Why is Lillee rated above Imran?

[smash84]

That's crap. The fact is you cannot know. Lillee broke down with a back injury on your there, though he took it with him.

Except England wherever Lillee went he either broke down or had a bowling average of 100+ or didn't care to tour

Why is it crap? It is reasonable to think that WI batsmen would be at home on their own pitches which were slower and less bouncy than Aussie tracks

 

[smash84]

More like regardless that they had different players? That he was faster then? The pitches may have suited him? You don't know. I don't know.

I don't see why this is such a difficult concession to make. I mean, I'm happy to say Imran's bowling record here isn't that great at least in part because he toured here a couple of times when he either could not bowl or wasn't at his best through injury or age.

Look, Michael Vaughan toured here in 03 and scored a zillion at an insane average. By the logic you want to use, had he never played Australia in England, you would say "Well, he averaged so much away, he surely would have gone better at home". He didn't, as we know. But say he was injured in 05, would you say it's logical he would dominate at home because he did out here?

The truth is you can't know.

I mean, if it's that hard a concession to make, you should pop over to the insecure thread.

tbf I stated it earlier that you can say that Lillee was a better bowler despite the statistics and people would be fine.

But I think the real disingenuous part is where people try to twist stats too much in order to prove that somebody is a better bowler.

 

[Burgey]

tbf I stated it earlier that you can say that Lillee was a better bowler despite the statistics and people would be fine.

But I think the real disingenuous part is where people try to twist stats too much in order to prove that somebody is a better bowler.

The Insecure Thread caters for that too.

We're a broad church.

 

[smash84]

The Insecure Thread caters for that too.

We're a broad church.

 

[Prince EWS]

Why is it crap? It is reasonable to think that WI batsmen would be at home on their own pitches which were slower and less bouncy than Aussie tracks

Bowlers don't always bowl to exactly the same standard though. People seem to be increasingly posting on CW as if small sample sizes are not problems in the slightest and that, instead of looking objectively across an entire player's career, we should break down their careers into really tiny segments and then extrapolate them as if performance was somehow linear, and as if we didn't actually have whole careers to work with already that we've just disregarded for some reason.

Sometimes it really is just a coincidence. Sometimes bowlers don't bowl at their best; they're human so performance will be variable. Extrapolating a small sample just because it was against common opposition or on common ground or whatever misses that fact. Even beyond that though, sometimes figures don't represent how someone bowled.. over a long career that always evens out into irrelevancy (especially since once you retire, what you can and could do doesn't matter anymore) but when we insist on breaking down careers into check points of performance against each individual country at each individual location etc etc and use samples so small that they're at extremely high risk for statistical insignificance, it starts becoming ridiculous IMO. By all means take the standard of the opposition into account when assessing a performance, but average against a certain country and average in a certain country are increasing becoming the most over-used, flawed and misleading measures used on here.

 

[smash84]

Bowlers don't always bowl to exactly the same standard though. People seem to be increasingly posting on CW as if small sample sizes are not problems in the slightest and that, instead of looking objectively across an entire player's career, we should break down their careers into really tiny segments and then extrapolate them as if performance was somehow linear, and as if we didn't actually have whole careers to work with already that we've just disregarded for some reason.

Sometimes it really is just a coincidence. Sometimes bowlers don't bowl at their best; they're human so performance will be variable. Extrapolating a small sample just because it was against common opposition or on common ground or whatever misses that fact. Even beyond that though, sometimes figures don't represent how someone bowled.. over a long career that always evens out into irrelevancy (especially since once you retire, what you can and could do doesn't matter anymore) but when we insist on breaking down careers into check points of performance against each individual country at each individual location etc etc and using samples so small that they're at extremely high risk for statistical insignificance, it starts becoming ridiculous IMO. By all means take the standard of the opposition into account when assessing a performance, but average v a certain country or average in a certain country are increasing becoming the most over-used, flawed and misleading measures used on here.

lol.....there goes my attempts at Ikki like dabbling in stats

 

[Prince EWS]

lol.....there goes my attempts at Ikki like dabbling in stats

I've been sitting on that one for a while really, and it's unfair that you had to be on the receiving on given you're faaaar from the biggest culprit, so I should apologise.


I won't though.

 

[smash84]

I've been sitting on that one for a while really, and it's unfair that you had to be on the receiving on given you're faaaar from the biggest culprit, so I should apologise.


I won't though.

It's alright. I know that you are a ****

 

[Top_Cat]

While we're on bugbears, could we address the use of the term 'sample'? Colloquially people are off the hook but for us in the statistics game, it's not just a subset of the data. Sampling implies a systematic, scientific and defensible process and the point isn't to exclude data which make the data look nasty to deal with but to have a dataset which is representative of the true distribution of values. Manipulation/transposition/transformation of the data generally comes after sampling, the sampling isn't the manipulation itself. And when people use the term 'sample size' what they're really talking about are 'cell sizes'.

Feel better.

 

[Burgey]

Think I'll sample a Grandfather Port tbh.

 

[hang on]

Its logical, if Lillee wasnt that successful against the WI on his own pitches, why would he be much better in their own den, regardless of that test in 73?

not necessarily. by that token, ponting should never have scored a run in sri lanka against murali after his travails against singh and co. in india.

 

[kyear2]

Tbf i stated it earlier that you can say that lillee was a better bowler despite the statistics and people would be fine.

But i think the real disingenuous part is where people try to twist stats too much in order to prove that somebody is a better bowler.

+1

 

[kyear2]

stats aside - and on this one, imran is definitely on par with lillee, which no amount of finagling will gainsay - lillee is certainly the second greatest fast bowler i have had the great privilege of watching. he was the real deal in terms of pace, strategy, aggression, adaptability and, of course, the moustache (it is movember, after all).

however, all said and done, marshall still remains primus inter pares.

+ 1

 

[abmk]

Its logical, if Lillee wasnt that successful against the WI on his own pitches, why would he be much better in their own den, regardless of that test in 73?

Logically this isn't correct. Also Lillee's record vs WI in Aus isn't bad at all , in fact it is pretty good considering the strength of WI batting line-up at that time .....

Though Lillee played only 'official' test in WI, he did play the WSC tests in WI ( 14 of them ) , took 67 wickets @ 26.87 ....

HowSTAT! World Series Cricket - Bowling Aggregates

HowSTAT! WSC Supertest Match List

Pretty much full strength WI sides for most part ......

 

[ankitj]

Bowlers don't always bowl to exactly the same standard though. People seem to be increasingly posting on CW as if small sample sizes are not problems in the slightest and that, instead of looking objectively across an entire player's career, we should break down their careers into really tiny segments and then extrapolate them as if performance was somehow linear, and as if we didn't actually have whole careers to work with already that we've just disregarded for some reason.

Sometimes it really is just a coincidence. Sometimes bowlers don't bowl at their best; they're human so performance will be variable. Extrapolating a small sample just because it was against common opposition or on common ground or whatever misses that fact. Even beyond that though, sometimes figures don't represent how someone bowled.. over a long career that always evens out into irrelevancy (especially since once you retire, what you can and could do doesn't matter anymore) but when we insist on breaking down careers into check points of performance against each individual country at each individual location etc etc and use samples so small that they're at extremely high risk for statistical insignificance, it starts becoming ridiculous IMO. By all means take the standard of the opposition into account when assessing a performance, but average against a certain country and average in a certain country are increasing becoming the most over-used, flawed and misleading measures used on here.

I wanted to say exactly this, but was failing to articulate.

While you can look at completeness of record etc, but one cannot generalize from small samples and start predicting stuff like Imran will always do better than Lillee against WI and other way round in England, based on their averages alone. Such predictive statements can be justifiably made only when there are qualitative justifications to accompany the statistical evidence.

For this reason, I really don't think performance of fast bowlers in subcontinent is such a huge deal. If pitches don't help, then good performances are not a function of skill, quite obviously. That Marshall or McGrath have great averages in SC and Ambrose doesn't might just be down to those random factors and/or unique circumstances that you mention.

On the other hand, examples of proper prediction based on something qualitative or skill related would be to predict that Sehwag will fail in SA and ENG or that Warne will fail against India (don't ask me to go to insecure thread for this). There are strong skill related reasons to make those predictions, stats just happen to substantiate them.

 

[ankitj]

I've been sitting on that one for a while really, and it's unfair that you had to be on the receiving on given you're faaaar from the biggest culprit, so I should apologise.


I won't though.

I was sitting on it too, as I said in last post

 

[smash84]

I wanted to say exactly this, but was failing to articulate.

While you can look at completeness of record etc, but one cannot generalize from small samples and start predicting stuff like Imran will always do better than Lillee against WI and other way round in England, based on their averages alone. Such predictive statements can be justifiably made only when there are qualitative justifications to accompany the statistical evidence.

For this reason, I really don't think performance of fast bowlers in subcontinent is such a huge deal. If pitches don't help, then good performances are not a function of skill, quite obviously. That Marshall or McGrath have great averages in SC and Ambrose doesn't might just be down to those random factors and/or unique circumstances that you mention.

On the other hand, examples of proper prediction based on something qualitative or skill related would be to predict that Sehwag will fail in SA and ENG or that would be to predict that Warne will fail against India (don't ask me to go to insecure thread for this). There are strong skill related reasons to make those predictions, stats just happen to substantiate them.

As somebody mentioned earlier that once you start moving sub 25 there is usually very little to choose between two bowlers. They are both usually very very good.

 

[ankitj]

As somebody mentioned earlier that once you start moving sub 25 there is usually very little to choose between two bowlers. They are both usually very very good.

Not sure about the 25 cut-off either. I will wager that an average of 20 would turn out to be stastically significantly different from average of 25. But that's a separate matter.

 

[smash84]

Not sure about the 25 cut-off either. I will wager that an average of 20 would turn out to be stastically significantly different from average of 20. But that's a separate matter.

You lose

 

[ankitj]

Edited