tag:blogger.com,1999:blog-22713811.post2312819027636704908..comments2023-05-18T10:02:56.564+02:00Comments on Pappus' plane - cricket stats: Opening partnerships, and a Kiwi recordDavid Barryhttp://www.blogger.com/profile/08378763233797445502noreply@blogger.comBlogger13125tag:blogger.com,1999:blog-22713811.post-83318001215630300442008-01-24T20:34:00.000+01:002008-01-24T20:34:00.000+01:00That's an interesting idea, Samir. It shouldn't b...That's an interesting idea, Samir. It shouldn't be too hard to calculate, so I'll write a post on it with the results either later tonight or tomorrow.David Barryhttps://www.blogger.com/profile/08378763233797445502noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-36979268373413791662008-01-24T20:21:00.000+01:002008-01-24T20:21:00.000+01:00Sorry, that should be 1.75 above.Sorry, that should be 1.75 above.Samir Choprahttps://www.blogger.com/profile/16145425333818168212noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-388978106900541212008-01-24T20:20:00.000+01:002008-01-24T20:20:00.000+01:00David,I don't know if you have the archives at han...David,<BR/><BR/>I don't know if you have the archives at hand in the correct format for this, but I'm very interested in trying to derive a statistic called "the-averaged-fall-of-wicket-quotient" for openers. Heres how it works. One of the two opening batsmen is obviously the first wicket to fall in any innings. His partner carries on till he too, is dismissed (in the limiting case, he carries his bat). Now, assign a number to opening batsmen corresponding to the fall of wickets. If you are the first to be dismissed, you get a 1. If you are the second, you get a 2, and so on. Add these up over all innings and divide by the total number of innings. So if a batsman is dismissed in four innings as being the first, second, first and third wickets to fall, his quotient is 1.5. It gives some indication of how deep the opening batsman lands up batting into each innings. I'm not quite sure of how to handle not-outs and instances of carrying the bat. The latter must be assigned some suitably high value, and the former needs to note how many other wickets fell in the same innings (and perhaps the result). <BR/><BR/>Just an idea.Samir Choprahttps://www.blogger.com/profile/16145425333818168212noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-4475025559186632812008-01-23T17:48:00.000+01:002008-01-23T17:48:00.000+01:00Hi Scorpi. Hobbs and Sutcliffe have easily the be...Hi Scorpi. Hobbs and Sutcliffe have easily the best average opening partnership of any regular opening pair (87,8), but because one averaged over 55 and the other over 60, you'd expect them to have a very high average opening partnership! Franklin and Wright overall weren't as good - average partnership 55 - but Franklin only averaged 23, so their average partnership is really quite remarkable.<BR/><BR/>I did my undergrad in maths and physics. Cricket statistics is just a hobby for me.David Barryhttps://www.blogger.com/profile/08378763233797445502noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-17493234904422689462008-01-23T16:57:00.000+01:002008-01-23T16:57:00.000+01:00hmm... this is totally mind-boggling crazy... math...hmm... this is totally mind-boggling crazy... math was a nightmare for me :)... I can't believe that Hobbs & Sutcliffe didn't end up on the top... very interesting. Are you a mathematician by profession or undertake anything related to the cricket world statistics?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-22713811.post-6580228648327865302008-01-20T13:00:00.000+01:002008-01-20T13:00:00.000+01:00No, I didn't exclude innings when they batted toge...No, I didn't exclude innings when they batted together when calculating individual averages. That's not really relevant to what I want to look at here, which is the average size of the partnership compared to the individual averages.<BR/><BR/>Now that I think about it, it might be relevant to consider the individual averages ONLY when batting with their particular partner. I'm happy with what I've done though, since the overall opening average gives a better reflection on how good the opener was.David Barryhttps://www.blogger.com/profile/08378763233797445502noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-57234595079323616362008-01-20T12:43:00.000+01:002008-01-20T12:43:00.000+01:00To be picky - while you hve use average as opener,...To be picky - while you hve use average as opener, did you exclude the partnership in question itself?<BR/><BR/>In case I worded that confusingly, what I meant is :<BR/><BR/>To predict the average partnership of, say, Langer and Hayden, based on their individual averages as openers when _not_ opening together.Optimistixhttps://www.blogger.com/profile/17927346995555238063noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-512850185742128252008-01-19T19:29:00.000+01:002008-01-19T19:29:00.000+01:00Greenidge-Haynes: 0.98, pretty much bang in the mi...Greenidge-Haynes: 0.98, pretty much bang in the middle, 49th.<BR/><BR/>Anwar-Sohail: 0.94, 59th.<BR/><BR/>Taylor-Slater: 1.10, 28th.<BR/><BR/>Marsh-Boon: 1.08, 33rd.David Barryhttps://www.blogger.com/profile/08378763233797445502noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-28354955761261350852008-01-19T18:56:00.000+01:002008-01-19T18:56:00.000+01:00Very interesting. Where do Greenidge-Haynes, and S...Very interesting. Where do Greenidge-Haynes, and Saeed Anwar - Aamir Sohail come in? How about Mark Taylor and Slater? They had some huge partnerships? So did Marsh and Boon, for that matter.<BR/><BR/>I wonder i fChauhan really could have been slower than Gavaskar :-)Optimistixhttps://www.blogger.com/profile/17927346995555238063noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-2133368511498634482008-01-17T20:14:00.000+01:002008-01-17T20:14:00.000+01:00That's a good question. Unfortunately I don't hav...That's a good question. Unfortunately I don't have any formal stats training (for the first time ever, I'm regretting not doing any stats courses in undergrad!), and this leaves me with some basic gaps in my knowledge of the subject.<BR/><BR/>I think to do a significance test I'd need to know the distribution of opening partnerships. It's not (I think) a simple case of saying that ratios above 1.2 are significant, because the cut-off value for, say, p = 0.05 will vary with the number of partnerships. (eg, a pair who bats once together and averages 200 will have a very high ratio, but this won't be statistically significant, because it's a one-off.)<BR/><BR/>I might be able to model this numerically and come up with some p-values, but off the top of my head there are some finicky details that I don't know how to solve....<BR/><BR/>Given how many times Chauhan and Gavaskar batted together, I'd guess that their ratio is statistically significant, but showing that seems to be a bit too hard for me.David Barryhttps://www.blogger.com/profile/08378763233797445502noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-31501668754333170462008-01-17T09:30:00.000+01:002008-01-17T09:30:00.000+01:00Thanks David.Is there a way to apply a test of sig...Thanks David.<BR/><BR/>Is there a way to apply a test of significance to the better ( or worse) figure?<BR/><BR/>I'm a bit silly with stats and my last big involvement with statistical analysis (doing it all by myself) was when I wrote my thesis some 20+ years ago! I used to apply a test of significance...some P values and t and the like...(am I on the right track?)...to see if the observed difference is significant or not.<BR/><BR/>And thank you for your views at TCWJ. I truly appreciate it.Soulberryhttps://www.blogger.com/profile/15830619858224129215noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-49537987373141285512008-01-14T21:20:00.000+01:002008-01-14T21:20:00.000+01:00Hey Soulberry. Chauhan and Gavaskar did significa...Hey Soulberry. Chauhan and Gavaskar did significantly better than you would expect, with a ratio of 1.14. Chauhan's average was less than 32, so to have an average opening stand with Gavaskar of over 50 was quite good.<BR/><BR/>If you rank the opening pairs by the ratio, they come 25th out of the 97.<BR/><BR/>Actually it could be a little bit higher, up around 22nd: for some reason my average partnership for them disagrees with that Cricinfo list (same number of runs, but I have their average as 52.81 whereas Cricinfo has 53.75). Perhaps there's a retired hurt affecting the numbers somewhere.<BR/><BR/>Either way, they did better than you would have expected.David Barryhttps://www.blogger.com/profile/08378763233797445502noreply@blogger.comtag:blogger.com,1999:blog-22713811.post-6037604846424242892008-01-14T21:01:00.000+01:002008-01-14T21:01:00.000+01:00How would Gavaskar-Chauhan partnership in tests ha...How would <A HREF="http://content-usa.cricinfo.com/ausvind/content/story/242788.html" REL="nofollow">Gavaskar-Chauhan partnership in tests</A> have fared according to the calculations? As expected, better, or worse?<BR/><BR/>Just curious, and a poor one with stats at that. Thanks.Soulberryhttps://www.blogger.com/profile/15830619858224129215noreply@blogger.com