As for Hydra, the two CC games Arno Nickel won were against an older version than the one that scored 5.5-0.5 against Adams (then world #7) seven years ago. It's hard to compare Houdini of today with Hydra of seven years ago, the development is fast in computer chess. In CC the human of course uses various chess engines, so there's no reason that any strong CC player equipped with Houdini would be at a disadvantage against the same engine, but that is another question.
Fischer vs. Houdini, WHO WOULD WIN??
Sort:
pfren wrote:
Engines play nonsense in specific closed positions, as they still lack positional envision. [...]
Surely it is only a matter of time before that specific positional knowledge/evaluation is programmed in?
pfren wrote:
Provided ample thinking time, Fisher would easily win in normal chess, while Houdini would quite possibly win in Fisher chess (Chess960).
Ironical, eh?
Well I find this statement rather interesting. Why is it that you think that an engine would probably be better in 960 but not in the normal starting position?
Tactics shouldn't be the problem since we allocate the human players enough time.
kasparov vs houdini. who would win?
pfren wrote:
Artsew wrote:
pfren wrote:
Provided ample thinking time, Fisher would easily win in normal chess, while Houdini would quite possibly win in Fisher chess (Chess960).
Ironical, eh?
Well I find this statement rather interesting. Why is it that you think that an engine would probably be better in 960 but not in the normal starting position?
Tactics shouldn't be the problem since we allocate the human players enough time.
Simply because chess960 does have a lot of complex positional setups, but some 99% of the GM's (the late Fischer included) are not terribly familiar with.
So, in chess960, the main weakness of the chess engines currently (proper evaluation of the strategical elements of a lot of stock positions), would be of lesser importance. OK, a super-GM when motivated would invent anti-computer positions even in chess960, but this requires quite a bit of work, and I guess noone will bother doing that without motivation (which of course is money).
ok. thx for your response. Too be honest, I don't know much about 3000+ strength chessprograms like Houdini.
However it does seem logical to me that programming it to avoid closed positions (perhaps even at the costst of a pawn?) would not be to hard, if it would be necessary that is. On the other hand I agree that "weaker" programs tend to have problems with closed positions.
The main reason that I responded was that I always thought that having the option to use a database would be a huge advantage for a program over the human. And they lack that option in 960.
Honestly, I don't like those topics when we compare 2 players of different eras/a human player to a machine. But to answer the main question, I think Houdini will win. Fischer is certainly a great chess player, but he can be unfamiliar with many modern lines and new ideas about the game, which Houdini "knows" perfectly well.
Feb 22, 2012
0
#89
fabelhaft wrote:
browni3141 wrote
There's a huge difference between CC and OTB, so of course bringing up CC is relevant to the discussion.
The discussion originally concerned the statement that Fischer would beat Houdini easily in normal chess and that other top players would do the same given enough thinking time to avoid tactical errors (without specifying how much time they need to avoid tactical errors). As for CC, Fischer had no experience of it and I have no idea how well he would have done, but statements about how easy it would be for him to beat Houdini should be taken with a grain of salt. Five years ago Ehlvest played a pawn odds match against Rybka and lost badly, this is how one of the games looked:
http://www.chessgames.com/perl/chessgame?gid=1471891
Ehlvest reached 2643 around then and had briefly been top 5 the previous decade so he is no weak player. Of course Houdini of today is a few hundred Elo points stronger than Rybka was five years ago, and not an opponent to be underestimated.
The OP was just a joke I think, not meant to be taken seriously.
The reason I say Hydra may have been stronger than a modern engine on a PC is because of Hydra's advantage in hardware.
I don't know if Nickel was allowed use of an engine or not. This page makes me think not. They say Nickel was "Unplugged".
http://www.chessgames.com/perl/chessgame?gid=1339073
Here are the other two games:
http://www.chessgames.com/perl/chessgame?gid=1411074
http://www.chessgames.com/perl/chessgame?gid=1339074
Edit: It seems both sides got assistance:
http://www.enotes.com/topic/Arno_Nickel
just analised the last game with Chessmaster GM ed. with a 1 minute / move thinking time and "even" a 2970 ELO engine with a those basic settings recognized 42. Rg8 (in the second game) as a terrible blunder. I just can't imagine adding Houdini that is >300 ELO stronger than CM. Again , if someone has a match won in CC against x64 bit version of Houdini running on at least dual core i would like to see it. I see numerous people trying to convince me chess engines play poor positional chess but so far i never saw a chess game (be that CC or not) in which the human won.
pfren , your move.
again , i would like to see that CC game where a human *defeated* a properly configured Houdini or Rybka computer.
come on whe will never know if fisher was the best and the greatest , he never played to kasparov , kasparov level is much much much higher than boris spassky, kasparov played to boris spassky and won easily
houdini could make fischer disappear
maybe that is what happened to fischer in 1973
by the way,
i read that he stayed at some canadian gm's house in san francisco
https://en.wikipedia.org/wiki/Peter_Biyiasas
There's a huge difference between CC and OTB, so of course bringing up CC is relevant to the discussion.
The discussion originally concerned the statement that Fischer would beat Houdini easily in normal chess and that other top players would do the same given enough thinking time to avoid tactical errors (without specifying how much time they need to avoid tactical errors). As for CC, Fischer had no experience of it and I have no idea how well he would have done, but statements about how easy it would be for him to beat Houdini should be taken with a grain of salt. Five years ago Ehlvest played a pawn odds match against Rybka and lost badly, this is how one of the games looked:
http://www.chessgames.com/perl/chessgame?gid=1471891
Ehlvest reached 2643 around then and had briefly been top 5 the previous decade so he is no weak player. Of course Houdini of today is a few hundred Elo points stronger than Rybka was five years ago, and not an opponent to be underestimated.