Duckworth/Lewis Method

[Mr Mxyzptlk]

Why don't you design a formula to replace that aweful D/L method?

 

[Neil Pickup]

Because I like the D/L method!

If you take the time to think it through and understand it, then you can see that it takes into account a lot of things and is as good as available at present.

 

[Bazza]

The D/L system is class. Revolutionary and hard to fault really. Sometimes it throws up something which at a glance looks ridiculous but it really is the biz.

 

[Mr Mxyzptlk]

I have found much fault with it over the years. For instance, I think it was India v West Indies in the Caribbean last year. WI started out having to chase 200+ off 50 then the rain and they ended up having to get the exact same amount of runs off 33 overs. That's a huge disadvantage.

 

[marc71178]

I have found much fault with it over the years. For instance, I think it was India v West Indies in the Caribbean last year. WI started out having to chase 200+ off 50 then the rain and they ended up having to get the exact same amount of runs off 33 overs. That's a huge disadvantage.

The WI still managed to come close to winning which shows that the D/L cost them the series!

Erm, looking at the card, there's no way the West Indian's can complain.

India made 260 all out in 50 overs (which the West indies bowled slowly and were docked 1 over for)

After 11 balls of the innings, there was a short break at 5-0 owing to rain, and 5 overs were lost - the target being revised to 248 from 44 overs.

That's nothing like the same number of runs needed in 33 overs!

 

[Mr Mxyzptlk]

My mistake, I meant WI-NZ

http://www-usa7.cricket.org/link_to...NZ_IN_WI/SCORECARDS/NZ_WI_ODI4_12JUN2002.html

 

[Neil Pickup]

Look at it this way... New Zealand had the six overs left at the end, and five wickets left. A superb opportunity to hike the runs up to 270+

The West Indies had 33 overs' time span left with ten wickets in hand. That means that they didn't have to pace the innings over 50, like New Zealand were doing until the rain intervened.

With 6 overs and 5 wickets left, there was approx 20% resources remaining. About 265 would have been the NZ final total.

With 33 overs and 10 wickets left, there's about 80% resources left. 80% of 265 is 212, giving a target of 213 off 33.

The figures (20 and 80) are approximations - the ones actually used were marginally different to give the target of 212 - this is how it works.

And, but for Scott Styris, the WI would have won.

 

[Mr Mxyzptlk]

Look at it this way... New Zealand had the six overs left at the end, and five wickets left. A superb opportunity to hike the runs up to 270+

The fact is that they didn't bat those 6 overs and the WI managed to keep them to a fairly reasonable total in the overs played. How is it possible that they would have to chase the same amount of such a drastic reduction of overs?? What justification is there for that??

The fact that WI managed to almost win despite is a good testament to the improvement of the team from a couple of years ago.

 

[marc71178]

West Indies could approach the innings knowing they had to bat for less time, so could be more aggressive as they still had 10 wickets left

When the NZ innings was curtailed, they had 2 settled batsmen in who had rebuilt the innings and could easily have scored a lot of runs in the last few overs.

Incidentally D-L doesn't take into account whether the team is 6 down because of a quick flurry of wickets or 6 down early on and rebuilt, but that's a different story.

 

[Mr Mxyzptlk]

I agree in there being a reduction but the reduction of overs was way too drastic for the situation.

 

[marc71178]

It was the amount of time left divided by the regulation time per over!

 

[Rik]

I don't really trust the DL method. I mean I've seen several matches which would have ended up close, become overwhealming wins by the team batting 1st. I've seen instances of a team chasing 100 then ending up having to get 130 in less overs...how can that be fair? I wouldn't mind seeing the team batting 2nd chase some rate decided by the run rate, that would provide some really exciting cricket...or what about the good old number of runs at that time in the match then double it or something...

 

[Bazza]

Rik no. Just no.

 

[luckyeddie]

A quick examination of the mathematics of the Duckworth-Lewis method has left me a little puzzled by your last post, Rik.

If a side batting first has been dismissed for 100, they have 0 resources left. How can the required target be higher?

Are you sure you are not confusing DL with the old method of 'discarding the least-productive overs'?

Unless, that is, the team batting first have themselves had their innings interrupted by rain too, causing revisions to their own resources at the time of the interruption(s) (which in effect will also change the way they will continue their own innings upon resumption of play).

 

[Rik]

A quick examination of the mathematics of the Duckworth-Lewis method has left me a little puzzled by your last post, Rik.

If a side batting first has been dismissed for 100, they have 0 resources left. How can the required target be higher?

Are you sure you are not confusing DL with the old method of 'discarding the least-productive overs'?

Unless, that is, the team batting first have themselves had their innings interrupted by rain too, causing revisions to their own resources at the time of the interruption(s) (which in effect will also change the way they will continue their own innings upon resumption of play).

No I am talking about innings that have been finished, ie 240-5. Then the rain comes with the team batting 2nd at 200-6 off 40 overs...and they end up having to score 80 odd off 5 overs. Doesn't really seem fair as they still could have won.

 

[marc71178]

No I am talking about innings that have been finished, ie 240-5. Then the rain comes with the team batting 2nd at 200-6 off 40 overs...and they end up having to score 80 odd off 5 overs. Doesn't really seem fair as they still could have won.

If that were the case, then it is likely the side batting first's innings had been interuptted at some point with those sort of scores.

 

[Rik]

If that were the case, then it is likely the side batting first's innings had been interuptted at some point with those sort of scores.

Yeah that's the confusing thing. There must be a better solution that the DL method, because it sort of takes away any chance a side 7 down with 50 to go in 10 overs has of pulling off a surprise win. Not only do we have too many ODIs, but the interruped ones are basically predictable too...

 

[marc71178]

Yeah that's the confusing thing. There must be a better solution that the DL method, because it sort of takes away any chance a side 7 down with 50 to go in 10 overs has of pulling off a surprise win. Not only do we have too many ODIs, but the interruped ones are basically predictable too...

It uses evidence gained from all ODI's played in history, which is as accurate a method of predicting what's going to happen as anything can be got.

 

[Rik]

It uses evidence gained from all ODI's played in history, which is as accurate a method of predicting what's going to happen as anything can be got.

So, because of the lack of shock wins by sides who look like they were gone but ended up winning, it takes away any chance of the team actually winning and makes the match a foregone conclusion? If you ask me that sorta takes away the key element which makes Cricket the game it is, unpredicability...

 

[Neil Pickup]

Find me one scorecard that backs up what you're saying here, Rik, and we will explain exactly why the D/L method is correct.

And as for unpredictability... erm, no argument here. You can't have a random variable in target setting!

Why not just toss a coin?