Indian pitches: a veritable graveyard for bowlers!

[Bongosamaj]

Are you still on drugs? :D

Are you disputing my statement ?

[yoda]

Are you disputing my statement ?

Yes, of course, DUH!

[Bongosamaj]

Yes' date=' of course, DUH![/quote'] On what basis ? So you are disputing that EE is very statistics heavy and that communications field is bereft of statistics. I see. Carry on then- i have nothing more to say to such blatant dishonest/ignorant statements from under-qualified people.

[Lurker]

totally wrong..Statistically Kapil is higher than Yuvi's sixes..Yuvi hit in 20/20 cricket where this kind of cricket is normal. and kapil did it in test match ...looking at different parameters...Kapil's 4 sixes in row is statistically greater than Yuvi's

Good for you if you look that way but unfortunately that isnt the case. Hitting 4 sixes at the time Kapil did was fantastic, no-one else did. Today it is common place. Folks who watch cricket would never understand what it meant for Kapil to have hit those 4 sixes. Why should they? Their are dime a dozen overs going for 28-30 plus runs these days. Just check the batting strike rate of batsmen today and compare them to 80s. Be prepared to hear how Kapil's 96 was nothing when compared to XYZ 105 plus. xxx

[yoda]

On what basis ?

Random thought, NOT!.

[Bongosamaj]

Random thought' date=' NOT!.[/quote'] Anyone who disputes the statement that ECE is the heaviest stats-oriented field outside of actual statistics major, might as well argue that the earth is flat, the sun spins around the earth, etc etc. Everything about communications ( a HUGE aspect of EE, be it radar, sonar, wireless or whatever) is statistics based and every course from 3rd year onwards will EXPECT you to be very proficient in stochastic modelling. So yes, you ARE arguing on purely basis of random thought and not enough qualification to speak on the matter.

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

Are you saying there is a chance that a Bowler with a Marshallesque Avg might actually be a Sami in reality and vice versa ?

No, i am saying that the gulf between them is vast enough that pure honest opinion is good enough a guage to rely on than the rather imprecise statistical tool you so love.

Prove it ... I can also quite simply type the opposite of what you say.

Your typing has zero authority in this field over mine . If you want to really debate me in statistics, be warned you are entering a field where i'd expect you to know 4th year statistical processes or not engage me. Anyways, the proof is very simple - for any statistical model you can come up with, i can come up with cricketing scenarios that is not counted in the statistical analysis. I've already provided two such anomalies in your statistical analysis: 1. Your Runs/Test guage is irrelevant without factoring in the FACT that if a side refuses to enforce follow-on and bats on (like Australia often does in the last 6 yrs), the runs/test WILL be higher than if they enforced the follow-on. Regardless of the bowling condition, the pitch condition, etc. that is a mathematical fact. Your analysis = ignorant of this FACT. 2. Your analysis is ignorant of the fact that margins of victory MUST also factor in the pitch equation for bowling & batting quality ARE NOT PROPORTIONATELY EQUAL IN EVERY TIMEFRAME. Consider India's inept displays in some tests to their competetive ones. Eg A : India : 300 & 320, Side X : 330 & 291/5. Total runs/test : 1241, rpw : 35.45 Eg B: India : 550 all out, Side X : 340 & 115. Total runs/test : 1005, rpw : 33.50 Eg A shows a much more even pitched contest, eg B shows a much more one-sided victory with nothing to do with the pitch. As a result, NATURE OF VICTORY must also be factored in and i am waiting to see how exactly your 'statistical knowledge' factors it in. These loopholes & flaws are proof enough that your statistical approach is flawed

And how many times do I need to tell you that IT WAS YOUR IDEA to look at stats on a decade by decade basis ? The problem with that is there were only 30 tests played in India during that decade .... which is not a good sample as a few highscoring games can skew the data. This is the reason why I prefer a larger time frame of atleast 2 decades.

Even based on 2 set of Data, it is clear that Pakistan has been harder to bat than India as well as West Indies. WI clearly is, as it ranks LOWER than india over the past two decades and Pakistan is Just ahead of India in this decade but FAR behind in the last decade (thus overall, harder in the TWO DECADES you want to consider). Yet your own comment is AGAINST the LOGICAL CONCLUSION I DEMONSTRATED FROM EXTRAPOLATION OF YOUR OWN STATISTICAL ANALYSIS. In short, you can't even draw proper conclusions from the goobledegook you present, let alone present something that isnt nonsense. And you want to debate statistics with me ? How audacious !

ust as long as you are willing to accord the same status to English pitches of 30s which have very similar numbers as India in 80s.

Show me the evidence and i will agree/disagree based on your evidence's veracity and the conclusions i draw from it. You've already shown with your 'pakistan & west indies easier than India' comment that you cannot even analyze data correctly, since your data alone leads to the opposite conclusion. So i am sorry, but i do not consider your judgement to be worth anything- at best, i am willing to give you the chance to present your data and let me see what the logical conclusion is.

because it blows a big hole in the Bradman = King theory as most of his runs came in the 30s.

no it doesn't blow a big hole at all. I've always maintained that Bradman's superiority is because of his COMPLETE DOMINANCE of his era which no other batsman ever had. I'd have considered your theory of 'pitches were very flat in the 1930s so bradman not greatest' if there were a few other batsmen averaging in the 60s and 70s in Bradman's times. Fact is, only one managed to and he played only 20 tests ( Headley). And Bradman posted an average literally more than 1.5x greater than Headley. Pitch, bowling, etc. does not come into the equation when the main point is disproportionate dominance = best ever.

And waht does it tell about you considering how you are yet to provide any sort of proof other than issuing blanket statements with no back up other than your own word ?

Blanket statements ? I've pointed out the logical FLAWS in your theory which you even acknowledged earlier and said you'll 'figure out' or 'misunderstood'. And if you want more proof, i'd ask you to do an error analysis and standard deviation of your statistical survey before you ask me about how good your theory is. For as any person who's done statistics worth his salt knows, statistical analysis is utterly meaningless without error analysis. Your figures count for squat until you present the scale of error in your analysis. Can you even do that ? If not, please withdraw all claims to knowing anything about statistics- for you do not if you cannot present the error analysis of your 'statistical analysis'.

[Holysmoke]

If you are doing 3rd or 4th year EE, then i call you a liar. Almost every course from 3rd year onwards is heavy on statistical analysis and the entire field of communications is based on statistical analysis.

I'm doing 4th year.I dont agree with you, so I am a liar :hysterical: Physics majors, and even Bioengineering majors use much more stats than ECE majors. Anyway, you carry on your bullshit.. nobody cares about the crap you spit out anyways...

[Bongosamaj]

Physics majors, and even Bioengineering majors use much more stats than ECE majors.

WTF! Entire f*cking communication related field- digital systems, analog transmission, power distribution, wireless comm, satellite work, etc etc. are ALL heavy on statistical analysis of the most advanced sort! Yes, biochem or physics do more BASIC STATISTICS ( error calculation & basic prob. theory) more, but high level statistics ? Sorry but this is just blatant lying/misderection for the sake of it. I can easily post here the curriculum of physics, biochem & EE in universities, down to the course outlines. If you see Physics or Biochem doing even HALF as much stat as EE, i swear i will eat my show and post the film of it on youtube.

Anyway, you carry on your bullshit.. nobody cares about the crap you spit out anyways...

Says the guy who wants to deport people and revoke citizenship based on some dumb sporting dream. Laugh!

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

there are no errors in this as I am not doing any sampling or modelling

False. Your data-set itself is a sample set. You are using sample data to base your conclusion and therefore there MUST be an error analysis to this!

Iam crunching the entire dataset using little more than common sense and common cricketing ABCD

Commonsense dictates that entire dataset of any dataset will have sampling error and hypothesis error, thus having an error analysis! Any hypothesis or statistical deduction MUST have an error analysis for it to be mathematically acceptable. That you claim your 'statistical outlook' to cricket is correct and yet, you cannot show any error range alone speaks about how you are trying to fit a square peg in a round hole!

And the reason I didnt provide those 2 stats u wanted was simply because you will refuse to see the light and will walk away with your usual "stats in cricket = meaningless" line ... so I didnt bother wasting time putting up that stat

No, because i don't think you CAN provide an error range for this hypothesis, simply because you do not know how to. I say stats in cricket is meaningless because they are and i speak from a position of authority when it comes to statistical analysis. And if you fancy your chances of proving me wrong, do an error analysis for your hypothesis!

and I see that there are others who think like me .

Whatever happened to your 'this isnt a popularity contest' spiel from just a few days ago ?

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

Once again there is no sampling .... whats presented is THE ENTIRE dataset ... in other words ... it takes into account every single RUN and WKT that is recorded in the Final score card of a Test Match for all Tests since 1877. Now unless you are saying there can be errors in those score cards itself I dont see how errors can creep in. I do not know how else I can simplify this further for you.

What nonsense ? The whole data set itself is a sample-set of events! The universal set is not exempt from standard deviation and hence all statistical analysis carries an error analysis. If you are disputing this, then i say you do not understand even the basics of statistics and it is a clear violation of DeMorgan's Theorem- the fundamental BASIS of statistcal analysis! Error is not in the scorecards- error is in your ANALYSIS. Every analysis in statistics is an IMPERFECT ONE. Which is what is the meaning of the word probability . Thus, the ERROR analysis is in reference to the THEORY/FORMULA used and what is the error in its predictive capacity. I don't know how much simplified i can make this, this is STAT 100- something you claim to have studied but apparently flout the first chapter of ! Read my words carefully, for these are worth engraving in stone, for its a statistical FACT: Every statistical analysis must account for error analysis, for the veracity of its theory. All statistics is probability analysis and stochastic modelling. All probability and modelling is an approximation, thus each and every probability study contains the error range. When you collect sample data in laboratory, there is error range simply from the capability of the instrument- regardless of how many data points you include. Similarly, since your statistical quantification is not 100% inclusive of all cricketing phenomena relevant to this analysis, you MUST present an error analysis if you wish us to take your statistics seriously. PS: Don't argue about error analysis - the very fact that you produced a statistical study and forgot to mention the error analysis shows either your ineptitude in the field of statistics or the inability of your statistical modelling of the situation. But bottomline is, error analysis MUST be present in EVERY statistical modelling. Each and every one. At the very least, provide us the standard deviation of your figures from the median. PPS: Why are you refraining from commenting on the fact that your ANALYSIS was incorrect in the case of West Indies and Pakistan in the 'two decades' you prefer to look at, in relation to India ?

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

ehh? A run is a event and so is a wkt .... and I have accounted for every single one of them.

SO WHAT ? You have not shown the error margin FOR YOUR THEORY ! Or are you claiming that your theory is 100.00% statistically accurate, thus being literally watertight ? This too, after i pointed out atleast two scenarios off the top of my head that IS NOT incorporated in your analysis ? Did you understand the ABCD of statistical analysis when you declared that 'cricket is the statistician's delight' and promptly proclaimed your understanding of statistics ? Or do you choose to proclaim you know statistics yet violate the most fundamental of statistical axioms ?

Dunno which stat u are refering to pls provide the post #

Post #65 proves categorically that your own analysis of your 'theory' is faulty ! Please answer this question : Do you or do you not think that every statistical analysis carries an error range with it? Keep in mind that answering in the negetive to the above question, absolutely establishes the fact that you know almost nothing about statistical analysis.(which has already been shown in this thread so far, but nevermind that).

[The Outsider]

Why are we getting into all this statistical jargon here, I don't understand. Cricket is a sport and no sport can be completely accurately modeled by statistics, for that matter nothing can be. But one would be willing to spend hours of statistical modeling and supercomputer time to model something important like weather to get good accuracy, but what's the point in doing it for a sport. Moreover, no one would be interested in such an analysis because people who follow cricket might have a some interest in numbers associated with the game, but would much rather watch the matches or play the game than spend hours trying to weed through sophisticated statistical modeling. At the same time, basic statistical parameters without overt importance given to them can be used to get general and intuitive trends and can make for some fun read. For example, taking the geometric mean of average and strike rate can give you a nice indicator of successful ODI batsmen, or like at the topic at hand just a simple runs/wicket over different countries can confirm your intuition that it's easier to score runs in Pakistan than in NZ on average. One can keep on dwelling for the perfect answer by making the analysis more and more complicated but the gains for the effort and complication are minimal in such a subjective thing as sport.

[Bongosamaj]

Why are we getting into all this statistical jargon here, I don't understand. Cricket is a sport and no sport can be completely accurately modeled by statistics, for that matter nothing can be.

We are getting into this statistical jargon because Boss thinks that statistics are an accurate way of representing cricket and cricket is a statistician's delight.

Moreover, no one would be interested in such an analysis because people who follow cricket might have a some interest in numbers associated with the game, but would much rather watch the matches or play the game than spend hours trying to weed through sophisticated statistical modeling.

Well there will always be the sad cases of Bheemie devoting more time to statistics than watching the game or understanding it. Hence the debate in the first place.

[Guest BossBhai]

-- Removed on request of the user --

[Bongosamaj]

lets say we are counting the no.of males & females in a neighbourhood does that need an error analysis when you can go meet each one of them ?

Yes, we need an error analysis depending on what your THEORY is.

You get a count and then you say females are x% of the population end of story

The error in a statistical analysis is from std. deviation in sample range AND std. deviation due to the LIMITATIONS OF THE THEORY. F=ma itself has an error range to it, regardless of what sample you use. Is that not known to you ? I am asking you to do an error analysis FOR YOUR THEORY. Unless ofcourse, you are claiming that YOUR THEORY is 100.00% true and thus includes EVERY SCENARIO in its modelling. Again, answer this question with a 'i think so/i dont think so' answer to make it very clear : Do you or do you not think that every statistical analysis carries an error range with it?