Luckiest and Unluckiest batsmen

[marc71178]

Yet the weight of people saying it (added to the continual lack of an agreement on what is a chance) is just dismissed as everyone else being wrong.

 

[Richard]

And I've shown why both are indeed wrong.

 

[Son Of Coco]

I'm looking at both.
It's vital to do so.

If you're looking at both by taking out players who have been dropped and still scored 0 you're adding a 'but' to the equation so it's not purely luck. It's become conditional.

 

[Son Of Coco]

No, it's just the only one you in your rather limited mind can think of.
There are all sorts of ways of random-selection.
If you actually took some notice of the matter, meanwhile, you might notice that the only random part of the matter was the selection - there the randomness ended and identification of patterns began.

You can be stubborn about it Richard, but asking your brother to pick IS NOT random. It's really as simple (or difficult) as that. That's 3 years of stats at uni and info from people who've been studying it a lot longer telling you that. No amount of "Am nots" when confronted with that fact will make a difference.

 

[Son Of Coco]

There is never a study which cannot be improved.
In anything, in fact, you can look at it and work-out how it could be made better.
You cannot point-out the flaws in mine and say therefore it's meaningless. And no, I'm not going to do the thing all over again, it took me long enough last time. Even if I did, as I say, there would still be things that could be done better.

There are studies that can be discounted due to a lack of effort on the experimenters behalf to even attempt to do simple things like randomise a sample however. On a basic level, the results mean nothing. If you were taking something away from your experiment then it would be the idea, if done properly, that you could compare batsman's luck. It would be hard to do in this instance though as not only is the sampling process flawed, there's no concrete definitions of what you were looking at and how you came up with your conclusions.

Basically, what we have is the ICC's report on the throwing caper. With no chance of seeing how the results were arrived at or what they mean.

 

[Son Of Coco]

No - it is not perfect.
If you wanted to achieve anything close to that you'd have to look at every reasonable-length career from about 1960 onwards (the earliest date that it'd really be possible).
And that'd take a year or so - of doing nothing but the study.
Something no-one, ever, would be prepared to undertake.

Look, I don't think you quite understand.
I have looked at different people, different interpretations of different dismissals, and compared them with my own - then I have formed an idea of what forms comparable descriptions.
And I've put that to use in this sort of study.

I wouldn't dream of saying it was - if I truly were asked to do something for such a prestigious purpose, I'd look at entire careers, regardless of how long it took. It'd be more than worth it.

Well, I wouldn't say no-one ever.....people have spent years doing studies on a wide variety of topics so it's entirely feasible that one day someone will do the same for this. It would throw up some interesting stats regarding whether a player can be unduly affected by luck during their careers.

I was under the impression that you'd read about the dismissals you hadn't actually watched, as I find it nearly impossible to believe you would have seen all the dismissals concerned - but not completely impossible. In this case some sort of observer bias has to come into it.

 

[Richard]

If you're looking at both by taking out players who have been dropped and still scored 0 you're adding a 'but' to the equation so it's not purely luck. It's become conditional.

I can see I'll have to go into the whole hog:
There are two things you need to look at: first-chance average and all-chance average. Both have their merits - the all-chance average takes all runs and all chances, so if you score 244 being dropped on 52, you get 244 runs and 2 dismissals. I prefer, myself, the first-chance one where you get what you gave your first chance on and no more.
For a first-chance average a 0 where you're dropped twice doesn't make a difference. For an all-chance one it does.

 

[Richard]

You can be stubborn about it Richard, but asking your brother to pick IS NOT random. It's really as simple (or difficult) as that. That's 3 years of stats at uni and info from people who've been studying it a lot longer telling you that. No amount of "Am nots" when confronted with that fact will make a difference.

I was not willing to waste paper by printing-out a load of stuff, cutting it up and dropping it into a container.
Believe me, I know perfectly well how random it was, and it was every bit as good as picking invisible names - no amount of throwing Uni research at me can change THAT.

 

[Richard]

There are studies that can be discounted due to a lack of effort on the experimenters behalf to even attempt to do simple things like randomise a sample however. On a basic level, the results mean nothing. If you were taking something away from your experiment then it would be the idea, if done properly, that you could compare batsman's luck. It would be hard to do in this instance though as not only is the sampling process flawed, there's no concrete definitions of what you were looking at and how you came up with your conclusions.

Basically, what we have is the ICC's report on the throwing caper. With no chance of seeing how the results were arrived at or what they mean.

And I'm not going to complain at that, it's a good comparison - but I maintain that it's not possible to do anything approaching a perfect study on this matter - so all we can do is identify lucky batsman, use first-chance scores when we know what we know and take it from there.

 

[Richard]

Well, I wouldn't say no-one ever.....people have spent years doing studies on a wide variety of topics so it's entirely feasible that one day someone will do the same for this. It would throw up some interesting stats regarding whether a player can be unduly affected by luck during their careers.

Like I said (somewhere) d'you really think I haven't scoured everywhere I can possibly find for something like that? Otherwise I'd have wasted my time and efforts doing what I did.
Of course I can't find anything approaching every study ever done, but you'd hope anything seriously major would be published.
I simply think that most people just don't realise how important it is.

I was under the impression that you'd read about the dismissals you hadn't actually watched, as I find it nearly impossible to believe you would have seen all the dismissals concerned - but not completely impossible. In this case some sort of observer bias has to come into it.

Like I say, I've used what I consider a very good system to study dismissals and should-be dismissals - I've read countless stuff from countless reporters, comparing it to what I know (and believe me, I've watched quite a bit of cricket in the last 4 years) and I can tell what various reporters mean when they talk about dropped catches and similar things.

 

[marc71178]

And I've shown why both are indeed wrong.

Except you haven't one bit.

It is impossible to cover for every batsman because a chance is not universally defined.
The system has no fair way to deal with sawn off batsmen.

 

[marc71178]

I was not willing to waste paper by printing-out a load of stuff, cutting it up and dropping it into a container.
Believe me, I know perfectly well how random it was, and it was every bit as good as picking invisible names - no amount of throwing Uni research at me can change THAT.

Because you know more than statisticians as well now.

Is there anybody you don't know more about their area of specialisation than they do?

 

[Richard]

Except you haven't one bit.

It is impossible to cover for every batsman because a chance is not universally defined.
The system has no fair way to deal with sawn off batsmen.

Yes, it does, I've stated that a thousand times - not-out if you're given out incorrectly or run-out by your partner.
Yep, you guessed it - that's open to subjectivity if you don't give it serious consideration, too - and it's equally possible to squeeze it if you were to force people to take it seriously.

 

[Richard]

Because you know more than statisticians as well now.

Is there anybody you don't know more about their area of specialisation than they do?

So they know, by assumption, that method X42ABN6 cannot be random?
Don't be ridiculous - the only way you can know is to have a full understanding of the individual situation, and I had that. I also have some knowledge of statistics and I do know about degrees of randomness and that 100% randomativity is totally impossible to achieve or even come close to.

 

[marc71178]

Yes, it does, I've stated that a thousand times - not-out if you're given out incorrectly or run-out by your partner.

And how is that fair?

Yep, you guessed it - that's open to subjectivity if you don't give it serious consideration, too - and it's equally possible to squeeze it if you were to force people to take it seriously.

Except on here we get different opinions of a lot of chances as to whether they were or not.

Are you saying we don't all take it seriously?

 

[marc71178]

So they know, by assumption, that method X42ABN6 cannot be random?
Don't be ridiculous - the only way you can know is to have a full understanding of the individual situation, and I had that. I also have some knowledge of statistics and I do know about degrees of randomness and that 100% randomativity is totally impossible to achieve or even come close to.

But you still ignore people who know more about it than you, because they've said you're wrong.

 

[Richard]

And how is that fair?

Err - they didn't get out, they get the credit for not getting out.
Understand?

Except on here we get different opinions of a lot of chances as to whether they were or not.

Are you saying we don't all take it seriously?

Yep, I am - how many people have dismissed out-of-hand the relevance of my first-chance score theorem?
Plenty - and coincidentally it's mostly those who insist on making disgreements where the truth is very obvious - the Lara 0-400*, for instance.

 

[Richard]

But you still ignore people who know more about it than you, because they've said you're wrong.

Yes, and they've got less understanding of the situation than me, because I did the thing - they didn't.

 

[Camel56]

So they know, by assumption, that method X42ABN6 cannot be random?
Don't be ridiculous - the only way you can know is to have a full understanding of the individual situation, and I had that. I also have some knowledge of statistics and I do know about degrees of randomness and that 100% randomativity is totally impossible to achieve or even come close to.

By some knowledge i hope you meant very basic knowledge because thats what you have shown. Basic knowledge of statistics is easy to come by, they teach it in high school. I know its a big assumption to make but i do assume you finished high school. More advanced knowledge of statistic and statistical relevance is learnt when you study statistics courses in university. You clearly havnt done that.

 

[Richard]

Nope, but I have a perfectly decent grasp of it, something you don't need University degrees to gain.