DoG's top 100 test bowler countdown: Adjust for wickets per innings

[kyear2]

Adjusting WPI is going to be very complicated. What is a better measure here to judge competition? Their averages or their SRs? The strength of your batting might make more difference than your bowling, as well as a bunch of other factors.

I think for a Hadlee/Murali, they benefit in WPI from low competition. But having to bowl dry/long drags their averages up. Hopefully these balance out for overall rating.

I've found, purely from observation, this not to be the case.

They still.do pick up wickets along the way and benefit from more unfettered access to the tail.

But we can agree to disagree.

 

[DrWolverine]

If there’s only one great bowler in the team, batsmen would be more careful against him and tend to take more risks against the lesser bowlers.

 

[Qlder]

In terms of adjusting wickets per innings, I meant adjusting the wicket volume in terms of the support (or lack thereof) a bowler had from his peers and how the percentage of team overs bowled. I did not mean adjusting the WPI by opposition strength of match conditions.

I don't see how this could be statistically solid as it seems too subjective, especially to downgrade/upgrade a whole career as each bowler was part of various strength attacks over their career.

Example: Lillee played most with Thomson (26 tests) and Max Walker (20), but also played a lot with Alderman (15), Lawson (11) and Pascoe (11) so any ranking of "bowler support" quality has to be subjective.

Interesting that Lillee would then be judged on who he bowled with but a quick look at opposition that bowled in the same matches shows in 10 Tests Lillee took 52 wkts while Imran took only 40 wkts in those same tests. Why should Lillee have those 52 wkts reduced and Imran's 40 wkts increased just because of who they bowled with?

In summary I definitely think a bowler should gets lots of credit for WPI, but any upgrading/downgrading based on who they bowled wth seems to be just a way to lower the score of high WPM bowlers like Murali, Hadlee and Lillee?

 

[Qlder]

Or maybe certain bowlers took many more wkts per innings then anyone else because their captain simply bowled them the most since they were the best bowler. Why get downgraded for that?

 

[DrWolverine]

If you are bowling alongside another ATG bowler, you have a luxury :
Shorter workload.
Lighter workload.
Fresher spells.

We should reward and not penalise such bowlers. They carry heavier workloads, endure more physical strain, shoulder more responsibility and are under more pressure.

 

[DrWolverine]

It is easier to take wickets when you are fresh and the batsmen are not settled than when you are bowling long spells and getting tired while they get used to you.

 

[Bolo.]

I've found, purely from observation, this not to be the case.

They still.do pick up wickets along the way and benefit from more unfettered access to the tail.

But we can agree to disagree.

You dont have unfettered access to the tail when your team isnt strong enough to get you to the tail twice. The proportion of non-tail wickets that go down is much higher for stronger teams. And you are bowling tired/at set bats/at times you are bowling badly/unsuited to conditions more.

Your point would be more valid for guys who are not one of the top team bowlers and dont get a crack at the tail. How valid... Im not sure. But not many of these guys are sitting amongst the top bowlers anyway.

 

[DCH]

Bret Lee 310 wkts S/R 53.3 wkt/inns 2.07
Zaheer Khan 311 wkts S/R 60.4 wkt/inns 1.88

Bret Lee took 42 wkts of batsmen averaging under 10 (tailenders), while Zaheer only 12. This resulted in better bowling strikerate and wkts per inns for Lee as compared to Zaheer.

@DOG How will it be adjusted ?

 

[Qlder]

Bret Lee 310 wkts S/R 53.3 wkt/inns 2.07
Zaheer Khan 311 wkts S/R 60.4 wkt/inns 1.88

Bret Lee took 42 wkts of batsmen averaging under 10 (tailenders), while Zaheer only 12. This resulted in better bowling strikerate and wkts per inns for Lee as compared to Zaheer.

@DOG How will it be adjusted ?

Why should it be adjusted? It just means Lee's express pace was better at removing the tail so why should he be penalized for it?

 

[sayon basak]

Why should it be adjusted? It just means Lee's express pace was better at removing the tail so why should he be penalized for it?

Tailenders are objectively worse batters than top order batters, that could be one of the reasons.

 

[HimeshChaturvedi]

Tailenders are objectively worse batters than top order batters, that could be one of the reasons.

But in batters rankings, we weren't penalizing players for scoring runs against the worse bowlers or part timers of the attack (and frankly, not possible to find this info for older games without ball by ball data).

 

[sayon basak]

But in batters rankings, we weren't penalizing players for scoring runs against the worse bowlers or part timers of the attack (and frankly, not possible to find this info for older games without ball by ball data).

Good point. But I think part time bowlers are less of a thing than tailenders. Bowlers almost always needs to bat, but part time bowlers (specialist batters) don't have to bowl very often, if at all.

 

[Thala_0710]

Tailenders are objectively worse batters than top order batters, that could be one of the reasons.

And hence @ankitj's discount factor was born

 

[Randomfan]

And hence @ankitj's discount factor was born

Makes perfect sense given many tails have been walking wickets. No way taking top order wickets are same as taking 10/11 depsite both counting in stats same way.

 

[DCH]

Why should it be adjusted? It just means Lee's express pace was better at removing the tail so why should he be penalized for it?

Since Brett lee picked higher share of tailenders, the average average of batsmen he dismissed is lower as compared to Zaheer.
Lee 30.89
Zaheer 34.19.
This means on an average Zaheer dismissed batter averaging 34.19 about 11% better than Lee.

Same in case of Wasim Akram and Shaun pollock.
Pollock dismissed on an average (30.02-higher share of top order batsmen) and Wasim (26.43-higher share of tailenders). But Akram has better SR and W/inns. Quality of batsmen dismissed by Pollock is 13% better than Wasim.

 

[Migara]

I've found, purely from observation, this not to be the case.

They still.do pick up wickets along the way and benefit from more unfettered access to the tail.

But we can agree to disagree.

Actually no. A pack of good bowlers get to the tail more often and tend to get tail end wickets more. They may not have high RPIs, but frequent access to tail end wickets will improve their Averages and SRs.

 

[Migara]

Why should it be adjusted? It just means Lee's express pace was better at removing the tail so why should he be penalized for it?

Or it shows Australia got to the tail more often than India

 

[Migara]

Lets take a pack of bowlers who average 9 wickets per innings, who are more or less equally good. We'll call them team A. Next team (B) has a WPI of 6, where one bowler has a WPI of 4 and others combined to have 2. If we consider both teams have equal quality batsmen, the following could be observed in average; Assume the averages of wickets from 1 - 10 is 35, 35, 40, 40, 35, 35, 30, 25, 20 and 15

Team A will get to 9th wicket of team B in average. . Since the batting average drops according to the position. The bowlers of team A will average 295 / 9 = 32.6. Taking the 10th wicket which averages the least is very high

Team B bowlers will average 220 / 6 = 36.4 instead. Taking the wicket that averages least is less likely here.

Now drop one of the bowlers from team 1 averaging WPI 2 and put the Team B bowler in. Now this combination will almost always get to the 10th wicket, and statistically, the bowler injected from team B has the highest probability to take that no 10 wicket, helping the average and SR the most. Now attacks that averaged 10 WPI is almost unheard of. So even if you combine 4 of the most prolific WPI bowlers, I iamgine it would only marginally go pass 10.

This is why support improves Avg and SR greatly, while reduce number of wickets less.

Edit: Barnes, Murali, Grimmet and Hadlee manages to take all 10 every time, but three spinners is an over kill.

 

[Johan]

Edit: Barnes, Murali, Grimmet and Hadlee manages to take all 10 every time, but three spinners is an over kill.

That is two spinners.

 

[kyear2]

Actually no. A pack of good bowlers get to the tail more often and tend to get tail end wickets more. They may not have high RPIs, but frequent access to tail end wickets will improve their Averages and SRs.

Take a look at the numbers for yourself. Readily available.