Comparing historical ratings

Sort:
Heinkel111

The chess rating system seems to be a good way to rank contemporary players against each other.

How can we compare players from history against players of today?

When I hear that Magnus Carlsen is the best player ever because he has achieved the highest rating it does not really make sense since the rating system only measures relative strength against contemporaries, not absolute strength.

In order to determine who are the best players from all of history would it not be possible to simply run the engine on all their moves from all their games and see who gets closest to the best engine move most often?

Players from across history could then be ranked against each other.

I would expect that chess playing skill levels generally improve over time since each generation has a greater database of opening theory etc than the last generation to study.

So even an engine-measured league table would be unfairly skewed towards contemporary players.

I would also expect that skill levels have increased rapidly since the advent of internet and engines.

Internet makes it easy to increase the amount of chess one plays (and practise makes perfect) since getting a game is much easier.

Engines make good quality post game analysis available to every player where in the past it was only available to proteges with a mentor.

Or maybe absolute rankings are meaningless since

1. Chess can be as much an art as a sport at the top level

2. Engines are too mechanical a tool to measure artists

Maybe this subject has already been done but I could not find the topic.

notmtwain
Heinkel111 wrote:

The chess rating system seems to be a good way to rank contemporary players against each other.

How can we compare players from history against players of today?

When I hear that Magnus Carlsen is the best player ever because he has achieved the highest rating it does not really make sense since the rating system only measures relative strength against contemporaries, not absolute strength.

In order to determine who are the best players from all of history would it not be possible to simply run the engine on all their moves from all their games and see who gets closest to the best engine move most often?

Players from across history could then be ranked against each other.

I would expect that chess playing skill levels generally improve over time since each generation has a greater database of opening theory etc than the last generation to study.

So even an engine-measured league table would be unfairly skewed towards contemporary players.

I would also expect that skill levels have increased rapidly since the advent of internet and engines.

Internet makes it easy to increase the amount of chess one plays (and practise makes perfect) since getting a game is much easier.

Engines make good quality post game analysis available to every player where in the past it was only available to proteges with a mentor.

Or maybe absolute rankings are meaningless since

1. Chess can be as much an art as a sport at the top level

2. Engines are too mechanical a tool to measure artists

Maybe this subject has already been done but I could not find the topic.

There are literally dozens of chess and other websites that have tackled this question.

Among them are:

https://en.chessbase.com/post/historical-chess-ratings-dynamically-presented

https//en.m.wikipedia.org/wiki/Comparison_of_top_chess_players_throughout_history

It has also been covered multiple times here on chess.com.

https://www.chess.com/forum/view/chess-players/strongest-players

 https://www.chess.com/forum/view/general/who-was-the-greatest-chess-player-of-all-time

 

kindaspongey

Carlsen-2961 Kramnik-2868 Kasparov-2816 Fischer-2775 Anand-2759 Karpov-2698 Capablanca-2664 Tal-2636 Spassky-2619 Smyslov-2618 Botvinnik-2602 Euwe-2547 Alekhine-2547 Petrosian-2543 Lasker-2498 Morphy-2409 Steinitz-2323

https://www.chess.com/article/view/who-was-the-best-world-chess-champion-in-history

https://www.chess.com/article/view/should-we-trust-computers

https://www.chess.com/article/view/were-players-in-the-1800s-terrible

https://www.chess.com/article/view/the-5-most-dangerous-chess-players-ever

https://www.chess.com/article/view/adolf-anderssen-mr-slice-and-dice

https://www.chess.com/article/view/adolf-anderssen-more-slicing-and-dicing

https://www.chess.com/article/view/behold-steinitz-the-austrian-morphy

DragonPhoenixSlayer

Engines cant accuratally measure the strenght of a chess player. I played a game earlier today according to the computer is better than a few games I analyzed by Garry Kasparov. Does that mean that I'm better than Kasparov? Does that mean that the game I played was better than the ones played by Kasparov? No. Computers can't tell the difference between a 15 move tactic that is impossible for a human to find and just simply hanging a piece. They don't know whats a position thats easy to play and one thats hard to play.

Chesserroo2

Finding the best move in some positions is harder than in others. What if your contemporary likes to drag you into complex positions? Then the computer ranks you lower than the other safer players. As for the opening, you could compare moves after move 10, though memorizing all those moves counts for something if they can refute the old ones.

Chesserroo2

DragonPhoenixSlayer wrote:

Engines cant accuratally measure the strenght of a chess player. I played a game earlier today according to the computer is better than a few games I analyzed by Garry Kasparov. Does that mean that I'm better than Kasparov? Does that mean that the game I played was better than the ones played by Kasparov? No. Computers can't tell the difference between a 15 move tactic that is impossible for a human to find and just simply hanging a piece. They don't know whats a position thats easy to play and one thats hard to play.

Agreed. Good moves make threats and make it difficult for your opponent to find the best response. Computers often don't simplify a winning position by trading down. Humans trade down because we are afraid we will make a mistake later and just want a position we can't mess up. Computers calculate that keeping the extra material on the board let's them win a bit faster.

Caesar49bc

It's hard to really compare players from different eras outside of the hypothetical "but what if the world chess champion from [insert year]  was 24 years old today?".

BeamingEel

HELLO

BeamingEel

how are you