Our dedicated chess computers in fact 300-350 elo weaker ??
Moderators: Harvey Williamson, Steve B, Watchman
Forum rules
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
Tests?This is more confusing. The test is not correct. I have a real rating of about 1900, and maybe less, and not 2300.
We need new tournaments like human-computer, or matches like IM Roman Vidonyak - Chess Genius Exclusive.
It is also necessary to divide the ratings by time control:blitz,rapid,tournament.
These are different ratings.
All the rest is just a discussion, there is no truth in them.
Now more opportunities.The Databases of players, methods of training, game online - all this is available.
This affects the overall level of players.
Now the grandmaster in 14-15 years is normal for a talented child.
https://en.wikipedia.org/wiki/Chess_prodigy
Why? There are more opportunities for growth.
We need new tournaments like human-computer, or matches like IM Roman Vidonyak - Chess Genius Exclusive.
It is also necessary to divide the ratings by time control:blitz,rapid,tournament.
These are different ratings.
All the rest is just a discussion, there is no truth in them.
Now more opportunities.The Databases of players, methods of training, game online - all this is available.
This affects the overall level of players.
Now the grandmaster in 14-15 years is normal for a talented child.
https://en.wikipedia.org/wiki/Chess_prodigy
Why? There are more opportunities for growth.
-
- Full Member
- Posts: 679
- Joined: Mon Aug 29, 2016 8:31 pm
- Location: Cheshunt, Hertfordshire, UK
I thought that was the object of the test,not to accept it as wrote,but to see just how a good dedicated computer would do in those positional tests.The elo part over 76 problems obviously cannot be taken seriously as the people who set them might have us believe,bring back the good old Aegon tournaments.
There are a lot of imponderables in equating the computer chess ratings when compared to strength versus humans not least of which is the engines style of play. For example as a 1950 ish elo player I find the Langs fairly easy to draw against but harder to beat, but they do seem to lack an understanding of tempo and time in my view and their passivity can make them vulnerable to slow positional attacks against which they will often just shuffle.
On the other hand I can play an old morsch program which is nominally 200 rating points weaker and find it an absolute nightmare since it will apply a lot of pressure with plausible attacking moves and actively seeking to open the game up.
This raises the question of humans adapting anti computer strategies based upon a knowledge of a programs style strengths and weaknesses even without the rather sad tactic of playing to exploit known holes in books. This obviously can confound ratings that are based on human computer games. A true rating would only be established by a series of double blind games where the human does not know not only what engine they are playing, but also even whether they are even playing an engine at all. This would be achieved by randomly mixing in human opponents. But ultimately understanding and exploiting an opponents weaknesses is very much a part of human chess as well, we all have weaknesses.
Given the complexities I don t think much would be achieved by a simple deduction from the rating lists if the aim was to create a closer correlation to human ratings. There are ample historical games for many of the dedicated computers against humans from which ratings can and have been derived and from which one can draw ones own educated conclusions
On the other hand I can play an old morsch program which is nominally 200 rating points weaker and find it an absolute nightmare since it will apply a lot of pressure with plausible attacking moves and actively seeking to open the game up.
This raises the question of humans adapting anti computer strategies based upon a knowledge of a programs style strengths and weaknesses even without the rather sad tactic of playing to exploit known holes in books. This obviously can confound ratings that are based on human computer games. A true rating would only be established by a series of double blind games where the human does not know not only what engine they are playing, but also even whether they are even playing an engine at all. This would be achieved by randomly mixing in human opponents. But ultimately understanding and exploiting an opponents weaknesses is very much a part of human chess as well, we all have weaknesses.
Given the complexities I don t think much would be achieved by a simple deduction from the rating lists if the aim was to create a closer correlation to human ratings. There are ample historical games for many of the dedicated computers against humans from which ratings can and have been derived and from which one can draw ones own educated conclusions
-
- Full Member
- Posts: 679
- Joined: Mon Aug 29, 2016 8:31 pm
- Location: Cheshunt, Hertfordshire, UK
Just for the record, this was the first time a dedicated computer got an I.M. norm against human competition. The computer is the Portorose 68030 and the tournament is the Neu Isenburger 1990.
1.B v I.M Schneider 2460 0-1
2.W v Radlica 1560 1-0
3.B v Norenberg 1590 1-0
4.W v I.M.Hresc 2335 1-0
5.B v I.M.Komijenovic 2470 .5-.5
6.B v I.M.Krant 2405 1-0
7.W v Schmidt -Schafer 2190 1-0
8.B v Lalic 2470 0-1
9.W v F.M Nikolaev 2340 1-0
10.W v G. M. Lau 2470 1-0
11.B v G.M. Smejkal 2525 .5-.5
Regards Steve C.
.
.
1.B v I.M Schneider 2460 0-1
2.W v Radlica 1560 1-0
3.B v Norenberg 1590 1-0
4.W v I.M.Hresc 2335 1-0
5.B v I.M.Komijenovic 2470 .5-.5
6.B v I.M.Krant 2405 1-0
7.W v Schmidt -Schafer 2190 1-0
8.B v Lalic 2470 0-1
9.W v F.M Nikolaev 2340 1-0
10.W v G. M. Lau 2470 1-0
11.B v G.M. Smejkal 2525 .5-.5
Regards Steve C.
.
.
-
- Full Member
- Posts: 679
- Joined: Mon Aug 29, 2016 8:31 pm
- Location: Cheshunt, Hertfordshire, UK
It's in selective search magazine 32, and would you believe it doesn't give the time controls ?But it does state that it got an I.M. norm at the tournament ,I don't believe I am mistaken the games would have been at tournament time settings,40 in 2,and does state that if it wasn't for the chess olympiad and world championships it would have made bigger news,and yes I know it's a chess computer magazine in the business of selling them,but there certainly are a considerable number of computer v rated humans at tournament time controls at the time.
- spacious_mind
- Senior Member
- Posts: 4000
- Joined: Wed Aug 01, 2007 10:20 pm
- Location: Alabama
- Contact:
- spacious_mind
- Senior Member
- Posts: 4000
- Joined: Wed Aug 01, 2007 10:20 pm
- Location: Alabama
- Contact:
Well this seems to confirm that at least in Blitz Chess, dedicated computers surpassed humans in the year 1991:
World Blitz Champion Michail Tal gets hammered by Mephisto Lyon 68030 by 15-9. Pretty conclusive.
Hail the new World Blitz Champion "MEPHISTO LYON 68030"!
Selective Search is full of great information!
Best regards
World Blitz Champion Michail Tal gets hammered by Mephisto Lyon 68030 by 15-9. Pretty conclusive.
Hail the new World Blitz Champion "MEPHISTO LYON 68030"!
Selective Search is full of great information!
Best regards
Nick
- Steve B
- Site Admin
- Posts: 10140
- Joined: Sun Jul 29, 2007 10:02 am
- Location: New York City USofA
- Contact:
the match was also reported in Computer Chess ReportsVolodymyr wrote:Hi Nick,
Tal lost,I do not believe in this. This is just information. Where are the games?
This is commerce.Paper suffers all.I do not believe!
Lyon 030 also defeated Spassky and Short in a Blitz tournament held at the same location as the individual match Vs Tal
you can read it here on page 1:
http://www.chesscomputeruk.com/Computer ... _1991a.pdf
Not Fake News Regards
Steve
- spacious_mind
- Senior Member
- Posts: 4000
- Joined: Wed Aug 01, 2007 10:20 pm
- Location: Alabama
- Contact:
Hi Volodymyr,
Don't think that Kasparov, Anand or Kramnik were exempt. Below is Blitz Tournament 3 Years later. Fritz 3 is only marginally better than Mephisto Lyon 68030.
I think the table below says it all. It's as a result of this that Grandmaster quickly stopped playing chess computers or computer chess at Blitz at around that time as they knew they would get beaten.
Best regards
ps...it is not propaganda its fact. Just ask the people who remember the old chess servers where computers played as well as humans and had incredible Ratings like 3200, 3300, 3400 as a result of online matches in Blitz.
Don't think that Kasparov, Anand or Kramnik were exempt. Below is Blitz Tournament 3 Years later. Fritz 3 is only marginally better than Mephisto Lyon 68030.
I think the table below says it all. It's as a result of this that Grandmaster quickly stopped playing chess computers or computer chess at Blitz at around that time as they knew they would get beaten.
Best regards
ps...it is not propaganda its fact. Just ask the people who remember the old chess servers where computers played as well as humans and had incredible Ratings like 3200, 3300, 3400 as a result of online matches in Blitz.
Nick
-
- Full Member
- Posts: 679
- Joined: Mon Aug 29, 2016 8:31 pm
- Location: Cheshunt, Hertfordshire, UK
- spacious_mind
- Senior Member
- Posts: 4000
- Joined: Wed Aug 01, 2007 10:20 pm
- Location: Alabama
- Contact:
Yes there are more opportunities for growth today. People have more time on their hands and technology to help them. But there were child prodigies even in the 19th century like Paul Morphy as just one example. (and no computers to help them study and memorize! ) I wonder how many of these prodigies will actually become Worldchampion. I only see one on that list so far with Magnus Carlsen.Volodymyr wrote:Tests?This is more confusing. The test is not correct. I have a real rating of about 1900, and maybe less, and not 2300.
We need new tournaments like human-computer, or matches like IM Roman Vidonyak - Chess Genius Exclusive.
It is also necessary to divide the ratings by time control:blitz,rapid,tournament.
These are different ratings.
All the rest is just a discussion, there is no truth in them.
Now more opportunities.The Databases of players, methods of training, game online - all this is available.
This affects the overall level of players.
Now the grandmaster in 14-15 years is normal for a talented child.
https://en.wikipedia.org/wiki/Chess_prodigy
Why? There are more opportunities for growth.
I wonder how a modern player would actually (remove studying these games on a computer because they exist from History prior to the encounter) play against someone like Anderssen who puts him into situations like this every time after an e4 or d4 move
https://lichess.org/vyMHodr8
It would be interesting to see as the modern player would be thrown out of his book knowledge after a few moves every time. All preparation and study out of the window
It took 150 years through computers to finally figure out where the real mistakes were in the above game example. The human never did really figure it out all by himself and dozens of books were written about this game over 150 years.
Also there are 6-8 Billion people in the world today so its a bit of a numbers game really too.
We are not so superior. People are still trying to figure out how the Pyramids were built with hands and wheel.
Best regards
Nick
It was a romantic era in chess.Romance and now there is,fans of sharp positions.
The victory of the Kramnik, in the style of Tal.This is Tal memorial 1995(Riga)
https://lichess.org/e3JsVNo9
A beautiful victory.
Collection of beautiful games
http://www7.zippyshare.com/v/OmQVFGcE/file.html
The victory of the Kramnik, in the style of Tal.This is Tal memorial 1995(Riga)
https://lichess.org/e3JsVNo9
A beautiful victory.
Collection of beautiful games
http://www7.zippyshare.com/v/OmQVFGcE/file.html