View Single Post
post #44 of (permalink) Old 03-04-2012, 05:14 PM
country flag duong
Registered User
Join Date: Jun 2008
Posts: 12,122

This OP about Wimbledon has some interesting and unexpected results fitting with some which other people have noted thanks to the newly published on a long period ATP Ricoh match stats.

I think of that thread : Stats prove that it was easier to return in the 90's than in the 00's

I think of the lower % of aces in 1994, also the much lower % of points won on second serve and that the best players on second serve were not actually Sampras as some think but Agassi and Courier.

I think that apart from Wimbledon, those ATP Ricoh stats should be exploited more by statisticians about that.

And further analyses should be generally done about those topics taking into account the many evolutions which occurred other than just the "speed" of the court (which in Wimbledon is rather about the bounce imo, highness as well as unpredictability, also in Wimbledon it depends first week or second week, maybe Nyctennisfan might make some interesting stats comparing first week and second week's stats in Wimbledon in both those eras, including for some specifi players like Nadal for instance)

Although sometimes these Ricoh stats are wrong : for instance I noted that the Ricoh stats for 2003 on return are clearly wrong when you compare them with the stats from 2002 and 2004, the results are just impossible then clearly their computer programs or their data had one problem in 2003.

PS : I'm a professional statistician myself but I'm not in favour of putting too much attention on significance tests, esp when they are published. It just makes it look too complicated for people, and here you could say that all of the results or nearly all of them are statistically significant (or you could just not publish the ones where the difference is not statistically significant).
More generally, I think your work was great, Nyctennisfan, but you paid too little attention to the way you published and presented the results, which makes the thread look too complicated for some people, and also the main results should be put more in evidence, and the results and analyses should be more separated from the method.
But I understand that when we do this kind of work for passion as I sometimes do myself, we think more for ourselves than for the others then we spend less time and attention about the way we present the results

useless old guy

Last edited by duong; 03-04-2012 at 05:28 PM.
duong is offline  
For the best viewing experience please update your browser to Google Chrome