DoG's Top 100 Test Bowlers Countdown Thread 100-1

[stephen]

Honestly it should be an era adjusted ratio.

For all bowlers who bowled in my career, what was the average career wickets (1 wpm minimum)? How do I compare to that tally? Penalise if I'm lower than 75% of that average. Reward if I'm over 125% of that average.

Do the same for years if you like so that players with short careers are punished and long careers rewarded.

Tweak the ratios until you're happy with the results.

The advantage of this is that players are compared to their contemporaries rather than players of all eras. Sid Barnes taking 188 wickets was a greater accomplishment than Stuart MacGill taking 208.

 

[stephen]

But I firmly believe that players shouldn't be rewarded heavily for longevity nor punished heavily for being made of glass. Some consideration should be made for Anderson's incredible longevity but I'd much rather have Marshall leading my attack.

 

[Days of Grace]

Thanks for the explanation, I can now see your perspective more looking from a statistical point of view. For the record I mean you absolutely no offence, have certainty enjoyed this whole exercise so far

No worries at all. As long as posters understand the work I've put into it and that it is impossible to please everyone's subjective viewpoint, they can criticize as much as they like :-)

 

[Days of Grace]

How about the case of Headley and Pollock? Both similar career runs, but Headley had a 24 year career vs. Pollock's 7 year career.

Headley gets 68 career points vs. Pollock's 30.

 

[OverratedSanity]

How about the case of Headley and Pollock? Both similar career runs, but Headley had a 24 year career vs. Pollock's 7 year career.

Headley gets 68 career points vs. Pollock's 30.

Headley probably wouldn't have had a 24 year career if not for the war tbh. He played 3 tests in his 9 year "career" after the war ended.

Equal weightage to years and matches makes sense imo. That said, you shouldn't let the results affect your formula just because a few rankings seem off (unless of course it's something really egregious like Andy Caddick > Ray Lindwall or something). It defeats the purpose of the list imo.

 

[OverratedSanity]

Do you account matches missed in longevity? Maybe if a player plays 20 years and plays 100 tests but misses 50 test matches his team played over that span for some reason or the other, you could take the percentage of matches he played as a variable in career points? Don't know if it makes sense, just thinking out loud.

 

[Days of Grace]

Do you account matches missed in longevity? Maybe if a player plays 20 years and plays 100 tests but misses 50 test matches his team played over that span for some reason or the other, you could take the percentage of matches he played as a variable in career points? Don't know if it makes sense, just thinking out loud.

I had that in a previous version. But the missed wickets when not selected or available makes up for it imo.

I will look into incorporating it for a future version, though.

 

[jaideep]

George Headley played 22 out of 45 tests(48.9%) played by the West Indies in his career span.I think few career points should be deducted for players who play less than say 65% or 70% of the total matches played by their country in their career span.

 

[ankitj]

But I firmly believe that players shouldn't be rewarded heavily for longevity nor punished heavily for being made of glass. Some consideration should be made for Anderson's incredible longevity but I'd much rather have Marshall leading my attack.

Reasons why you would much rather have Marshall is well covered in other factors that are also part of this methodology. I am fairly certain Marshall is going to be ranked ahead of Andersen. So I don't see what's the point of your point. What injustice are you worried about?

 

[Days of Grace]

George Headley played 22 out of 45 tests(48.9%) played by the West Indies in his career span.I think few career points should be deducted for players who play less than say 65% or 70% of the total matches played by their country in their career span.

Or just give more weight to runs or wickets taken.

 

[jaideep]

Or just give more weight to runs or wickets taken.

I think it is fine as it is with wickets and career length in same ratio.Dont change it for 1 example like George Headley's.

 

[ankitj]

Headley probably wouldn't have had a 24 year career if not for the war tbh. He played 3 tests in his 9 year "career" after the war ended.

Equal weightage to years and matches makes sense imo. That said, you shouldn't let the results affect your formula just because a few rankings seem off (unless of course it's something really egregious like Andy Caddick > Ray Lindwall or something). It defeats the purpose of the list imo.

Agree with OS. Don't tweak the ratios until you're happy with the results. Otherwise the purpose of the exercise is defeated. The list is good as it is. It was good even without the 2 changes you made mid way. And those changes anyway didn't make much difference to results showing that we are arguing on margins. Your methodology is already bringing out some great insights. E.g. how much greater was Fazal's career than I had thought.

 

[Brian Lara]

How about the case of Headley and Pollock? Both similar career runs, but Headley had a 24 year career vs. Pollock's 7 year career.

Headley gets 68 career points vs. Pollock's 30.

Perhaps rather than taking the span of a career from first to last year, use the number of years they played in. You could also weight the years closer to what you consider a player’s prime if you wanted.

 

[Days of Grace]

I think it is fine as it is with wickets and career length in same ratio.Dont change it for 1 example like George Headley's.

Yes, you're right. I'm worried too much about outliers, which are few and far between.

In the end, if you missed a number of matches you lose out on wickets. And if you played a lot of matches in a short time you lose out on career length. So they balance out, I think.

 

[Bolo]

Do you account matches missed in longevity? Maybe if a player plays 20 years and plays 100 tests but misses 50 test matches his team played over that span for some reason or the other, you could take the percentage of matches he played as a variable in career points? Don't know if it makes sense, just thinking out loud.

It definitely works for this example. Headley gets 49% of 22 years. Pollock gets 70% or 100%, whatever it was, of 7 years.

I had that in a previous version. But the missed wickets when not selected or available makes up for it imo.

I will look into incorporating it for a future version, though.

The missed runs from missed matches were enough to make up for it when you were heavily favouring runs (wickets) at the start of this thread, but with an equal rating it should probably be reassessed.

 

[Days of Grace]

We are going around in circles here. Should I put it back to a 2:1 ratio for wickets:career length?

 

[stephen]

Yes, you're right. I'm worried too much about outliers, which are few and far between.

In the end, if you missed a number of matches you lose out on wickets. And if you played a lot of matches in a short time you lose out on career length. So they balance out, I think.

You'll always find things to change. I have enjoyed the list so far and am more keen to see it continued rather than tinkering around the edges with players moving up or down a place or two.

 

[srbhkshk]

Theoretically at least - I do feel Wickets should be weighed more than career length.

 

[jaideep]

We are going around in circles here. Should I put it back to a 2:1 ratio for wickets:career length?

Dont change it back to 2:1.1:1 ratio for wickets:career length is better.

 

[Bolo]

4-1 was too much. 1-1 or 2-1 are both ok, but obviously both will have problems. You are either overrating Rabada or Headley. You can't get around this by simply adjusting this. If you use OS's suggestion though, Headley could become effectively 2-1 with Rabada 1.1?-1, which penalizes each according to their own statistical weakness. It's not perfect, but it's better.

Anyway, just finish the list as it sits please. While we can debate the methodology as we go along, this is really just fine tuning. Push on and get a congruous list now that we are into the big names, reassess at the end and adjust if needed.