 |
|
10-25-2012, 09:36 PM
|
#16
|
|
Registered User
Join Date: Apr 2006
Age: 27
Posts: 554
|
Re: ELO Ratings
Quote:
Originally Posted by Litotes
What happened to clay rankings? Suddenly average rating for top-100 is several hundred points higher than anywhere else.
|
On clay, the favorites tend to win more matches since the serve has less impact. It's a mathematical property of the points system in tennis.
Last edited by amirbachar : 10-26-2012 at 01:32 PM.
|
|
|
|
Sponsored Links
|
Advertisement
|
|
10-27-2012, 07:34 PM
|
#17
|
|
Welcomed User
Join Date: Mar 2012
Location: Norway
Age: 40
Posts: 8,284
|
Re: ELO Ratings
Quote:
Originally Posted by amirbachar
On clay, the favorites tend to win more matches since the serve has less impact. It's a mathematical property of the points system in tennis.
|
Not really. Nadal and Borg has a higher percentage won on clay than anyone else on any court. But they are the only ones. Federer, for instance, has 83,3% won on hardcourt (530-106) and 77,1% won on clay (178-53). Yet his clay score is much higher than his hardcourt score. Something is amiss here.
|
|
|
10-27-2012, 08:05 PM
|
#18
|
|
Registered User
Join Date: Apr 2006
Age: 27
Posts: 554
|
Re: ELO Ratings
Quote:
Originally Posted by Litotes
Not really. Nadal and Borg has a higher percentage won on clay than anyone else on any court. But they are the only ones. Federer, for instance, has 83,3% won on hardcourt (530-106) and 77,1% won on clay (178-53). Yet his clay score is much higher than his hardcourt score. Something is amiss here.
|
There is no point in comparing Federer's ratings on clay and on hard, as they are on a different scale.
The ratings may only be used to calculate probabilities within the same surface.
|
|
|
10-27-2012, 09:41 PM
|
#19
|
|
Welcomed User
Join Date: Mar 2012
Location: Norway
Age: 40
Posts: 8,284
|
Re: ELO Ratings
Quote:
Originally Posted by amirbachar
There is no point in comparing Federer's ratings on clay and on hard, as they are on a different scale.
The ratings may only be used to calculate probabilities within the same surface.
|
OK, if you're satisfied then I'm not complaining. I'm only trying to be helpful and probably not very good at it. Just compare your current clay ranking with your all time high one and see if there's not potential for improvement somewhere. There is basically noone currently within 600 points of their PB.
|
|
|
10-27-2012, 10:05 PM
|
#20
|
|
Registered User
Join Date: Apr 2006
Age: 27
Posts: 554
|
Re: ELO Ratings
Quote:
Originally Posted by Litotes
OK, if you're satisfied then I'm not complaining. I'm only trying to be helpful and probably not very good at it. Just compare your current clay ranking with your all time high one and see if there's not potential for improvement somewhere. There is basically noone currently within 600 points of their PB.
|
First of all, I really appreciate your help.
About the peak ratings, they were calculated using different method. I guess I should update them, because otherwise, there are not a source for compassion... Hopefully I will do that tomorrow.
|
|
|
10-28-2012, 04:39 PM
|
#21
|
|
Registered User
Join Date: Jun 2010
Age: 22
Posts: 3,326
|
Re: ELO Ratings
Great Work!
Ever thought of doing the same for indoor/outdoor?
The thing I find really tough to digest in this is Borg at 18 on clay! Moreover, he has almost the same rating as DelPotro
|
|
|
10-29-2012, 02:56 PM
|
#22
|
|
Registered User
Join Date: Oct 2012
Age: 51
Posts: 150
|
Re: ELO Ratings
How do you treat matches where one of the players retires during the match? And what about matches with walkovers?
|
|
|
10-29-2012, 09:28 PM
|
#23
|
|
Registered User
Join Date: Apr 2006
Age: 27
Posts: 554
|
Re: ELO Ratings
Quote:
Originally Posted by G.100sic
How do you treat matches where one of the players retires during the match? And what about matches with walkovers?
|
Every set that has started counts, there for ret. in the middle of a set counts as a loss of the set, and walkovers don't count at all.
|
|
|
10-29-2012, 10:06 PM
|
#24
|
|
Registered User
Join Date: Oct 2012
Age: 51
Posts: 150
|
Re: ELO Ratings
Quote:
Originally Posted by amirbachar
Every set that has started counts, there for ret. in the middle of a set counts as a loss of the set, and walkovers don't count at all.
|
Yeah, my idea exactly the same.
|
|
|
11-07-2012, 07:08 AM
|
#25
|
|
Registered User
Join Date: Apr 2010
Age: 27
Posts: 364
|
Re: ELO Ratings
I didn't know the ELO system, I've a couple of questions from ignorance ..
What means the date on the right of the score?
Is there any way to use another factor depending on the round and the tournament to be played? Not the sama a 1R of Halle that a Wimbledon final ..
Thanks and great job
__________________
PATRICK RAFTER
PAW: Rank 24 (23)
Best Results:
2010: W: ROLAND GARROS Q WIMBLEDON
Suicide Tennis: Rank 67 (67)
Best Results:
2010: Q WIMBLEDON F Master 1000 Roma
|
|
|
11-08-2012, 07:52 AM
|
#26
|
|
Registered User
Join Date: Oct 2012
Age: 51
Posts: 150
|
Re: ELO Ratings
Quote:
Originally Posted by Rafter_forever
I didn't know the ELO system, I've a couple of questions from ignorance ..
What means the date on the right of the score?
Is there any way to use another factor depending on the round and the tournament to be played? Not the sama a 1R of Halle that a Wimbledon final ..
Thanks and great job
|
I took the liberty of answering this... 
Since those are peak rank lists, I suppose that the date is the date when the actual peak rating was achieved by that player.
As for the second question...
Elo rating system is all about head-to-head performance. No pressure factors like tournament size and value, or at which stage of the tournament is a match played are taken into account.
And Elo system is named after its inventor, Arpad Elo, a person, so it is not an acronym.
|
|
|
11-08-2012, 08:36 AM
|
#27
|
|
Registered User
Join Date: Apr 2010
Age: 27
Posts: 364
|
Re: ELO Ratings
jaja, Yes, the first question I understood then make it..
Second, yes, I know that, but my question is if changing the original system could be made. For example, changing the K factor
__________________
PATRICK RAFTER
PAW: Rank 24 (23)
Best Results:
2010: W: ROLAND GARROS Q WIMBLEDON
Suicide Tennis: Rank 67 (67)
Best Results:
2010: Q WIMBLEDON F Master 1000 Roma
|
|
|
11-08-2012, 08:38 AM
|
#28
|
|
Welcomed User
Join Date: Mar 2012
Location: Norway
Age: 40
Posts: 8,284
|
Re: ELO Ratings
Quote:
Originally Posted by Rafter_forever
jaja, Yes, the first question I understood then make it..
Second, yes, I know that, but my question is if changing the original system could be made. For example, changing the K factor
|
Easy. In Chess, where the Elo system originated, there are lots of different Ks. No problem whatsoever, if you want this. OP obviously doesn't.
|
|
|
11-08-2012, 02:17 PM
|
#29
|
|
Registered User
Join Date: Oct 2012
Age: 51
Posts: 150
|
Re: ELO Ratings
Quote:
Originally Posted by Rafter_forever
jaja, Yes, the first question I understood then make it..
Second, yes, I know that, but my question is if changing the original system could be made. For example, changing the K factor
|
Of course that it is possible, but why, what's the point doing that in Elo system?!
Official ranking system already does that, and this is the other way to measure performance.
And as amirbachar pointed out, this system gives better predictability of head-to-head accounts.
|
|
|
11-09-2012, 12:10 AM
|
#30
|
|
Registered User
Join Date: Apr 2010
Age: 27
Posts: 364
|
Re: ELO Ratings
Quote:
Originally Posted by Litotes
Easy. In Chess, where the Elo system originated, there are lots of different Ks. No problem whatsoever, if you want this. OP obviously doesn't.
|
Thanks, the issue is that should recalculate everything, but would be cool, there are limits to the value of K?
Quote:
Originally Posted by G.100sic
Of course that it is possible, but why, what's the point doing that in Elo system?!
Official ranking system already does that, and this is the other way to measure performance.
And as amirbachar pointed out, this system gives better predictability of head-to-head accounts.
|
Easy, I don´t know much about chees, but tennis not is like other sport like boxing for example, because many times, the top players use some tournamentes as preparation for Grand Slam on M1000, and many times they lose matchs that they would not lose in other tournaments..
__________________
PATRICK RAFTER
PAW: Rank 24 (23)
Best Results:
2010: W: ROLAND GARROS Q WIMBLEDON
Suicide Tennis: Rank 67 (67)
Best Results:
2010: Q WIMBLEDON F Master 1000 Roma
|
|
|
 |
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
|