Oh, please. The reason these matches aren't held, is because no professional player wants to waste their time and energy grinding away in a futile battle against a computer.
PEP, in the position below, white has a completely, 100% winning move. A human player found it OTB. I am curious how long will it take for your setup to find it.
Ok, I'm not seeing it. 23.Qxe5 fxe5 24.Rf1 Rc7... where is the win for White? Looks like a draw to me. Maybe it is a draw, and that's why Houdini isn't finding a win. Why should we believe the claim that this is a win? Someone said it was zugzwang after 25.Bd1 Re7, and indeed that is true, but Black doesn't have to play 25...Re7. My Houdini found a nice rook sac instead, and it didn't take more than a minute.
So, you tell me? Does this position show the weakness of computers, or does it show that Houdini has found a brilliant variation which has been overlooked by others?
PEP, in the position below, white has a completely, 100% winning move. A human player found it OTB. I am curious how long will it take for your setup to find it.
Ok, I'm not seeing it. 23.Qxe5 fxe5 24.Rf1 Rc7... where is the win for White? Looks like a draw to me. Maybe it is a draw, and that's why Houdini isn't finding a win. Why should we believe the claim that this is a win? Someone said it was zugzwang after 25.Bd1 Re7, and indeed that is true, but Black doesn't have to play 25...Re7. My Houdini found a nice rook sac instead, and it didn't take more than a minute.
So, you tell me? Does this position show the weakness of computers, or does it show that Houdini has found a brilliant variation which has been overlooked by others?
I'm no master, but I don't see what can be done to stop the plan of, b3, bxc4, c5, c6, c7, Rf7 (or make a "pass" move if the queen guards it. Black runs out of moves), Rg8+, Rd7+, Rd8
I hope I'm not just missing something obvious.
Edit: I will play it against Houdini to see what's up.
Adam: I concur with browni, Black is out of moves... though I would first like to move my king to g3 to avoid any surprise checks.
Apparently that line was wrong, but trying other variations Houdini eventually concedes that white's position is winning. 27. a4 a6 28. Ba2 (planning b5)
For all the (mostly outdated) insults about computers not understanding positional chess, Carlsen, Kramnik, Anand etc always seem pretty interested in how Houdini evaluates a position in post match interviews.
It's not so difficult to beat Houdini in correspondence chess. Computers still lack certain elements of positional understanding, and they can certainly be outplayed by a strong player. On rapid/blitz games though, it is a totally different story.
This is an interesting point that I've discussed with other human players. My contention was that a strong player, such as yourself, could beat an engine in correspondence play. My opinion was in the minority. Most thought the monstous calculating potential of a chess engine would carry the day even against very good players in a correspondence game!
If it was an OTB game I highly doubt his opponent will have had an eval to be trusting of. Also pfren it doesn't flatter you to be rude to people who are debating something with you ("and any dumb guy (not calling names, although it's quite clear who the dumb guys in this thread are) will take this for granted").
(apparently Houdini told him that he was totally lost).
Figured he meant Houdini analysis from after the game. Also plausible. But you're right I probably misunderstood since he plays a lot of corre games.
This debate is pointless. people made chess engines. somebody told the engine that move a is better than move b and why. without people then engines would totally suck. Therefore people are clearly better.
Ummm... since Pinochio's engine rejected trading queens, here is a new one- but here no hints for Pinochios.
White to play and win. Post your solution as pgn, it must be thorough (covering all possible black defences).
Good luck with your engine for the next couple of weeks. Just a small hint: a Grandmaster can settle this within ten minutes' thought.
why would I lie about a thing you can check for yourself in 5 minutes ?
houdini 1.5 is free, download it and leave it running for 1 minute on your commodore 64.
As for your other position houdini 1.5 running on my old laptop for some minutes likes Kh2 as well... seems youve deeply analyzed this position, it might be easier if you post your "miracle line" rather than posting huge trees and every possible form of defence for all lines from engines.
Ive played my hand anyway: Kh2 , lets see if houdini can refute your miracle line?
another point to be made is ...
it was the humans decision to resign , not the computers.
if the computer plays on, it probably would find the miracle draw variation. (given that it exists!)
The computer doesnt see it yet, but it will continue to play without making big mistakes, cut those lines out (ie it knows not to move the king, and that it should be active with the knight , which cuts down a huge amount of calculation to be done).
This pfren guy is always finding ways to sneak in human elements to make the computers look bad.
If you want a clear mis-evaluation by a computer, I always like to use the Traxler counterattack example. Every computer I have seen has evaluated 5.Nxf7 as the clearly superior move (even saying numbers like +2.00 or more) when it is clearly an equal move by theory (or even lost in the anti-computer variations).
The point is the position can be drawn. The opponent resigned when he saw the eval climbing over +2.00, but it's still a draw with best play. Refer to my previous post demonstrating the uselessness of white's queenside pawns.
ye Ive played thru many of the variations now and houdini ends up finding ways as black to get rid of the king side pawns and draw by just holding and blockading the queenside pawns as you mentioned. (pfren, I mean pinochio, was trying a misdirection with the: "white to move and win").
Houdini 1.5 still has an eval of +3 when there are only the 2 queenside pawns + B vs N remaining.
This is obviously incorrect but its also irrelevant as it keeps making the best moves.
The evals also mean different things for different engines.. for some engines its a material count, for others its a probability of how likely it is to win the position.
The only thing that matters is the engine finds the draw, not its evaluations leading upto that. As engines improve these known drawn positions will be programmed in anyway.
If you are so sure about your discovery, then feel free to accept a $ 1,000 challenge. I will play black, your Houd... err, I mean you, white, and you have to win the game and $ 1,000, else you lose $ 1,000.
Ok, well, I'm going to call bullshit on you. There is no way either of us can compel each other to pay the bet, so your entire offer of a wager is just nonsense. We might as well make the gamble for a billion dollars. Do you accept, or you are just cheap talk and that is that?
Besides, I never said for certain that White could win. I think white could, there is certainly a very strong chance. All I said was that Kh2 is the correct move for White, and you apparently agree. There was no need for White to calculate. Kh2 is the best move. Whether it wins or not is irrelevant.
Ultimately, what I see here, is you claiming that this position cannot be properly assessed by a computer, and claiming that you know the best moves for Black which will refute a computer's analysis... are you willing to put your reputatation on the line by playing this out? I am. Let's do it. Show me that you can stop White. If you can, great, you'll make your point and we'll all learn something. Otherwise... what are you even talking about???
Dude, even I could draw this with black now (in centaur CC that is). I just had to key in on the two passers being useless.
I'm not saying you can't. I've looked at it some, it's not an easy win for White, assuming it is a win. But pfren's original post involved a specific position, and he said there was a win for White which the computer supposedly can't find. I don't know what that win is. He said, "White to play and win." I think Kh2 has got to be the first move. I'm still waiting for him to move. Its his turn. The position may very well be a draw, but he did claim that there was a win for White. That's his claim, not mine.
If the position is a draw, then why did Pfren say "White to play and win"? That's just dishonest.
It's not that simple. After 44.f4 e4 45.Bc4 g5 white may attempt to use his king via c3, after properly positioning his bishop (to prevent the kingside duo to move too far down the board), when Black is lost if he plays passively, or by taking on f4. But he does have a plan: taking on b5 with knight in some variations.
There is also the line 44.f4 e4 45.f5, where Black saves his ass by the cunning 45...Nd5 46.Kg3 h5! inviting the bishop to a mined square.
This is why a GM would need ten minutes to evaluate this, else a couple of them would suffice. In evaluating such positions, humans are still much superior to even the most advanced engines, sice they have the ability to think schematically.
The computer will draw the f4 line as well as the Kh2 line (without any guidance)... if you think otherwise play it out with houdini and report back with your win.
First of all since when do computers play passively? Never.
Second, computers can and do think schematically...
here are just a FEW examples
Lol at these engine fans, an engine uses algorithm to find moves then how can it "think"? WTF happened to philosophy? Is it already dead.
is the point that the move actually isn't kh2? Because that would be surprising.
If that's the case, I'll definitely learn something. However, regardless of time control, I wouldn't hesitate to play Kh2 here. It's just the only rational move.