Possible highest rated players of all time list.

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Kurt Meyer
Posts: 18
Joined: Mon Jun 19, 2017 2:37 pm

Re: Possible highest rated players of all time list.

Post by Kurt Meyer » Wed Jun 28, 2017 4:09 pm

Agree: It deserves an Ig Nobel prize...

:roll:

Branko Radovanovic
Posts: 65
Joined: Sat Sep 13, 2014 2:12 pm

Re: Possible highest rated players of all time list.

Post by Branko Radovanovic » Wed Jun 28, 2017 4:37 pm

Leo wrote:Karpov defeated Annand in a Championship match so there is a discrepancy right there.
There is no discrepancy, actually. The quote says: "These models can then be used to compute the win/draw/lose probability for any given match between two players." So, all outcomes of a match are possible - including that of a weaker player winning - they just have different probabilities.

BTW, Kenneth W. Regan used to argue that, contrary to what many believe, Elo ratings of modern players are not inflated (https://www.cse.buffalo.edu/~regan/pape ... 12IPRs.pdf). Looking at Alliot's rankings, though, it seems that even if modern-day masters do generally play better than their predecessors, their ratings are inflated to some extent (e.g. Kasparov being ranked lower than Fischer).

Leo
Posts: 921
Joined: Fri Sep 16, 2016 4:55 pm
Location: USA/Minnesota
Full name: Leo Anger

Re: Possible highest rated players of all time list.

Post by Leo » Wed Jun 28, 2017 6:22 pm

Branko Radovanovic wrote:
Leo wrote:Karpov defeated Annand in a Championship match so there is a discrepancy right there.
There is no discrepancy, actually. The quote says: "These models can then be used to compute the win/draw/lose probability for any given match between two players." So, all outcomes of a match are possible - including that of a weaker player winning - they just have different probabilities.

BTW, Kenneth W. Regan used to argue that, contrary to what many believe, Elo ratings of modern players are not inflated (https://www.cse.buffalo.edu/~regan/pape ... 12IPRs.pdf). Looking at Alliot's rankings, though, it seems that even if modern-day masters do generally play better than their predecessors, their ratings are inflated to some extent (e.g. Kasparov being ranked lower than Fischer).
I guess I meant to say the ratings here are not etched in stone be any means.
Advanced Micro Devices fan.

Leo
Posts: 921
Joined: Fri Sep 16, 2016 4:55 pm
Location: USA/Minnesota
Full name: Leo Anger

Re: Possible highest rated players of all time list.

Post by Leo » Wed Jun 28, 2017 6:24 pm

Kurt Meyer wrote:Agree: It deserves an Ig Nobel prize...

:roll:

What does IG mean?
Advanced Micro Devices fan.

Leo
Posts: 921
Joined: Fri Sep 16, 2016 4:55 pm
Location: USA/Minnesota
Full name: Leo Anger

Re: Possible highest rated players of all time list.

Post by Leo » Wed Jun 28, 2017 6:35 pm

Branko Radovanovic wrote:
Leo wrote:Karpov defeated Annand in a Championship match so there is a discrepancy right there.
There is no discrepancy, actually. The quote says: "These models can then be used to compute the win/draw/lose probability for any given match between two players." So, all outcomes of a match are possible - including that of a weaker player winning - they just have different probabilities.

BTW, Kenneth W. Regan used to argue that, contrary to what many believe, Elo ratings of modern players are not inflated (https://www.cse.buffalo.edu/~regan/pape ... 12IPRs.pdf). Looking at Alliot's rankings, though, it seems that even if modern-day masters do generally play better than their predecessors, their ratings are inflated to some extent (e.g. Kasparov being ranked lower than Fischer).
Very interesting paper.
Advanced Micro Devices fan.

Dirt
Posts: 2851
Joined: Wed Mar 08, 2006 9:01 pm
Location: Irvine, CA, USA

Re: Possible highest rated players of all time list.

Post by Dirt » Wed Jun 28, 2017 6:54 pm

Leo wrote:
Kurt Meyer wrote:Agree: It deserves an Ig Nobel prize...

:roll:

What does IG mean?
It's just word play. Ignoble means not noble.
Deasil is the right way to go.

Kurt Meyer
Posts: 18
Joined: Mon Jun 19, 2017 2:37 pm

Re: Possible highest rated players of all time list.

Post by Kurt Meyer » Wed Jun 28, 2017 7:15 pm

About Ig Nobel prize

http://www.improbable.com/ig/

Leo
Posts: 921
Joined: Fri Sep 16, 2016 4:55 pm
Location: USA/Minnesota
Full name: Leo Anger

Re: Possible highest rated players of all time list.

Post by Leo » Wed Jun 28, 2017 7:24 pm

Dirt wrote:
Leo wrote:
Kurt Meyer wrote:Agree: It deserves an Ig Nobel prize...

:roll:

What does IG mean?
It's just word play. Ignoble means not noble.
I get it.
Advanced Micro Devices fan.

MikeGL
Posts: 931
Joined: Thu Sep 01, 2011 12:49 pm

Re: Possible highest rated players of all time list.

Post by MikeGL » Thu Jun 29, 2017 2:34 am

Leo wrote:1 Carlsen
2 Kramnik
3 Fischer
4 Kasparov
5 Anand
6 Khalifman
7 Smyslov
8 Petrosian
9 Karpov
10 Kasimdzhanov
11 Botvinnik
12 Ponomariov
13 Lasker
14 Spassky
It is a very good and very thorough paper, but there are also weaknesses on this. Stockfish is very strong and would flag a repetition move as very bad and a blunder if the winning move is not played outright. There are position where a player needs to improve his clock by repeating moves thereby nearing the 40-move limit for a new time control and if you repeat position twice, SF would flag it as blunder even though the strong/optimum move is still available and will be played later after the repetition.

You will notice this if you watch live games at chessbomb where inferior moves are colored red by stockfish, and all repetition moves are always colored red even if a player has already seen the strongest optimal move.

I am sure this was not taken into consideration on that study. I was listening to some tutorial of Karpov and he was also playing some repetitive moves just to gain time, and those moves will be considered a blunder by SF8 when, in truth, it is not.

The weakness is evident when the World Champion Capablanca who has the fewest professional loses in history of chess is not listed on top 14. Maybe partly because, there's another weakness. Capablanca would simplify into a winning endgame with 1 pawn up rather than play accurately the complex lines in a winning position. Simplifying might not be the best, like choosing a +6.00 complex midgame winning line versus eval of +2.00 endgame with 1 pawn up would be considered wrong by SF.

corres
Posts: 2903
Joined: Wed Nov 18, 2015 10:41 am
Location: hungary

Re: Possible highest rated players of all time list.

Post by corres » Thu Jun 29, 2017 8:05 am

All chess engines have only one strong side: chess tactics. So to make comparisons between chess player with the help of a chess engine (Stockfish or not Stockfish) gives great advantage to chess players playing tactically and cause great drawback to chess players playing strategically.
Moreover without endgame database chess engines play endgames very irregular manner. To recite the endgame knowledge of a chess engine from Capablanca or other endgame experts is a stupid thing.
I think this article is only a play with statistics and Stockfish.

Post Reply