Humans v Houdini chess engine (Elo 3300)

Hah!

I am thinking the count of refused matches is even on both the human and computer side. Kasparov wanted a rematch when he thought Deep Blue was being helped by humans (this was 90's tech by the way don't tell me this proves humans > engines today because it doesn't) and he was refused. IBM wanted a third Deep Blue vs Kasparov match and he refused.

Humans know they're going to be beat so they have very little interest in a high profile man vs machine game. I'd like to see a GM play three matches against an engine. One on a laptop (or better yet, a smart phone), one on the best desktop computer money can buy and one against the unleashed wrath of a top-10 of 2013 supercomputer.

Another interesting matchup would be a team of GM's against a supercomputer. 10 minutes per move. I want to know how many Grandmasters it would take to beat your "positionally stupid" engines.

Did this happen? Sounds like it would settle the issue.

pfren: How do you defend yourself against first criticizing a person for talking about supercomputers without naming one and then you yourself refering to games without naming any? It seems like you're dodging the issue fairly nimbly.

If a beginner randomly picks houdini moves (just the top move every time) then I could see it falling into a worse position or at least a draw vs a player who knows what they're doing and has a weak computer to check tactics for them.

If a master uses a few programs and a fast computer to help with CC it would be a different story.

Well of course the time control matters. I was saying something similar myself not too long ago. The hardware matters just as much. You need a good combination of engine and hardware. For centaur players, you need a good combination of engine, hardware and knowledge. A smart player is a substitute for a good deal of brute force, but no player is smart enough to beat a full-fledged Houdini or Rybka on a very strong computer.

@pfren: That didn't really answer my concern. The point of a debate is that it DOES matter whether or not someone believes you. That is the whole point. Watch:

I think they are 500-foot tall solid gold sheep on the moon. They eat rainbow watermelons the size of Stadiums and talk endlessly about ancient Martian litterature. You don't believe me? Well I don't ------g care if you don't because it's true.

What??

Do you also play as a centaur (C) at ICCF? OOpps I forgot it. Already, majority of the players are centaurs at ICCF. Please don't get this message as a humiliation to centaurs. I respect them a lot. It's a tough job I guess. The problem is (as explained by many good-hearthed people above) we don't want to play against neither centaurs nor pure-engine. This is why majority of chess webservers still strictly monitor engine-usage for rated games. They are following demand. Currently people are demaning pure human play, in other words cheat-free matches they want. Is that too much demand? We're not saying neither centaur chess nor pure engine chess must be removed from the chess history. We just want space we deserve.

@FirebrandX: Thank you for this. I can see why 15. Qc1 does in fact look strange. My guess is the more standard moves are 15. Bxc4 or 15.Qf4 or something of the like.

In the Vote 960 Chess game currently going on, white (the team I am on) is insisting the game is won yet a few from black are bringing "evidence" from their engines that the game is drawn. We have quite thoroughly analyzed the rest of the game and we could probably all go away and leave behind a little note saying if 42. ... A then 43. B, if 42. ... C then 43. D and win. Yet the engines say it's a draw.

Either we're missing something or, more likely, the engine is. I can't really see what remarkably revolutionary move we're playing and I doubt an engine is seeing something in a few minutes that several 2000 players have been missing for the last month.

If I were to use an engine for that endgame, I could see which part of our plan the engine is missing, or to which of black's responses we haven't given enough consideration and that, I suppose, is where the strength of centaurs comes from. That and their horse bodies.

The engines definitely have a weakness of some kind and it's obviously in positional value (I was advocating that myself). I just think that if Deep Blue vs Kasparov was so close in 1997, then fifteen years later the ludicrous advances in computing will have closed that gap and then some.

I think the brute force of a computer will far outweigh the positional intuition of a human. In your game, I believe the computers will have missed 15. ...e5. Would you say that innaccuracy cost black the game or was there some other mistake or innaccuracy played later? I'm sure a computer could have come up with much stronger tactics than what you encountered.

As a follow-up, I think it would be quite interesting to have a particular engine play against itself starting from a few key positions around move 15. Like, check white's other possible 15. moves and a few of blacks possible 15. ... moves and just see if that small inaccuracy there was enough to take black apart or if a few other things have to happen to completely cost black the game.

Haha! Don't worry FEDTEL, IM Pfren is someone who makes other stupid by the way he argues. Though I know we can agree on a fact: 98% of the time he is right.

hahaha!, I don't want to be rude, but yeah, review your post first

Soo you're basicly saying you won't prove what you're saying? I wonder who's living in a fairytale here.

That are many people that are reading this discussion that, while already agreeing with you, would appreciate some more examples.

I really enjoyed studying the position you posted earlier, it was really interesting.

Medium crab bisque.

It's funny that you assume I'm an enginefanboy when I never stated my opinion. 

And btw, apperently it's meaningless to ask the fool for proof: He will just state he's the better and thereby is right. :) 

I'm not even sure you are sarcastic.

Did anyone get a book of pithy sayings for Christmas?

Mature guy caught in action!

I can hardly follow the path of argument anymore. It seems this one just keeps diving off into sideline topics. If the question is whether humans are better than chess engines, the question's already answered; it's no. Can a human beat a chess engine? Yes of course, not in an overall match, but I'm sure out of 100, Carlsen or one of our top GMs might get a win or some wins against Houdini 3.

Engines can still make mistakes, nothing plays chess perfectly and I have my doubts that anything ever will, but just because Houdini can make an error, does not mean it is not stronger than any human alive. Its overall consistency and tactical ability is just overwhelming.