is there a maximum theoretical elo • page 1/3 • General Chess Discussion • lichess.org

if God were playing chess, would his ELO be bounded?
He would never lose, but against a similar being would draw, depending on whether chess is a theoretical win/draw/loss. If we assume that perfect play leads to draw then God would never lose but sometimes draw, and would therefore have a bounded ELO, as iiuc a given win percentage against a given opponent, directly translates to a fixed ELO difference . If chess is a win (e.g. for white) then even here there may be a theoretical limit since gods would either win or lose to each other 50/50 (depending on whether they have to play white/black) , and would have better odds against mortals (even tho they might lose occasionally to mortals who happened to play a perfect game as white).

Does a maximum elo obtain, and if so what is it?

if God were playing chess, would his ELO be bounded? He would never lose, but against a similar being would draw, depending on whether chess is a theoretical win/draw/loss. If we assume that perfect play leads to draw then God would never lose but sometimes draw, and would therefore have a bounded ELO, as iiuc a given win percentage against a given opponent, directly translates to a fixed ELO difference . If chess is a win (e.g. for white) then even here there may be a theoretical limit since gods would either win or lose to each other 50/50 (depending on whether they have to play white/black) , and would have better odds against mortals (even tho they might lose occasionally to mortals who happened to play a perfect game as white). Does a maximum elo obtain, and if so what is it?

applebanana

#2

If we go by FIDE rules where maximum change in rating is 10 after 2400, then maximum rating possible seems to be about 500 points greater than the next best player after which you get no rating for winning. I have no idea what I am talking about :) I am merely parroting what I observed from tinkering with different numbers in here
http://www.3dkingdoms.com/chess/elo.htm.

If we go by FIDE rules where maximum change in rating is 10 after 2400, then maximum rating possible seems to be about 500 points greater than the next best player after which you get no rating for winning. I have no idea what I am talking about :) I am merely parroting what I observed from tinkering with different numbers in here http://www.3dkingdoms.com/chess/elo.htm.

sheckley666

#3

Don't you get at least one point for a win, regardless how low your opponent was rated?

applebanana

#4

@sheckley666 nope. You can see that by checking with the link i gave

Toadofsky

#5

An upper bound can only be established if a player has drawn or lost a game.

jeremyrutman2

#6

if one could determine a fraction of perfect games that e.g. Carlsen can play against God, then it would be possible to relate God's rating to Carlsen's. Lacking this I don't see to how to find it. One thing that jumps to mind is, as a player approaches perfection the loss fraction should go down even if the win fraction doesnt rise , (ie more draws and fewer wins/losses assuming play against approximately equal opponents). This I imagine is what you'd see from the stats for players as func. of rating .

If it helps : the elo expected score is :
Eb=1/(1+10^((Ra-Rb)/400) for players of rating Rb, Ra.

and from there (following the wikipedia on elo )

Ea/Eb=10^(Ra/400) / 10^(Rb/400)

if one could determine a fraction of perfect games that e.g. Carlsen can play against God, then it would be possible to relate God's rating to Carlsen's. Lacking this I don't see to how to find it. One thing that jumps to mind is, as a player approaches perfection the loss fraction should go down even if the win fraction doesnt rise , (ie more draws and fewer wins/losses assuming play against approximately equal opponents). This I imagine is what you'd see from the stats for players as func. of rating . If it helps : the elo expected score is : Eb=1/(1+10^((Ra-Rb)/400) for players of rating Rb, Ra. and from there (following the wikipedia on elo ) Ea/Eb=10^(Ra/400) / 10^(Rb/400)

VertSangBleu

#7

if god played chess he would be crushed. i mean, his creations are far from being optimal so he is pretty bad at what he does. the creation is a mess. he would be destroyed by children

Interesting...

2000 points difference is roughly 1 win in 1 million games. Say a normal person can at best play 100 K games in a lifetime. So 1900 difference is a win in 500 K. 1800 difference is a win in 250 K. 1700 difference is a win in 125 K.

Say Stockfish is 3400 and AlfaZero is 200 points stronger. Then the hipotetical living being that could beat AlfaZero in each and every game while having a near human lifespan would be 3400+200+1700 = 5300 rating points. And I don't think that ratings above that score would actually mean anything.

2000 points difference is roughly 1 win in 1 million games. Say a normal person can at best play 100 K games in a lifetime. So 1900 difference is a win in 500 K. 1800 difference is a win in 250 K. 1700 difference is a win in 125 K. Say Stockfish is 3400 and AlfaZero is 200 points stronger. Then the hipotetical living being that could beat AlfaZero in each and every game while having a near human lifespan would be 3400+200+1700 = 5300 rating points. And I don't think that ratings above that score would actually mean anything.

Sneakmasterflex

#10

First of all the OP is assuming a. that god exists b. that he/she plays chess, for both assumptions there is no proof whatsoever. Secondly, a Fischer in his prime would probably beat god at chess. That is easier to assume since we actually have proof that Fischer exists/existed(depending on your views of a potential afterlife. Thirdly, chessplayers knew for a long time, there is only one Goddess of Chess, and her name is Caissa. Her name is Caissa. Her name is Caissa. Her na.....