Indian pitches: a veritable graveyard for bowlers!

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

now what is the error analysis I can do for this ?

THE ERROR ANALYSIS FOR THE LIMITATIONS OF YOUR THEORY! A theory without error analysis is a law...like einstiens LAW OF RELATIVITY. Your idea is NOT a law, it is a statistical theory with limitations- unless you do an error analysis FOR YOUR THEORY, it has zero credibility.

I sure as hell dont know to apply the standard error of mean, confidence limits and such ... so there enlighten me.

So you admit you know nothing of statistics yet you claimed to know it better than most and bragged about an insane amount of degrees and credentials just simply as an ego-trip ?

[Guest BossBhai]

-- Removed on request of the user --

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

Lets see some numbers ...

Some numbers to WHAT ? I said that YOUR THEORY is incomplete- hell, i mentioned in post # 108 atleast TWO GLARING FLAWS in your theory. That YOUR THEORY does not incorporate atleast those two (off the top of my head, there could be more) scenarios, leads to the conclusion that YOUR THEORY IS NOT ALL-INCLUSIVE, IT HAS FLAWS IN ITS MODELLING. Now, whether those flaws are GREAT or SMALL can only be determined by DOING AN ERROR ANALYSIS of your theory. So go on and do an error analysis OF YOUR THEORY. You'd have completed the error analysis, if you were capable of it, in much less time than you've spent trying to avoid doing so. i wonder why.

[Bongosamaj]

ohh and my "theory" i.e the RPT = adding all scores and dividing by no.of tests and likewise for Bowling avg = adding all scores and dividing by sum of wkts

No that is not your theory. Those are the instruments (the formulae) to your theory.

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

there is nothing else that goes inside the "theory" I can assure you ... zip nada zilch ..

The theory is to do with which place is easier/harder to bat in. RPT/WPT/WPR are your TOOLS towards the theory. No wonder you draw incorrect conclusions to your own theory when you cannot even discern theory from the tools used for that theory. What happened to the response to post # 116 ? I pointed out to you where YOUR OWN CONCLUSIONS FROM YOUR OWN THEORY ARE IN ERROR. So why are you arguing about the veracity of statistical analysis in the context of cricket, a far weightier subject, when you cannot even draw the logical conclusions from the theory you cook up in the first place ?

[The Outsider]

Lets see some numbers ...

Boss, taking your hypothetical example, I think what Bongo is trying to say is : Test 1 : 900 runs scored 35 wickets fell Test 2 : 900 runs scored 40 wickets fell Those are facts, straight numbers, there are no errors involved yet. Errors will enter into the picture when let's say based on the above numbers you lay down a hypothesis that the pitch used in Test 2 was more difficult to bat on than in Test 1. That might be true but it will depend on a host of variables like maybe the umpiring in Test 2 was worse than Test 1, maybe the quality of batsmen featuring in Test 2 were worse than in Test 1, maybe the quality of bowler featuring in Test 2 was better than Test 1, maybe atmospheric conditions made a difference to the outcome etc. etc. Those things will prevent you from making a simple statement like the pitch in test 2 was more difficult. Now, over a large number of tests played a lot of these factors would "even out". But cricket being a sport we know that there is no perfect evening out which will happen. Tendulkar given out wrong would appear the same as Sreesanth getting a life from a statistical sense of it being a wrong umpiring decision, but we know from a cricketing sense that's not true. So while there would be surely be a narrowing of these error bars as your sample size increases but anomalies like a great team playing for 15 years, or one team regularly copping bad decisions etc. would still leave uncertainty in your results which in such an imprecise and subjective thing like sport is very difficult to quantify. Perhaps, running Monte Carlo simulations will get us to quantifying the errors after intricate modeling. But yeah, broad trends can obviously be seen without having to do an intricate analysis.

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

But yeah, broad trends can obviously be seen without having to do an intricate analysis.

Yup. Those who do not try to see cricket through the narrow and utterly imperfect prism of statistics only and instead take time to understand the game already know that India is one of the easiest places to bat in. And guess what ? Boss's own stats supports that conclusion.

[Bongosamaj]

so why dont u show me how the error analysis is calculated for the data I provided ?

So are you saying you don't know how to for your theory ? Do you understand the FACT that error analysis is to do with error margin in THEORY part as well as in the DATA part ? Do you understand the simple fact that even if your DATA has zero error, your THEORY has errors in it and MUST be guaged ? Did you understand a word of what Swetabh posted ? I find it surprising, given your claim to so many qualifications and a superman-esque education level. While i do consider myself well educated in sciences and mathematics (predominantly calculus & statistics), i find it ridiculous to believe that i am having to spell out the basics of statistical analysis to someone who claims the level of education as you do.

[Guest BossBhai]

-- Removed on request of the user --

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

Else there should never be any such thing as stats in any sport and the samis should have their right to sit alongside the marshalls ...

There shouldn't BE any serious considerations to stats when talking really how good someone is on a truely sporting level. If you cannot see Tendulkar's greatness without the aid of stats, i am sorry to say, you don't understand cricket at all.

. and if you observe all RPW values tend to hover around the tight range of 32-35 in most cases which in itself suggests that the data is very consistent and there isnt a skewing effect.

Swetabh is talking about the error bars due to the INHERENT GUAGES IN YOUR THEORY NOT INCLUDING ALL POSSIBLE FACTORS. And you are still hung up on error in data ? Have you ever done a single paper on error analysis in university or at work ?

but Mr. CC here needed an escape clause and hence all the meandring ensued as usua

Nobody is meandering. I am ASKING you to provide an error analysis for your statistical analysis, since EVERY SINGLE STATISTICAL ANALYSIS MUST CARRY AN ERROR ANALYSIS (ERROR DUE TO THEORY AND ERROR DUE TO DATA COLLECTION) FOR IT TO COUNT FOR ANYTHING. Without any idea to how precise your theory in iterself is, there is no basis on claiming how good your theory is. Its just that simple or is it not clear to you yet that you are fighting a losing battle here, trying to defend the indefensible ( that sporting competence can be very accurately prescribed by statistics).

[Bongosamaj]

did you bother to check all the different scenarios which were accounted for earlier in the thread?

Yes and your claim that all the different scenarios were accounted for is false. As i pointed out to you in previous posts (which i see you've ignored), i've identified ATLEAST TWO scenarios that your theory does not take into account, thus having an ERROR in your THEORY. I am asking you for a proper analysis of it, since you beat your chest so much about being a collector of degrees. Is it still not clear to you or do you wish to try and hide some more from the fact that your idea of trying to use stats to talk about how easy/hard it is to bat someplace is a pretty lame approach ? Why are you not addressing the fact that you analyzed your theory incorrectly and i've even pointed out the post # where you did so ?

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

I have and would have put them up but for your dogged "Regardless of what comes out stats = lies" line of thinking and your propensity to meander. I explained this yesterday itself to you.

So, you mean to say that you are sticking away from the analysis because i've already said that i do not consider your dogged approach to stats to have any value ? Well in that case, you should have no problem agreeing to the request made by someone else ?

how many times do i need to adress that

once would suffice. Explain to us how you managed to draw the incorrect conclusion from your own set of data & hypothesis.

your propensity to meander

Make no mistake, this conversation has meandered because of YOU, not me. We were doing just fine in this thread's progress until you decieded to hide away and not do an error analysis to your own theory. And as i said, if you were capable enough to do it, surely, you'd have done so, for it'd have taken less time than what you've spent here arguing on why you won't do it. So lets not pass the buck here !

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

yes trying to save myself invaluable time .... do u mind ?

In that case, i am sure you have no problem if someone else without such a 'dogged aversion to stats' as you say, asked you for the error analysis ? As i said, i find it odd that you are spending more time trying to hide away from doing the analysis than actually doing it and taking the ballgame forward. Hell, if you show your error range to be minimal ( 2% for eg), you will be vindicated yourself, as i'd have no alternative but to accept the fact that your stochastic modelling is pretty accurate. C'mon, you apparently hold the literal equivalent of 3 4-year university degrees, all apparently in high tech & sciences - this should be child's play for you to find what is the error range for your theory and once and for all, settle the argument of ' statistical analysis is relevant/not relevant in cricket in a fine detail'. You will also then prove your presented theory (which you incorrectly analyzed btw) to be correct. I wonder why you stand at the doorsteps of proving your argument correct and yet spend more time trying to avoid doing so ?