1 edit
Originally posted by ih8sensI think this romanticism about computers being materialistic and all will go on whatever computers do in the next let's say, 100 years. what's "quite wrong..." is people criticising engines and computer play with no actual knowledge or experience about it. I've said this many times on this forum, the romantic statements about computers come from a decade behind, but people love the myths about human intuition, the soul, spiritual artist against the cold-blooded, soul-lacking technician etc, so I'm now beginning to see the ideology behind the opposition is so much greater than I thought in these kinds of discussions.
quite wrong...
Game 3967407 - The double sac. The blind spot of most engines.
and about the positional thing... it's cuz they are materialistic.. their positional play is simply setting up for tactics.. positional play is about knowing from experience where each piece should be.. engines never gain experience.
Originally posted by schakuhrWhat I meant was, I'd like to have the compensation expressed in natural language like "and black gains the control of center with tempo" etc. I cannot see any positional plus for black in those lines.
He meant the line 1. e4 d5 2. exd5 Nf6 3. c4 and then 3. ..c6 or e6.
I can't speak for 3. ..c6 because I've never looked at it, but 3. ..e6 is a correct gambit (Icelandic gambit it's called).
1 edit
After 1 e4 d5 2 ed Nf6 3 c4, all of the books that I am familiar with say that after 3...c6 dc 4 Nc6, Black has more than enough compensation for the pawn. Black leads in development and White can't free his position by moving his d-pawn to d4. Databases give Black a plus score in this position. Grandmaster Larry Evans in "The Chess Opening for You" likes Black in this position. Evans points out, however, that both Lasker and Steinitz liked White in the position! Experience has shown that Evans is probably right and Lasker, et. al. were wrong.
BTW, probably the best move after 3...c6 is 4 d4, which will often transpose into a Panov-Botvinnik Variation of the Caro-Kann Defense.
3...e6 probably isn't as strong, since after 4 de Be6, White can play 5 d4!, which he can't do in the 3...c6 line.
1 edit
Originally posted by diskamylQuick development, a lead in development, pressure against c4 and on the d-file.
What I meant was, I'd like to have the compensation expressed in natural language like "and black gains the control of center with tempo" etc. I cannot see any positional plus for black in those lines.
After 1. e4 d5 2. exd5 Nf6 3. c4 e6 4. dxe6 Bxe6 there are 2 different types of positions: white plays d4 or white does without d4.
5. Nf3?! Qe7! this is a common theme in this gambit. black threatens Bxc4+, the only way to prevent this is 6. Qe2 which seriously hampers white's development. In a recent game at my club a 1900-rated player played 6. Be2 and he probably already was worse after 6. ..Bxc4 7. d3 Ba6.
A better try here seems 5. Be2 Qe7 6. Qb3 c5 when black has compensation with a clamp on d4 and more central influence. the d2-pawn becomes very weak when pushed to d3, but otherwise the c1-bishop can't be activated.
5. d4 Bb4+ 6. Bd2 Qe7 7. Bxb4 Qxb4+ 8. Qd2 Nc6!
Let's see. black is far ahead in development, he can castle next move. White's king is still stuck on the (opened) e-file.
A few sample lines: 9. Qxb4 Nxb4 10. Na3 O-O-O 11. d5 Bf5. black is probably already winning here. Rhe8+ is next up, and its not clear how white can prevent the black pieces from destroying his position.
Or: 9. d5 O-O-O 10. Nc3 Bg4! this prevents white from castling for the time beeing.
We're following the game Kuijf-Hodgson here. White eventually succumbed after some beautiful attacking play from black: 11.f3 Rhe8+ 12.Be2 Bf5 13.0-0-0 Na5 14.g4 Bg6 15.Nh3 Nd7 16.Nb1 Qb3
17.Bd3 Qxa2 18.Qb4 Re2 19.Bxe2 Nb3+ 20.Qxb3 Qxb3 21.Rd2 Qe3 0-1
Originally posted by rubberjaw30That would require thinking. If a machine could actually think it would have passed the Turing test and the maker would have won a huge prize. Chess engines have everything coded in, they don't learn like humans do, so it won't matter how often a computer witnesses a particular tactic or position. It will never learn to avoid the trap or use it if it cannot do so already.
aren't computers able to "learn from experience" by noticing patterns in their databases?
for example...
numerous games involve an h-pawn storm on the king...
shouldn't a computer be able to say, hey! this could be a good idea!
1 edit
Originally posted by diskamylStudying computers is what got me back into chess. In the two years since I've picked up the game again that's about 50% of my chess time playing computers.. the rest being on RHP.
I think this romanticism about computers being materialistic and all will go on whatever computers do in the next let's say, 100 years. what's "quite wrong..." is people criticising engines and computer play with no actual knowledge or experience about it. I've said this many times on this forum, the romantic statements about computers come from a decade be ogy behind the opposition is so much greater than I thought in these kinds of discussions.
I know computers and chess engines pretty well... they ARE materialistic and they are NOT all that good. Engines are fantastic defenders but they aren't really great attackers (in comparison to their supposed ELO).
I know this post is about to be ripped to shreds but frankly, I don't care. My experience with engines has taught me this and I will stand by it regardless.
2 edits
Originally posted by ih8sensIf computers "are NOT all that good", as you say, then how come one of the world's top ten players, Michael Adams, could only manage one draw in a six-game match? The best computer absolutely CRUSHES the best human player.
Studying computers is what got me back into chess. In the two years since I've picked up the game again that's about 50% of my chess time playing computers.. the rest being on RHP.
I know computers and chess engines pretty well... they ARE materialistic and they are NOT all that good. Engines are fantastic defenders but they aren't really great attacker ...[text shortened]... on't care. My experience with engines has taught me this and I will stand by it regardless.
Originally posted by diskamylwell, that's just where you're wrong. I've been messing around with computers from the 80's and studying them academically from the 90's. I could build a basic superscalar processor from logic gates, code a crude operation system, compilers for higher level languages and networking layer and the programs on top of it. it wouldn't be a great system because I'm not an expert at all that, but you know, I can make it work. I do know what I'm talking about when it comes to computers, and I also have gathered some understanding of chess. and all I know says you're the one with the romantical perception about computers. until there's a paradigmn shift, all the (top level) chess programs are essentially the same: they relay on exhaustive brute force search, pruning and set of 'positional' rules for evaluation, and that's that. you can make such a search slightly faster, pruning more effective and tweak the rules a bit, but you can't change it's essence. machines don't think and they can't understand.
I think this romanticism about computers being materialistic and all will go on whatever computers do in the next let's say, 100 years. what's "quite wrong..." is people criticising engines and computer play with no actual knowledge or experience about it.
Originally posted by gaychessplayernone of you have answered my question:
I totally agree! But, Teichmann was right and computers have proved him right: "Chess is 99% tactics!"
why are computers so far behind witht their positional understanding?
for example: in the King's Gambit Accepted, don't computers give black a plus, when history shows that white has plenty of compensation for the pawn, and black usually has to give the pawn back to equalize?