Luckiest and Unluckiest batsmen

[Richard]

Why not?

A random sample cannot tell you a great deal about every player, only a few "random" people.

And given the randomness of it, it'll give you a pretty good idea of the general picture.

 

[Richard]

Erm, no it is not - the larger the sample the more value to be attached - you cannot work out the "average luck" without looking at all batsmen.

No, you can't - but you don't need to.
The difference between a random-sample and a wholesale-sample won't be that large, it's one of the basic rules of maths.

 

[tooextracool]

Sehwag is a lucky batsman - maybe Smith has benefited equally from his luck, but he's sure had less of it.

how is sehwag lucky? you've conveniently not given us sehwags fc average because you know that hes been less lucky than smith. at least in sehwags case,his fc average makes him go from very good to good. in smiths case it goes from very good to mediocre.

 

[tooextracool]

No, you can't - but you don't need to.
The difference between a random-sample and a wholesale-sample won't be that large, it's one of the basic rules of maths.

and one of the basic rules of math, is that a random sample becomes more and more accurate as you add more people to it. you could have a random sample of 1 person or 10 people, both of which will probably not give you anywhere near the right accuracy, especially with you picking and choosing unrandomly.

 

[marc71178]

how is sehwag lucky? you've conveniently not given us sehwags fc average because you know that hes been less lucky than smith. at least in sehwags case,his fc average makes him go from very good to good. in smiths case it goes from very good to mediocre.

It's all part of the "random" sample.

 

[Richard]

how is sehwag lucky? you've conveniently not given us sehwags fc average because you know that hes been less lucky than smith. at least in sehwags case,his fc average makes him go from very good to good. in smiths case it goes from very good to mediocre.

I've given you my best shot at Sehwag's f-c average in the period concerned. You could probably do a better job, given that I'd imagine they screen most India games in India.
Sehwag is lucky, though, for the simple reason that he gets lots of let-offs.

 

[Richard]

and one of the basic rules of math, is that a random sample becomes more and more accurate as you add more people to it. you could have a random sample of 1 person or 10 people, both of which will probably not give you anywhere near the right accuracy, especially with you picking and choosing unrandomly.

Well the sample I used was 15 batsmen in 2002.
I know that's not as accurate as 500 batsmen in the whole of the 1980s and 1990s.
But it's not massively different to having used 40 batsmen from 2000-2004 or something along those lines.

 

[marc71178]

Well the sample I used was 15 batsmen in 2002.
I know that's not as accurate as 500 batsmen in the whole of the 1980s and 1990s.
But it's not massively different to having used 40 batsmen from 2000-2004 or something along those lines.

Erm 15 people for one year is nowhere near as good a sample as 40 for 5 years.

15 people out of the number who played in one year is a minor percentage and not statistically significant.

 

[Son Of Coco]

And as I've said, on a random-sample comparison certain players benefit more than others.
Hence you don't need to do it for every player ever.

It's not a random-sample though Richard, you're picking and choosing.

 

[Son Of Coco]

As I say, I can't be expected to compare every single player, it'd take 20 years, a random-sample comparison is perfectly reasonable to demonstrate that some are more lucky than others.

Picking 10 out of possibly 400 players isn't a large enough sample to demonstrate anything, especially when there's been no random-sampling process put in place. Hand-picked isn't random sampling. You also have to make people aware of the exact process you've used, what results were found, and what you think that means. Not just say 'I've looked at it in detail and this is what I've found' - if you do this you start resembling the ICC.

 

[Son Of Coco]

That's the point I was making - it is not realistic to expect it to be done.

And if you ask me taking a random-sample of players, researching as thoroughly as possible the luck they received over a certain period and collating a first-chance average for them, then comparing it to the two or three players you've identified as quite clearly very lucky is a valid thing to do and as close to perfect as it's possible to expect anyone to go. And certainly far more valid at indicating a batsman's pedigree than the scorebook-average.

But as we've pointed out before Richard, your method seems to be very subjective. You need to eliminate as many confounding variables as possible and give clear cut definitions of what you've used to measure 'luck' etc.

Well, if you're a statistician it's very reasonable to expect it to be done, and there's a distinct possibility it already has to some extent (I'll have a look around later).

As we've said before, picking out players that you think are lucky and then going 'I told you so, they ARE lucky' isn't random-sampling.

 

[tooextracool]

I've given you my best shot at Sehwag's f-c average in the period concerned. You could probably do a better job, given that I'd imagine they screen most India games in India.
Sehwag is lucky, though, for the simple reason that he gets lots of let-offs.

hes had just as many let offs as smith has had, and hes averaged more than smith despite that. the fact is that if sehwag is a mediocre player because of his fc average then so his smith

 

[tooextracool]

Well the sample I used was 15 batsmen in 2002.
I know that's not as accurate as 500 batsmen in the whole of the 1980s and 1990s.
But it's not massively different to having used 40 batsmen from 2000-2004 or something along those lines.

no the problem is that you hand-picked 15 batsman, instead of picking 15 at random. theres a difference. and neither is 15 batsmen anywhere near conclusive enough a sample to use it as a general rule.

 

[Camel56]

I would have said Geoff Lawson was one of the most unlucky batsmen ive seen. To cop that ball from Patrick Patterson right in that face, smashing his jaw in the process was most unlucky.

 

[Son Of Coco]

I would have said Geoff Lawson was one of the most unlucky batsmen ive seen. To cop that ball from Patrick Patterson right in that face, smashing his jaw in the process was most unlucky.

haha, I think that was a mixture of not really being able to bat and raw pace. Oh, and not having a grill could add an element of stupidity.

 

[thierry henry]

Haha, well if it was that obvious why wasn't it posted before?

You'll have to remind me re: the purpose of the thread, or at least of your opinion as to what this purpose is.

Well, I may be days late, but....

the thread is discussing which batsman are the luckiest, therefore it is to be assumed that people posting here have accepted the possibility that luck does exist for batsman. Obviously some people don't believe in this luck, therefore this thread is not for them. It's not a "discuss whether batsman can possibly be lucky over a long period of time" thread.

 

[Son Of Coco]

Well, I may be days late, but....

the thread is discussing which batsman are the luckiest, therefore it is to be assumed that people posting here have accepted the possibility that luck does exist for batsman. Obviously some people don't believe in this luck, therefore this thread is not for them. It's not a "discuss whether batsman can possibly be lucky over a long period of time" thread.

It is all relevant big fella. The thread wasn't titled "Luckiest/Unluckiest Batsmen (posts only from people who believe in luck in the first place)". My original post was in reply to something that was already on here and I'm sure the starter of the thread would be aware that luck, or the abscence thereof, would be open to debate. Threads evolve, and this one has lead to the discussion that is being held here at the moment. There was no requirement that anyone posting here either did or didn't believe in luck for them to be able to participate.

 

[Zinzan]

Whilst one could argue that through a players career these things even themselves out....I genuinely believe some batsmen just seem to have the luck of the Devil..


Any thoughts ? Have I missed any one obvious...

Plse note...I'm not downplaying the ability or skill of any of those above "lucky" batsmen. I just feel they seem to have more than there fair share of luck.[/QUOTE

Correct, I was merely stating my opinion and expected it to be debated

 

[Richard]

It's not a random-sample though Richard, you're picking and choosing.

Picking-and-choosing would be using a stratified sample to try and tell myself what I wanted to find.
I've not done that.

 

[Richard]

But as we've pointed out before Richard, your method seems to be very subjective. You need to eliminate as many confounding variables as possible and give clear cut definitions of what you've used to measure 'luck' etc.

Well, if you're a statistician it's very reasonable to expect it to be done, and there's a distinct possibility it already has to some extent (I'll have a look around later).

As we've said before, picking out players that you think are lucky and then going 'I told you so, they ARE lucky' isn't random-sampling.

No, that's not what I did - I compared them to randomly selected batsmen whom I'd done the same for.
As I've said, I flatter myself I'm normally pretty good at working-out what should and shouldn't be out - and so would anyone else be if they actually bothered to try.