**Official** England in Namibia and Zimbabwe

[marc71178]

However, as we've seen so many times, in actual matches things pan out differently every time. It's not logical to use the thing when discussing actual happenings.

Except that DL is modelled on just about every game in the past 20 years or so, thus taking this into account and showing what happens usually from that situation, thus making it a logical predictor of what would happen.

 

[Prince EWS]

1 game does not disprove a career record of 27 wickets in 19 FC games and 12 in 27 OD games.

Well, I did say I didnt think he was a good longer-version player, and as for his List A stats, they are heavily based on his international stats and a stint he had three years ago at the age of 16.

 

[Richard]

A generalisation, in other words.
When every game is different.

 

[marc71178]

It maps all games and shows the usual outcome.

Just because that doesn't fit in with your mythical world, you've decided it's wrong.

 

[Richard]

No, I've not decided it's wrong - I've decided you can't use it to say "this would have happened" - because each game is different.
Nonetheless, if anyone can develop a better way of creating a result, I'd like to hear it!

 

[marc71178]

This is talking about the validity of wickets in hand though.

Something it quite clearly shows.

 

[Richard]

Oh, yes, it does.
It does not, though, mean that wickets in hand invariably result in a massive increase in scoring-rate.

 

[Neil Pickup]

Under regression and expectation rules, the number on the DL table shows the expected result given two predictor variables (runs & wickets):

ie: T = a×r - b×w + e

I know this is inaccurate because the decrease related to wickets down isn't linear, and there are probably indices and logarithms involved, but that's unimportant.

The final variable, "e", denotes the random error term which has a mean of zero - therefore the expected total (E[T]) is the value in the DL table.

 

[Richard]

Good, good.
What does this mean in terms of the wickets lost not being able to be given as a certainty given that all matches are individuals?

 

[Neil Pickup]

Nothing is a certainty in anything but the correlation is strong enough for it to be a very reliable predictor.

Obviously, the closer the team is to using their resources (wickets/overs) the more accurate.

 

[Richard]

Yes, incredibly obviously.
The fact that more runs are almost invariably scored when more wickets are in hand does not mean that taking just 2 or 3 wickets in the first 40 overs is anywhere near so certain to see 100 or so scored in the last 10 as some suggest.
And if you want to find some figures that show the relative probability of it, fine by me!

 

[Neil Pickup]

The tables don't use raw numbers but percentages. A quick bit of manipulation on the line for 10 overs remaining shows:

10	9	8	7	6	5	4	3	2	1
47%	46%	45%	42%	39%	35%	30%	22%	13%	5%
295	292	289	285	279	271	259	244	226	210

Where the percentages express the increase you can expect in your total from the 40-over mark to the 50-over mark depending on how many wickets left, and the numbers then convert this to a benchmark score should the 40-over score be 200/X

 

[Richard]

Which don't, of course, take any account of the quality of the bowling left to bowl the last 10 overs, but there you go - no system can cover everything.

 

[marc71178]

But they have taken into account just about every ODI played ever, or is that just a minor point?

 

[Richard]

Err, no, it's a fairly major point.
In those ODIs there will be a fair bit of deviation.
Maybe Neil could find the SD of all the stuff he found earlier?

 

[marc71178]

Regardless of the deviation, it shows that it happens much more often than you give credit, thus making wicket-taking a key part of an ODI.

 

[Richard]

A key part, yes - but not as important as bowling economically, and certainly not neccessary if bowling economically.

 

[Prince EWS]

As I said earlier in another thread, obviously it would be great if all your players could bowl economically and continually take wickets, but these players are rare to find. Hence, I think you have to balance your team out. I think you should have some strike bowlers, and some economy bowlers, and its up to the captain to use them correctly.

On the isolated case of Prosper Utseya, he is clearly an economy bowlers, but he would take more wickets than he has if the other bowlers were better, and put pressure on the batsmen to attack Utseya.

 

[Richard]

As I said earlier in another thread, obviously it would be great if all your players could bowl economically and continually take wickets, but these players are rare to find. Hence, I think you have to balance your team out. I think you should have some strike bowlers, and some economy bowlers, and its up to the captain to use them correctly.

It's all a great theory, and it works fine if you can resist the temptation to tinker.
The trouble is (as Gough's case has recently illustrated perfectly) the minute one of your "economy" bowlers has a bad game or two, he immidiately gets jumped on. People will, almost without fail, call for wicket-takers to be brought in, that the bowler who's supposed to be bowling economically isn't doing his job.

 

[Prince EWS]

It's all a great theory, and it works fine if you can resist the temptation to tinker.
The trouble is (as Gough's case has recently illustrated perfectly) the minute one of your "economy" bowlers has a bad game or two, he immidiately gets jumped on. People will, almost without fail, call for wicket-takers to be brought in, that the bowler who's supposed to be bowling economically isn't doing his job.

But it really doesnt matter what the people say if the selectors do their job.