And that is how it is supposed to be
Click to expand...

This is what Duckworth and Lewis say

Shouldn’t the revised target take account of the quality of the players at the crease when play is stopped and of those who still have to bat? And should not account also be taken of the number of overs the top line bowlers will still have to bowl when play is resumed?

Although it is quite true that the extent to which the effective resources of the batting and bowling sides are depleted by a stoppage depends on the identities of the individual players affected, there is no way in which such factors could be incorporated into an objective rule for revising targets. It would require both teams to identify, before every match, the way the total quality of their sides, in respect of both batting and bowling, is divided between the individual team members. Furthermore, it would be necessary to input details of who was still to bowl and to bat and to perform the calculation based on this before a revised target could be computed. As well as leading to contention, such a procedure would be quite impractical to implement.
Click to expand...
 
just a hypothesis.

could an algorithm may be devised where initially weightage is put same for all 11 players (let's say 1.0) and then multiplying it by their last 10 performances (in terms of batting) putting a milestone for n number of runs. And that score could be used as a standard determining the value of the batsman?

As for T20 goes, the deviation curve in probability will be quite steep (both upwards and downwards). To the statistics experts, isn't there any way to make the curve less steep so that they remain close to the mean line?
 
Canford Cliffs said:
So DLS is a mathematical formula designed by 2 british statistians but I am pretty sure most pundits don't understand how it works. This method has screwed many teams over the years with the most infamous screwjob being against SA in the 1992 WC.

Even if we analyze today's T20 game, Australia scored at a run rate of 6.95 for 19 overs and then it rained. How come the req run rate assigned to chasing team is 11.5 per over? What kind of calculation is this?

We have plenty of intelligent modern mathematicians. Its time to put this colonial age old DL method in the bin and adopt a new one.
Click to expand...

The 1992 match was pre DLS.

As for the system itself, it may be th best method currently available. Rain days no longer exist, viewers (rightfully) demand results, so not sure what else can happen, unless every rain affected game is labeled a draw.

What is your suggestion?
 
Itachi said:
just a hypothesis.

could an algorithm may be devised where initially weightage is put same for all 11 players (let's say 1.0) and then multiplying it by their last 10 performances (in terms of batting) putting a milestone for n number of runs. And that score could be used as a standard determining the value of the batsman?

As for T20 goes, the deviation curve in probability will be quite steep (both upwards and downwards). To the statistics experts, isn't there any way to make the curve less steep so that they remain close to the mean line?
Click to expand...

That's an interesting way to look at it, but the volatility of the last 10 performances is likely to be very high which will in essence reduce the applicability. We'd probably need a bigger sample size to make sense of it, but again if it's just T20 internationals, then you'd have to be looking at data sets 3-4 years old.

A simplistic way to correct for the deviation curve is to look at second-order volatility (gamma), i.e. the volatility of volatility.

But I tend to agree with the authors of DLS, there seems to be very limited utility of making this adjustment. The system has been working for a while, I don't really see any compelling reason to change it. There was a debate three years ago, I think, led by an Indian statistician Jayadevan who proposed an alternative to DLS, but this was rejected because of some inconsistencies.

Some cricinfo articles for those who wish to delve deeper:

http://www.espncricinfo.com/magazine/content/story/881493.html

http://www.espncricinfo.com/magazine/content/story/890877.html
 
Canford Cliffs said:
So DLS is a mathematical formula designed by 2 british statistians but I am pretty sure most pundits don't understand how it works. This method has screwed many teams over the years with the most infamous screwjob being against SA in the 1992 WC.

Even if we analyze today's T20 game, Australia scored at a run rate of 6.95 for 19 overs and then it rained. How come the req run rate assigned to chasing team is 11.5 per over? What kind of calculation is this?

We have plenty of intelligent modern mathematicians. Its time to put this colonial age old DL method in the bin and adopt a new one.
Click to expand...

Otherwise you can chose rrr 7 but will get 3 wickets in hand. i,e 1/4th of total wickets for 1/4th of overs.
 
Opens a rant thread on DLS and his "best" example was a match where DLS didn't even exist but was proposed right after as a solution to the infamous farce he thinks was because of it lol

Irony is delicious
 
Canford Cliffs said:
So are you saying DLS is all fair and square? I am no mathematician...so cant suggest an alternatives. But surely there are many IT guys who can come up with better formula.

Remember that SA game in 92 WC? SA were on course for a victory with 22 runs needed in 13 balls. It rained then for 12 mins and the equation was 22 runs in 1 ball :)))

I am amazed in this era, this rule still exists. Its one of the many things that is wrong with cricket compared to other sports.
Click to expand...

You're actually completely wrong about the bold reference. Duckworth Lewis was not applied to that match and Duckworth Lewis came into existence after it.

Is Duckworth Lewis perfect? No.

Do I know of anything better than Duckworth Lewis? No.

So I will accept DLS on the basis that despite it's imperfections it is the best system available.

With regards to the point about the run rates. When there is a chance of rain most countries choose to bat second so they know what the target is if it rains. And most countries would rather chase 150 in 20 overs than chase 300 in 50 overs.

It's pointless to complain without actual suggestions. I think everyone knows about the limitations of the DLS system, you haven't made some sort of insane scientific breakthrough there.

On a separate note, the rules are the same for all countries. It's not as if DLS was derived to make it harder for India to win.
 
It is not DLS, but rain that seems to deny India whenever they look to be in a strong position to win a match.
 
Yes, I always hated DLS.

That WC semi final with ABD and Miller in, SA could have scored much more than what D/L set up NZ for a chase. It is non-sense rule really, crap all the way.

Should have been scrapped way back or some kind of modification definitely needs to be done. Need to re-calculate and introduce a better model perhaps
 
New Zealand have clinched their three-match T20 series against Bangladesh after beating visitors Bangladesh by 28 runs in their rain-affected second T20 international on Tuesday.

Sent in by Bangladesh, New Zealand reached 5-173 in the 18th over when their innings was cut short by rain.

Bangladesh were set a revised winning target of 170 from 16 overs under the Duckworth Lewis Stern system and finished 7-142 after losing their first match.

The series win was New Zealand's seventh at home this season; four of those were in T20s, two in Tests and one in ODIs.

Amid bizarre scenes, Bangladesh began their run chase before match referee Jeff Crowe officially calculated their revised total.

The umpires stopped the match with Bangladesh 0-12 after 1.3 overs, waiting for official notification of the new target.

There was a long delay while Crowe worked feverishly at his computer. At one point he had what appeared to be an angry exchange with Bangladesh coach Russell Domingo.

Bangladesh at times looked on pace to succeed in what would have been an extraordinary chase which required them to score mostly at 12 runs per over.

Soumya Sarkar made a half-century from 25 balls and added 81 for the second wicket with Mohammad Naim (38) to raise the hopes of the touring team.

The pair took 19 runs from the sixth over bowled by Ish Sodhi and 20 from the seventh from Adam Milne to briefly drop the required run rate below 10.

But part-time off-spinner Glenn Phillips played a vital role, allowing only four runs from the eighth over and nine runs off the 10th. Sodhi gave up five runs off the ninth and the run rate climbed again.

Soumya fell to Southee for 51 in the 11th over and Bangladesh's task became insurmountable.

New Zealand's innings moved in fits and starts, propelled at the end by Phillips who made an unbeaten 58 from 31 balls - his 50 from 27 deliveries - and Daryl Mitchell, who made 34 from 16.

https://www.cricket.com.au/news/mat...2021-03-30?mode=amp&__twitter_impression=true
 
It obviously isn't perfect but a better method has not yet been devised. As soon as a potentially better method is devised, it will be tested and if it works better it will be implemented.

Until then we have two choices:

1. Just abandon any game that cannot be finished because of weather.
2. Continue to use D/L Method.
 
The controversy in today's match wasn't so much with the DLS method but the fact that the match referee didn't finalize the target in time for the Bangladesh innings to start, which is atrociously incompetent. They played out almost a couple of overs before the match was stopped to get clarity on the target.
 
D/L was never perfect. It has costed SA plenty of times. But I am not sure if they have a solution to it.
 
Last Monetarist said:
The controversy in today's match wasn't so much with the DLS method but the fact that the match referee didn't finalize the target in time for the Bangladesh innings to start, which is atrociously incompetent. They played out almost a couple of overs before the match was stopped to get clarity on the target.
Click to expand...

Yes exactly. Match referee is clearly an incompetent buffoon who should be called up and his performance put under immediate review. Major screw up, the kind that should really get people sacked but we should be nice and give him another chance.
 
(Reuters) - Bangladesh coach Russell Domingo cut a frustrated figure after his side started their run chase in Tuesday's second Twenty20 against New Zealand without knowing what tally they needed to win and criticised the conduct of the match officials.

A downpour brought a premature end to New Zealand's innings on 173 for five following which the big screen at McLean Park and the ICC website said Bangladesh's target under the Duckworth-Lewis-Stern system was 148 in 16 overs.

Play was then halted nine deliveries into Bangladesh's chase and they were apprised of their new target -- 170 runs in 16 overs, which was later changed to 171 after the 13th over, according to ESPNCricinfo.

The visitors eventually fell well short of the tally they needed to keep the three-match series alive.

"I don't think I have been involved in a game before where batters go out and don't know what the target is," Domingo said. "Nobody had any idea of how many we needed after five or six overs.

"I don't think the game should have started until it was finalised, before there was a clear indication of what is required, and what we needed at certain stages. I don't quite think the conduct was good enough this evening.

"(When we met the match referee) they were still waiting for calculations. If you are waiting, the game can't start. What was explained was they normally get it one or two balls in... No excuses, but this is just very frustrating."

A spokesperson for the ICC said an operational issue at the ground meant the DLS sheets, which contain the target scores at different points of the innings, could not be provided to the teams.

"The target score was verbally communicated to the umpires at the start of the innings," the spokesperson told media outlet. "However, play had to be halted after 1.3 overs as the teams requested for the DLS sheet.

"Play resumed once the sheets were provided to the teams."

The final match in the series takes place at Eden Park in Auckland on Thursday.

https://cricket.yahoo.net/news/bangladesh-coach-domingo-upset-dls-093022830
 
Canford Cliffs said:
So DLS is a mathematical formula designed by 2 british statistians but I am pretty sure most pundits don't understand how it works. This method has screwed many teams over the years with the most infamous screwjob being against SA in the 1992 WC.

Even if we analyze today's T20 game, Australia scored at a run rate of 6.95 for 19 overs and then it rained. How come the req run rate assigned to chasing team is 11.5 per over? What kind of calculation is this?

We have plenty of intelligent modern mathematicians. Its time to put this colonial age old DL method in the bin and adopt a new one.
Click to expand...

1997 was the colonial age?

That 1992 semifinal match was decided by applying the Most Productive Overs method. As this was widely criticised, Duckworth-Lewis-Stern was introduced instead.
 
How did this work out

I didn't see the match but according to espncricinfo New Zealand scored 212 in 50 over but West Indies was given 41 overs to chase same score according to d/l method.
So if they were reducing 9 overs so shouldn't be they have reduced 20 odd runs..
How did d/l work out in this case
 
DL method is necessary for rain-affected games.

But, I agree that it is faulty and it can be remodeled or replaced.
 
sweep_shot said:
DL method is necessary for rain-affected games.

But, I agree that it is faulty and it can be remodeled or replaced.
Click to expand...

Actually the reason I am asking this as yesterday match showed us that score made by batting first in 50 overs is being taken equivalent for chasing side in 41 overs.
How D/L have arrived to this calculation, don't make sense.
If WI were asked to chase 190 odd in 41 I can understand, removed 9 overs, removed 20 runs.
But giving same score?
Let say there is important WC match semi or final is going on. Team batting first scored 300 in 50 overs, and rain comes, so will they ask chasing team to score 300 in 41 overs. According to yesterday calculation looks like D/L will do the same thing.
 
Before somebody say chasing team knows from start and they can plan according to it, yes but same score with 9 over less?
Shouldn't they have deducted 15 to 20 runs to balance the equation.
 
You must log in or register to reply here.
Back
Top