Wickets per match - how important?

[Days of Grace]

Hi guys,

Just working on my Test Bowlers' Ratings. It's taking a while.

Right now, I'm stuck on wickets per match. I have adjusted bowling averages across the board, but I'm wondering if it's worth adjusting wickets per match.

I have adjusted the bowlers before WWI. It's quite simple. As 10% more wickets fell per match on average prior to WWI, then I simply took 10% off each pre-WWI bowler's WPM average. For example, Sidney Barnes has been adjusted down to 6.30 from 7.00.

The problem I have is comparing, for example, Richard Hadlee and Malcolm Marshall. Hadlee has the higher WPM, but that is basically due to him being so far ahead of the other NZ bowlers during his time playing.

Marshall, on the other hand, competed with other greats such as Garner, Holding, Ambrose, etc.

So, how to adjust those WPM? An almost impossible task, and is does it need to be adjusted anyway, since one could argue that Hadlee had to carry a bowling attack and Marshall always had other bowlers helping him by always keeping pressure ont the other batsmen.

Any thoughts?

[Goughy]

IMO, very important.

There are reasons why it may differ slightly but it is a meaure of production. Something average doesnt measure.

[Migara]

Wickets per innings has a greater importance IMO.

[Days of Grace]

You see, Migara, I have debated between Wickets per innings and wickets per match myself.

I'm sticking to the WPM at the moment because, whilst it favours those bowlers who get to bowl in 2 innings a match most of the time (their batting side is not following on), if the bowler is not good enough, he is not going to get to get the ball to bowl with, is he?

[Migara]

Originally Posted by Days of Grace

You see, Migara, I have debated between Wickets per innings and wickets per match myself.

I'm sticking to the WPM at the moment because, whilst it favours those bowlers who get to bowl in 2 innings a match most of the time (their batting side is not following on), if the bowler is not good enough, he is not going to get to get the ball to bowl with, is he?

The part in the bold is my concern. If batsman are not ggod enough to set targets, so a match ends in 3 innigs, nothing a bowler could do.

[Days of Grace]

Then it should be wickets per team innings, not wickets per innings in which a bowler actually partcipated.

[Migara]

Originally Posted by Days of Grace

Then it should be wickets per team innings, not wickets per innings in which a bowler actually partcipated.

Exactly. That's what I am at.

[bagapath]

used to keep 4 as the cut off point for determining greatness in bowling when it came to WPM. realized too many greats like akram, bedi, lindwall, botham, willis and gibbs miss out on that criteria then.

may be 2.25 wickets per innings should be a good starting point. with four bowlers in all test teams that kind of a WPI rate would get you nine wickets. a part timer or a run out should take care of the other wicket. all i know is this is a pretty important stat but dont know how to use it.

an old database on channel 4 site, much before cricinfo and wisden merged, used to allow one to generate these kind of lists where you can practically write your own formula to filter players. too sad it is gone and statsguru has not got that kind of flexibility - or may be it does and i haven't worked it out yet.

[SJS]

Lets put it this way. Here is a list of the top ten bowlers as far as wickets per Test are concerned. You need to try and understand what that list says to you.

1. Barnes, Sydney F
2. Lohmann, George A
3. Muralitharan, Muttiah
4. Turner, Charles T B
5. Grimmett, Clarence V
6. O'Reilly, William J
7. Blythe, Colin
8. Lillee, Dennis K
9. Peel, Robert
10. Hadlee, Richard J

I have taken out Dale Steyn whose career isn't over.

Now put the same bowlers (over a 100 Test wickets) by average and the top ten are :-

1. Lohmann, George A
2. Barnes, Sydney F
3. Turner, Charles T B
4. Peel, Robert
5. Briggs, John
6. Blythe, Colin
7. Wardle, John H
8. Davidson, Alan K
9. Marshall, Malcolm D
10. Garner, Joel

Now here are the ten at the bottom of that list (with the one with the lowest wickets per Test index being at number 1. Unlike in the case of Dayle Steyn, I havent taken out Kallis from this list since I do not expect any fantastic change in Kallis's figure by the time his career is over.

1. Hooper, Carl L
2. Shastri, Ravishankar J
3. Kallis, Jacques H
4. Illingworth, Raymond
5. Bailey, Trevor E
6. Abdul Razzaq
7. Rhodes, Wilfred
8. Emburey, John E
9. Boje, Nico
10. Johnson, Ian W

Now look at the bottom ten by averages:-

1. Hooper, Carl L
2. Boje, Nico
3. Shastri, Ravishankar J
4. Mohammad Rafique
5. Giles, Ashley F
6. Edwards, Fidel H
7. Wright, Douglas V P
8. Emburey, John E
9. Tufnell, Philip C R
10. Malcolm, Devon E

I think one can see what these lists tell.

How much a side relies on a bowler/or how much he was above his team mates (bowlers) is best indicated by the high wickets per Test and vice versa. Thus in spite of his fabulous record as a bowler, Marshall does not figure in the top ten (he just missed out) because the team had other strike bowlers too while Hadlee in the weaker attack of New Zealand has more wickets per Test.

The strike rate indicates , of course, the propensity of a bowler to take wickets. Generally a higher strike rate should mean a more attacking bowler. Thus faster bowlers will invariably have lower strike rates and amongst spinners, leg spinners will also have lower strike rates.

Average is a good medium because it is a product of the economy rate and the strike rate.

One can, with just a bit of careful study and by giving time make more out of the stats by looking at all the figures.

Our problem is we are interested in knowing who was better thats all. There can be no single answer to that.

For New Zealand, Hadlee was priceless because they had no one else. Maybe if he was bowling in the West Indian pack with Marshall, Holding and Garner, his figures would have been different. Would that make him a better or worse bowler - of course not. If we just managed to rid ourselves of the obsession to conclusively place everyone on a measurable scale, we would be able to appreciate the greats of the game much better. Otherwise our debates are reduced to seeking self glory from the deeds of those we support by a blind adherence to just those statistics that support our stance.

[wfdu_ben91]

I fail to see how wickets per match has any influence when comparing 2 great bowlers with all circumstances considered. The percentage of top-order batsman dismissed would have greater relevance, IMO.

[zaremba]

Originally Posted by SJS

For New Zealand, Hadlee was priceless because they had no one else. Maybe if he was bowling in the West Indian pack with Marshall, Holding and Garner, his figures would have been different. Would that make him a better or worse bowler - of course not.

= the point.

[fredfertang]

Interesting point about Hadlee's level of support - if he'd been part of a pack all supporting each other it may be he'd never have felt the need to cut down his pace - if he hadn't cut down his pace he may never have become the great bowler he undoubtedly was and presumably wouldn't have had the longevity he did

[JonnyB]

Originally Posted by Days of Grace

Then it should be wickets per team innings, not wickets per innings in which a bowler actually partcipated.

Agreed, good article by the way.

[Days of Grace]

Good points made, especially by SJS.

Just thinking from my ratings' point of view, how I can adjust the WPM statistic.

I first thought I could adjust it by looking at the amount the % of overs bowled by a bowler in their team's innings.

But then I thought that a lot of great bowlers bowl about an equal amount of overs compared to their somewhat inferior teammates.

There must be some other way. I'm thinking that if you have more teammates who are of a similar or higher level to you, you will face more competition for wickets.

Is there an objective analysis that can work this out?

I'm pondering.