The Outsider Posted January 20, 2008 Share Posted January 20, 2008 In India's matches when Bucknor has been umpire, India have a W/L ratio = 0.83 During that period in all matches, that India have played their W/L ratio = 1.28 During that period when India has played without Bucknor, their W/L ratio = 1.33 Thank you Stevie Wonder.:finger: Link to comment
Brainfade Posted January 20, 2008 Share Posted January 20, 2008 Good one, prof. What sample sizes are we talking about? Link to comment
Ram Posted January 20, 2008 Share Posted January 20, 2008 Honestly, we shouldnt even bother looking at such stats, it reveals a very defeatist mindset, IMO. We should see how many catches we took/dropped, how many 100s we scored etc in victory defeat. External factors beyond our control should never take our time and energy. Link to comment
The Outsider Posted January 20, 2008 Author Share Posted January 20, 2008 Good one' date=' prof. What sample sizes are we talking about?[/quote'] Not huge obviously and the thread is only half serious. 22 tests with Bucknor and around 140 in total. Link to comment
The Outsider Posted January 20, 2008 Author Share Posted January 20, 2008 Honestly, we shouldnt even bother looking at such stats, it reveals a very defeatist mindset, IMO. We should see how many catches we took/dropped, how many 100s we scored etc in victory defeat. External factors beyond our control should never take our time and energy. Getting rid of the "external factor" ensured our victory at Perth. If the "external factor" was around we would have been 0-3 down. On the other hand if the "external factor" had been removed earlier we would have been 2-1 up. Link to comment
Ram Posted January 20, 2008 Share Posted January 20, 2008 God knows why I posted such a serious reply to a thread mostly in jest ! :D Link to comment
Anakin Posted January 20, 2008 Share Posted January 20, 2008 This is silly? How exactly Link to comment
DesiChap Posted January 20, 2008 Share Posted January 20, 2008 Facking Sad but true. The biggest gain of the tour "we have seen the last of Stevie "wonder" Bucknor Link to comment
Sachinism Posted January 20, 2008 Share Posted January 20, 2008 even without the stats we all know bucknor has got it in for india always looking to screw us over Link to comment
observer1 Posted January 20, 2008 Share Posted January 20, 2008 There is only a 1:50 chance you could be wrong, Herr Professor.... Analyze a 2x2 contingency table india win india lose Total bucknor present 83 100 183 bucknor absent 133 100 233 Total 216 200 416 Fisher's exact test The two-tailed P value equals 0.0181 The association between rows (groups) and columns (outcomes) is considered to be statistically significant. Link to comment
Dhondy Posted January 20, 2008 Share Posted January 20, 2008 There is only a 1:50 chance you could be wrong, Herr Professor.... Analyze a 2x2 contingency table india win india lose Total bucknor present 83 100 183 bucknor absent 133 100 233 Total 216 200 416 Fisher's exact test The two-tailed P value equals 0.0181 The association between rows (groups) and columns (outcomes) is considered to be statistically significant. Hi, monkey speaking here. This anthropoid begs to point out slight misuse of stats. Firstly, Fisher's exact test is used for actual values, not proportions. Secondly, even if this were actual values, albeit an impossibility for one man to umpire so many Tests, you would be advised to use the chi-square test, as none of the row or column values are less than 5. Thirdly, there is an obvious bias there, as Bucknor wouldn't have umpired against WI, against whom India would have won very often, and being a senior umpire, would have been chosen to umpire in matches that India played against Australia or SA, where they were likely to do poorly. A case in point can be seen by comparing the win:loss ratios of any team when Simon Taufel and Russell Tiffin have umpired (not simulataneously). Link to comment
The Outsider Posted January 20, 2008 Author Share Posted January 20, 2008 Yeah, the sample size is obviously an issue here and with such small numbers (5 wins, 6 losses with Bucknor), it's not much more than a fun poking exercise. In addition to the bias Dhondy pointed out, there will also be the bias of him not being chosen for less important matches that India played against Zimbabwe and Bangladesh. Link to comment
observer1 Posted January 20, 2008 Share Posted January 20, 2008 Dont blame me hanumanji. Blame graphpad I only input Professor's values - gp was the one which said that Fisher's was better than chi squared - me, i know no difference other than between a nail and a hammer. Anyway, to satisy your simian tastes, here is another banana.... Analyze a 2x2 contingency table win lose Total bucknor present 83 100 183 bucknor absent 133 100 233 Total 216 200 416 Chi-square with Yates correction Chi squared equals 5.186 with 1 degrees of freedom. The two-tailed P value equals 0.0228 The association between rows (groups) and columns (outcomes) is considered to be statistically significant. Link to comment
Dhondy Posted January 20, 2008 Share Posted January 20, 2008 Dont blame me hanumanji. Blame graphstat I only input Professor's values - it was the one which said that Fisher's was better than chi squared - me, i know no difference other than between a nail and a hammer. Graphstat is obviously written by an ape in an even lower state of evolution than the hanuman (maybe a chi-mpanzee?). Fisher's is only used in preference to Chi^2 when the individual values are Link to comment
observer1 Posted January 20, 2008 Share Posted January 20, 2008 pliss to be read edited post ye neanderthal Link to comment
The Outsider Posted January 20, 2008 Author Share Posted January 20, 2008 Graphstat is obviously written by an ape in an even lower state of evolution than the hanuman (maybe a chi-mpanzee?). Fisher's is only used in preference to Chi^2 when the individual values are Of the entire sample or the distribution in the table? Just curious. Link to comment
Dhondy Posted January 20, 2008 Share Posted January 20, 2008 Dont blame me hanumanji. Blame graphpad I only input Professor's values - gp was the one which said that Fisher's was better than chi squared - me, i know no difference other than between a nail and a hammer. Anyway, to satisy your simian tastes, here is another banana.... Analyze a 2x2 contingency table win lose Total bucknor present 83 100 183 bucknor absent 133 100 233 Total 216 200 416 Chi-square with Yates correction Chi squared equals 5.186 with 1 degrees of freedom. The two-tailed P value equals 0.0228 The association between rows (groups) and columns (outcomes) is considered to be statistically significant. Same problem. Actual values needed. Link to comment
Dhondy Posted January 20, 2008 Share Posted January 20, 2008 Of the entire sample or the distribution in the table? Just curious. The individual row or column values. Link to comment
Dhondy Posted January 20, 2008 Share Posted January 20, 2008 Same problem. Actual values needed. Actually, bugger that. You can use the Chi squared test of proportions. Link to comment
The Outsider Posted January 20, 2008 Author Share Posted January 20, 2008 Wins with Bucknor : 5 Losses with Bucknor : 6 Wins without Bucknor : 44 Losses without Bucknor : 32 Link to comment
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now