The Duckworth-Lewis system ... can someone please explain to me ...

[Crystal Web]

I'd say the addition of the runs was a bonus more than anything else. Remember we were on the back foot for a large portion of our innings and had to rebuild our innings. It required a low risk, disciplined strategy, whereas D/L accounted for us adopting a slightly more "vigorous" strategy. I'm not sure we'd have batted any differently had we known beforehand that the game would be reduced to 43 overs. There was still a lot of confidence in our bowlers at that time. At the end of the day, D/L did not cost us this game. We were simply beaten by the better team on the day who took advantage of the opportunities we gave them, while we didn't capitalise on those offered to us. When two top teams meet, that's usually the deciding factor.

 

[Tacet]

I'm not going to comment on whether 17 runs for foreknowledge is a fair tradeoff - I honestly don't know. One place where I do feel the D/L system penalizes the first side to bat is interrupted innings. AB often takes a long time to get his eye in and to up his strike rate. He was starting to increase his strike rate when it started raining. After the rain he seemed to bat a bit slower again. I don't think D/L really takes this into account.

 

[AfricanTech]

I'm not going to comment on whether 17 runs for foreknowledge is a fair tradeoff - I honestly don't know. One place where I do feel the D/L system penalizes the first side to bat is interrupted innings. AB often takes a long time to get his eye in and to up his strike rate. He was starting to increase his strike rate when it started raining. After the rain he seemed to bat a bit slower again. I don't think D/L really takes this into account.

No, Miller took over

Oh, and at everyone shouting at Picard, the man is trying to find some solace so just leave him be.

D/L has screwed us more times than it has helped us (in my opinion not backed up by any kind of objective stats), so ja, we hatez it, my precious.

 

[JackWhite]

It doesn't matter. We should have been able to defend 298 in 43 overs. If you cant do that, then the other team simply deserved it.

 

[Bighit]

They did not add 17 runs for the 7 overs, enough people have explained that here. They added 17 runs to penalise NZ for knowing that they had 43 overs from outset.

Correct.

For the others the DL method isn't perfect, but to make sure that we have a game finishing in one day, we need some sort of system in place. If you want to say the DL method is horse cr ap, then provide a new system with formula to correct this.

Then, someone else has posted saying that you get the standard version and professional version and that before, you used to double the score at 30 overs to estimate a score at 50 overs, now it's closer to doubling the score at 35 overs; it's worth noting that the professional edition of the DL method does have small adjustments made to it every year so the formula and values used now is different to that which would have been used in 2003.

Overall, did the better team win? I think so.

 

[Fulcrum29]

I'd say the addition of the runs was a bonus more than anything else. Remember we were on the back foot for a large portion of our innings and had to rebuild our innings. It required a low risk, disciplined strategy, whereas D/L accounted for us adopting a slightly more "vigorous" strategy. I'm not sure we'd have batted any differently had we known beforehand that the game would be reduced to 43 overs. There was still a lot of confidence in our bowlers at that time. At the end of the day, D/L did not cost us this game. We were simply beaten by the better team on the day who took advantage of the opportunities we gave them, while we didn't capitalise on those offered to us. When two top teams meet, that's usually the deciding factor.

This is the only post I'm agreeing with.

 

[Ho3n3r]

Because we had lost 5 wickets.

Remember they also faced 43 overs


The system does not allow for our maxed out last overs that we often deliver.

Only the wickets and runs before the interval affects the calculation. Thus, at the time we lost only 3 wickets.

 

[Alton Turner Blackwood]

Guys. we didn't lose 7 overs, the game was reduced by 7 overs for each side. So both teams played 43 overs.

The problem is, we didn't know this that until it started raining. So in that aspect, D/L definitely favoured the Black Caps and the 17 runs added to compensate us for this is ridiculous.

If I were Dave Richardson, in the next board meeting I'd bring up a change - this is the 21st century we have much, much more advanced computational systems available.

Here's what I'd do.

Develop a neural network, which accepts the following as input:
From the batting team
Total overs faced
Run rate
Strike rates of the current batsmen
Averages of the current batsmen
Strike rates of the remaining batsmen
Averages of the remaining batsmen
It should accept the above rates for the current opponent and also against other opponents
A fuzzy logic value indicating how prone each batsman is to get out against each of the opposing bowlers as well as his ability to score against said bowler

From the fielding team
Strike rate of all the remaining bowlers
Economy rate of all remaining bowlers
Fuzzy logic value for line and length
Pace
Again, the system should accept these rates against the current opponent and also other opponents
A fuzzy logic value for each bowler against each opposing batsmen individually - call it Bunny. For example Warne's against Cullinan would be extremely high, something like 0.9000, meaning that Cullinan is extremely likely to go out to Warne.

General
Assign a fuzzy logic value to the frequency of dot balls
Assign a fuzzy logic value to the playing conditions (windy, drizzle, heat, etc)
A home, away or neutral indicator
Win/loss ratio

Now, feed all the above information into the system and develop a computational mathematical model.

Then using machine learning, over the next year, for every single player, in each and every game, record these details ball-by-ball. This is so we can teach the system that certain combinations of inputs will almost always result in a particular outcome.

You will end up with something similar to the Key to the Match (or is it Game, meh) which the Ausies are using in CWC 2015, only much more advanced. So much so, that it'll be able to tell the likely outcome of a game before a single ball was bowled.

Taking yesterday's game as an example.
Using the above you'd have been able to predict the behaviour of AB and JP from the 44th over onwards.

 

[Ho3n3r]

People seem to be oblivious to the fact that the D/L system, in this case, just calculated what we would've had after 38 overs, had we known from the start that the game was only 43 overs long, instead of 50 overs.

Thus, we had 204/3, while aiming to bat 50 overs. Thus, it has determined that we would've had 221 after 38 overs, had we been aiming to bat only 43 overs. Anything that happened after the resumption is irrelevant, as we were fully aware of the match situation then.

So the calculation seemed pretty fair to me.

 

[Fulcrum29]

Then, someone else has posted saying that you get the standard version and professional version and that before, you used to double the score at 30 overs to estimate a score at 50 overs, now it's closer to doubling the score at 35 overs; it's worth noting that the professional edition of the DL method does have small adjustments made to it every year so the formula and values used now is different to that which would have been used in 2003.

Duckworth/Lewis (D/L) is revised every two years, the last revision was done in November 2014. Steve Stern, an Australian, did the revision and updated the Pro edition to the Duckworth-Lewis-Stern (DLS) method, the Pro edition is now named the Stern edition. However, in all terms it stays the D/L method.

Here is an article on Steve's impression on his DLS custodianship :

http://www.couriermail.com.au/sport...wis-stern-system/story-fnii08h3-1227216578105

Now Picard has someone to blame

Here is a case study on the latest revision: https://radar.brookes.ac.uk/radar/items/7fdf8cb7-7904-48a0-b791-440ba8d6373d/1/

---

I must say... this, Z(u,0, λ)= Z0 F(w)λn(w)+1 {1- exp ( - bu/[λn(w)F(w)])}, will make an excellent password.

 

[Bighit]

I have just read it, but no matter what formula you use or introduce, nothing will compensate for the unpredictability of the ODI game and the formula. For instance, DL would never have taken into account the hat trick of Duminy, or even the hatrick of Langevelt against the Windies in 2007 or so. DL would have screwed us over 6 love in that.

I still show more interest in test matches though than any sort of limited overs and I'm happy that DL isn't in that.

 

[Fulcrum29]

I have just read it, but no matter what formula you use or introduce, nothing will compensate for the unpredictability of the ODI game and the formula. For instance, DL would never have taken into account the hat trick of Duminy, or even the hatrick of Langevelt against the Windies in 2007 or so. DL would have screwed us over 6 love in that.

I still show more interest in test matches though than any sort of limited overs and I'm happy that DL isn't in that.

The very same counts for the opponents.

 

[SaiyanZ]

I think it's fair to say the D/L revised target was understated as it doesn't take into account that ABD was at the wicket. Arguably the best batsman in the last 15 overs in recent times. D/L works with averages and ABD is far above average.

 

[Ho3n3r]

I think it's fair to say the D/L revised target was understated as it doesn't take into account that ABD was at the wicket. Arguably the best batsman in the last 15 overs in recent times. D/L works with averages and ABD is far above average.

This is unfortunately true. Didn't think of it that way.

And Miller was to come as well...

 

[AfricanTech]

I think it's fair to say the D/L revised target was understated as it doesn't take into account that ABD was at the wicket. Arguably the best batsman in the last 15 overs in recent times. D/L works with averages and ABD is far above average.

Well said.

 

[raind33r]

Oh right, let's just discount the wet pitch and ball.
The bowling is negated because of slippery ball. Bowler has to grip it just a bit tighter - it does affect the bowling!

DWL does not take into account the conditions for the 2nd team.

Game should have been replayed when both teams had the same field condition.

 

[SeRpEnT]

It doesn't matter. We should have been able to defend 298 in 43 overs. If you cant do that, then the other team simply deserved it.

I agree 100%.
If you cannot defend that then you just dont have good enough bowlers or very crappy fielding or a combination of both (which was the case yesterday).

 

[animal531]

Yeah, I'm not a fan of DL (especially when they had extra time the next day to play on)...but we were let down by our (at best) average bowling and very poor fielding.

 

[Beavis]

What is so sad or even grudge one more... is the fact that imo our fielding was not that bad as everyone makes it to be.... two missed runouts and a dropped catch... did you guys watched the game how they threw their bodies around fully committed.

How many dropped catches did Australia had yesterday, but because they still won nobody is bothered with their poor fielding. Guess one have to blame something right?

 

[OzzieCapie]

What is so sad or even grudge one more... is the fact that imo our fielding was not that bad as everyone makes it to be.... two missed runouts and a dropped catch... did you guys watched the game how they threw their bodies around fully committed.

How many dropped catches did Australia had yesterday, but because they still won nobody is bothered with their poor fielding. Guess one have to blame something right?

One catch dropped by WK that I saw but two out of two direct hits from side on