Latest Updates:
Page Index Toggle Pages: [1] 2 3 
Topic Tools
Very Hot Topic (More than 25 Replies) Modern Engines and Classic Games (Read 9846 times)
Heuristic
Full Member
***
Offline


I Love ChessPublishing!

Posts: 103
Joined: 05/14/20
Re: Modern Engines and Classic Games
Reply #36 - 05/18/20 at 19:17:38
Post Tools
TonyRo wrote on 04/20/20 at 12:44:42:
I remember quite a few years back I worked my way through Larsen's Best Games on Forward Chess for the 3rd or 4th time. I was pretty blown away by how accurate his play and analysis was, even all the way back to the 60s and 70s. I think sometimes people are quick to dismiss the older generations when talking about how good players are now, but those guys were no joke at all.


Tthose players in the 60s and early 70s were phenomenally strong - Spassky, Petrosian, Larsen, Taimanov etc. A lot stronger than they were given credit for. Karpovs following era was actually much weaker even though he faced several of those guys passed their prime.
  
Back to top
 
IP Logged
 
trw
YaBB Moderator
*****
Offline


I Love ChessPublishing!

Posts: 1414
Joined: 05/06/08
Gender: Male
Re: Modern Engines and Classic Games
Reply #35 - 05/18/20 at 17:25:38
Post Tools
cathexis wrote on 04/19/20 at 16:30:12:
Greetings!

Has anyone ever tried running the most brilliant & classic games of the past through modern chess engines to see how they'd fare againest Stockfish, etc.?

Major quibble: Of course if ALL the moves aren't the same then you could say it's no longer the same game. True. And if you allow the engine to play whatever then it certainly would change the game into lines of its preference, especially if you task the engine with the losing side. But it would be interesting to load mid-game positions and see what the engines thought of the moves the great masters of the past had chosen, especially in the absence of regrettable and obvious blunders. I suppose also I could just load these games myself. But I'm curious if there's ever been a more "formal" study of this and what were the results?

Stay safe, be hopeful, and thanks in advance,

Cathexis



My friend did this with every world champion and dumped the results into a SQL database... it was truly fascinating to see the error rates etc. He was using to establish a definition of a blunder ie how much centipawns change is required for a WC to go from winning to drawn (defined as ? in his database) or winning to losing (defined in his database as ??).
  
Back to top
 
IP Logged
 
VGA
Junior Member
**
Offline


I Love ChessPublishing!

Posts: 97
Joined: 08/13/17
Re: Modern Engines and Classic Games
Reply #34 - 05/18/20 at 02:05:14
Post Tools
Super-GMs of the modern era play more and better opening theory moves, that increases the accuracy significantly.  They also play the opening stage of the game more quickly and have more time remaining for the rest of the game.

One of my wishes when it comes to chess it to see the classical time controls reduced by about 20% to account for this. It should make for more decisive games, too.
  
Back to top
 
IP Logged
 
cathexis
God Member
*****
Offline


No matter where you go,
there you are.

Posts: 608
Joined: 03/03/20
Re: Modern Engines and Classic Games
Reply #33 - 05/17/20 at 13:36:52
Post Tools
Kidding a bit, but it's tempting to consider whether Mr. Carlsen is rated the most accurate by machine evaluation standards because he is the most like a machine himself? Or because, of all the candidates on that list he has had the greatest percentage of his total years to work with said engines? You know, gaming the engine. Roman did a video on that. Looks old though. And please! No flaming, I am halfway just kidding. No 'dis to the WC.

https://www.youtube.com/watch?v=p4x_iz-nu_c
  
Back to top
 
IP Logged
 
Heuristic
Full Member
***
Offline


I Love ChessPublishing!

Posts: 103
Joined: 05/14/20
Re: Modern Engines and Classic Games
Reply #32 - 05/17/20 at 09:28:02
Post Tools
cathexis wrote on 04/19/20 at 16:30:12:
Greetings!

Has anyone ever tried running the most brilliant & classic games of the past through modern chess engines to see how they'd fare againest Stockfish, etc.?

Major quibble: Of course if ALL the moves aren't the same then you could say it's no longer the same game. True. And if you allow the engine to play whatever then it certainly would change the game into lines of its preference, especially if you task the engine with the losing side. But it would be interesting to load mid-game positions and see what the engines thought of the moves the great masters of the past had chosen, especially in the absence of regrettable and obvious blunders. I suppose also I could just load these games myself. But I'm curious if there's ever been a more "formal" study of this and what were the results?

Stay safe, be hopeful, and thanks in advance,

Cathexis


It has been done and it was found that the Magnus Carlsen was the most accurate WC followed by Vladimir Kramnik and Gary Kasparov.

https://www.chess.com/article/view/who-was-the-best-world-chess-champion-in-hist...
  
Back to top
 
IP Logged
 
cathexis
God Member
*****
Offline


No matter where you go,
there you are.

Posts: 608
Joined: 03/03/20
Re: Modern Engines and Classic Games
Reply #31 - 04/28/20 at 22:02:40
Post Tools
My words may have been inadequate due to lack of experience but the response was awesome - and enlightening!

My only whine would be that the path that Rene so beautifully outlined can seem rather vague and far off in terms of exactly how one gets from here to there. But I don't fault anyone for that.

My Thanks!
  
Back to top
 
IP Logged
 
Stigma
God Member
*****
Offline


There is a crack in everything.

Posts: 3265
Joined: 11/07/06
Gender: Male
Re: Modern Engines and Classic Games
Reply #30 - 04/28/20 at 19:36:27
Post Tools
ReneDescartes wrote on 04/28/20 at 17:46:35:
Juicing is taking illegal hormones and performance-enhancing drugs.

Just a slang term for doping, then. Thanks.

Comparing that to opening preparation seems far-fetched. One is illegal (though there's no telling how many do it anyway and get away with it), while the other has been an integral part of chess since the first writings on the game. The closer equivalent of juicing in chess would be cheating with engines during play.

Though the openings arms race today can become overwhelming. You'd think the game should really be more about two minds competing at the board than about who has burnt the most hours and computing power on homework. So while I enjoy my favorite openings, I also welcome the slowly increasing interest in Fischer-Random (AKA Chess960) as a fresh alternative.
  

Improvement begins at the edge of your comfort zone. -Jonathan Rowson
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1236
Joined: 05/17/10
Gender: Male
Re: Modern Engines and Classic Games
Reply #29 - 04/28/20 at 17:46:35
Post Tools
Juicing is taking illegal hormones and performance-enhancing drugs. What AOC just said is right, but advanced.

The practice in chess is called opening preparation, and it has been important for centuries. This entire website and enterprise exists only because of it. It is one of the fine traditions enriching chess culture, not a compromising or ethically dubious practice. It is important not to understand it too crudely. Memorization is involved, but it is not as simple as that.

It isn't a matter of extending one's thinking capacity beyond seven moves. Players such as Pillsbury, Reti, Nimzovich, and Botvinnik did opening research to develop and refine whole themes and strategies previously unknown in chess. Ceding White a classical center and kingside advantage in return for a huge mobile pawn mass on the queenside (many lines of the Slav, such as the Tolush gambit, Botvinnik variation, etc.); allowing White to create a massive center only in order to destroy it later and emerge in the cleared position with better-placed pieces (the Gruenfeld); not castling at all when the center is blocked (e.g., the Leningrad Nimzo-Indian); deliberately incurring weaknesses in the belief that the opponent won't be able to spare the moves to exploit them (the Sveshnikov Sicilian): these are some of the fruits of this research. The names of the openings record the rich history of such creative opening research. The main creative ideas endure, although the exact memorized sequences and subordinate ideas evolve over time.

When you study opening theory at home, you must absorb these historical acquisitions while you memorize the moves. Now, AOC was referring to doing additional "home preparation"--doing original investigations and creating additional little pieces of theory of one's own--but you must understand that such 'home-cooked" moves are just little additional pieces of icing on top of a huge memorized historical opening cake. In order to use the icing, first you must own the cake. Accumulating such pieces is the way the mass of opening theory evolved, but each innovation only makes sense against the whole previous background.

Take it all with a grain of salt. Some of the greatest players in history -- Lasker, Capablanca, Carlsen -- dominated even though many of their opponents knew more theory than they did. Traps aside, both opening memorization as a whole and original home preparation only help players who (1)understand with what strategies they should follow up the memorized moves at every point should the opponent jump out of theory (the "ideas behind the openings"), (2) are strong enough to convert into a win the slight advantages they gain with their memorized sequences, and (3) are fairly equal in strength with their opponents. The advantages in question are generally worth less than a pawn and don't exist at all if, once  the opponent leaves theory, one doesn't follow the strategies implied by the moves. However, the basis of these strategies--a great part of their content, to one who knows nothing of them--is the general body of valid opening principles, and indeed chess principles as a whole. If you want to understand any of this stuff, home-prepared or not, you need to learn these principles.

The book which you got, by the way, gives the whole package. It explains--in increasing order of importance--the moves of 1.e4 e5 openings, the associated strategic and tactical ideas, and correct general opening principles. When you read it you will personally be reaping the fruits of hundreds of years of home preparation. Enjoy it with a glass of juice and a good conscience.
« Last Edit: 04/29/20 at 17:12:53 by ReneDescartes »  
Back to top
 
IP Logged
 
an ordinary chessplayer
God Member
*****
Offline


I used to be not bad.

Posts: 1672
Location: Columbus, OH (USA)
Joined: 01/02/15
Re: Modern Engines and Classic Games
Reply #28 - 04/28/20 at 16:03:58
Post Tools
cathexis wrote on 04/28/20 at 14:01:19:
an ordinary chessplayer wrote on 04/27/20 at 16:30:57:
I think there is some value to using engines on master games. Of course there was a limit to what the master was able to see over the board. The point of using an engine is to help identify that limit.


Could this usage imply that some stronger players are,in effect, trying to brute force themselves into an advantage by ... (speculation snipped)

I didn't imply anything like what you inferred. I was talking only about going over classic games, where the players weren't using an engine, and comparing their moves to the engine's choice. I used the word limit, but strong players are not limited to a "6 or 7 move forward analysis capacity". In a single variation they can calculate an unlimited distance ahead. I know I certainly can, although I might not be as fast as a GM. Where they (and I) have trouble is when there are a lot of branches which all need to be compared.

This is off-topic, but since you brought it up: For gaining an opening advantage over the board, it's useless to prepare a long sharp variation in the hope that the opponent won't be able to find all the moves. Good players can easily cope with that. It's more important to find a "new" move where the opponent has a lot of choices but no clear way to choose between them. Faced with that task, even the strongest opponent can quickly go wrong. Such a "new" move doesn't have to be an improvement, aka "Theoretical Novelty". It can just be different (a "wrinkle"), and ideally it should give the opponent a plausible way to go wrong. The key is to prepare it well, and this is where the engine is great, because it can quickly generate all the replies and give a line against each one. So while the opponent is thinking and wondering if they are getting it right, you are still in book.
  
Back to top
 
IP Logged
 
Stigma
God Member
*****
Offline


There is a crack in everything.

Posts: 3265
Joined: 11/07/06
Gender: Male
Re: Modern Engines and Classic Games
Reply #27 - 04/28/20 at 14:10:44
Post Tools
cathexis wrote on 04/28/20 at 14:01:19:
Sort of the chess equivalent of top athletes "juicing" to get a higher level of performance than they could from traditional fitness training. If so, if that analogy is correct and in practice, it leaves me a little confused from an idealistic perspective.


What does "juicing" mean here? I googled it and got instructions for making healthy juice, but I sense you're talking about something else. Protein powder, perhaps? Or some special way of mixing up the training itself?
  

Improvement begins at the edge of your comfort zone. -Jonathan Rowson
Back to top
 
IP Logged
 
cathexis
God Member
*****
Offline


No matter where you go,
there you are.

Posts: 608
Joined: 03/03/20
Re: Modern Engines and Classic Games
Reply #26 - 04/28/20 at 14:01:19
Post Tools
an ordinary chessplayer wrote on 04/27/20 at 16:30:57:
I think there is some value to using engines on master games. Of course there was a limit to what the master was able to see over the board. The point of using an engine is to help identify that limit.


Could this usage imply that some stronger players are,in effect, trying to brute force themselves into an advantage by rote memorizing longer lines in specific variations? For example, they are already strong players with a 6 or 7 move forward analysis capacity who then take narrow variations they are already good at and then use engines to construct an "algorithm" of even deeper moves. They rote memorize that product and then use their skills to (hopefully) steer the game down a path that's winning for them. Or has this become the new normal for top competitors already? Sort of the chess equivalent of top athletes "juicing" to get a higher level of performance than they could from traditional fitness training. If so, if that analogy is correct and in practice, it leaves me a little confused from an idealistic perspective.

Just curious,

Cathexis
  
Back to top
 
IP Logged
 
Jupp53
God Member
*****
Offline


be

Posts: 988
Location: Frankfurt/Main
Joined: 01/04/09
Gender: Male
Re: Modern Engines and Classic Games
Reply #25 - 04/27/20 at 20:47:13
Post Tools
cathexis wrote on 04/27/20 at 13:48:27:
Please don't label me a troll!

I accept the takeback after the mouseslip. Wink

(I had to do something similar short time ago and appreciate the pardon. We have a good coummunity here.)
  

Medical textbooks say I should be dead since April 2002.
Dum spiro spero. Smiley
Narcissm is the humans primary disease.
Back to top
 
IP Logged
 
an ordinary chessplayer
God Member
*****
Offline


I used to be not bad.

Posts: 1672
Location: Columbus, OH (USA)
Joined: 01/02/15
Re: Modern Engines and Classic Games
Reply #24 - 04/27/20 at 16:30:57
Post Tools
I think there is some value to using engines on master games. Of course there was a limit to what the master was able to see over the board. The point of using an engine is to help identify that limit. Identifying that limit has implications for my own play. To a certain extent I can rely on published annotations, or do my own analysis, but the engine is an additional tool. Actually I learn the most when I do my own analysis first, then look at the published annotations, then do more of my own analysis, then check with an engine, and finally perhaps even more of my own analysis. One valuable thing learned from this exercise is being able to triage complications as (1) solvable; (2) not solvable but playable; (3) unknowable and too difficult to be played.
  
Back to top
 
IP Logged
 
an ordinary chessplayer
God Member
*****
Offline


I used to be not bad.

Posts: 1672
Location: Columbus, OH (USA)
Joined: 01/02/15
Re: Modern Engines and Classic Games
Reply #23 - 04/27/20 at 16:08:30
Post Tools
We're all familiar with that thread. The less said about it the better.
  
Back to top
 
IP Logged
 
cathexis
God Member
*****
Offline


No matter where you go,
there you are.

Posts: 608
Joined: 03/03/20
Re: Modern Engines and Classic Games
Reply #22 - 04/27/20 at 13:48:27
Post Tools
Please don't label me a troll!

I will always take peronal accountability for my remarks, but I value my membership here. It was entirely my fault ! I had two linked pages open at the same time and thought I was replying to the other one. My bad! I was reading an older page on Erik Kislik.
"Opening Selection by Erik Kislik."
https://www.chesspub.com/cgi-bin/chess/YaBB.pl?num=1538332239/22

And that one was 10 pages long and full of vitriol at the end. Eventually, GM Tony Kosten used his admin authority to just shut the whole thread down. Read for yourself. However, it was MY mistake to post a reply here. I do apologize. But again, please don't put that label on me. I learned a long time ago that the forums are not for the faint of heart. Actions are always accountable, true. But reputations matter also. Again, my bad.
  
Back to top
 
IP Logged
 
Page Index Toggle Pages: [1] 2 3 
Topic Tools
Bookmarks: del.icio.us Digg Facebook Google Google+ Linked in reddit StumbleUpon Twitter Yahoo