< Earlier Kibitzing · PAGE 7 OF 8 ·
|Feb-26-09|| ||FiveofSwords: The only way we make computers decent at chess is really by hardwiring these long term concepts into it. Expiraments with programs that dont have positional concepts hardwired has only produced pretty weak programs. It goes without saying, however, that since humans have come up with these 'rules' to guide the computer in its evaluations, only humans would fully understand these rules, understand how to take advantage of them strategically, and understand when there are exceptions (for example the not-really-exposed 'karpov' king). Like I said I have no doubt that if some human wanted to they could specialize in anti-computer play, and probably beat up pretty good on computers. But it just doesnt sound liek much fun to do this cause computers dont care.|
|Feb-26-09|| ||chessman95: <gangstaman>
<the computer is basing it's view of the value of a piece only on certain predetermined factors.> Yes, but the how much each factor impacts the relative value of a peice is yet to be determined. Through practice and learning the computer will weight each factor according to the impact it has learned it has, and then apply the weighted system to the peice, so the factors aren't really predetermined; the computer just knows what they are.
<For example, let's say you see a line for yourself that generally works out in your favor. However, you are able to see that your opponent has a particular response to this line which renders it inferior to your alternatives. You can see its inferiority, yet you play it anyway hoping your opponent does not?> This is a bad example for what I was saying. I'm not saying that I will set a trap even if the correct move gives me a disadvantage. I'm saying that often regular computer engines will play a move that I don't get at all because it sees that in the other lines if I play perfectly then it will have a disadvantage. Regular computer engines always assume optimal play when evaluating a position. AI engines on the other hand can be made to evaluate all positions, even if they have a possible but complex refutation. So they will score not only the "perfect" positions, but also factor in the moves that humans will more likely play, and will get away with 99% of the time. They are much more usefull in evaluating positions as humans are likely to play them.
|Feb-26-09|| ||chessman95: <FiveofSwords>
<The only way we make computers decent at chess is really by hardwiring these long term concepts into it.> True, but as I have said with AI engines, you can still do this, but instead of giving each "concept" a relative value, leave that up to the computer to learn with practice, which can lead to good evaluations of positions.
<I have no doubt that if some human wanted to they could specialize in anti-computer play> Absolutely! I am actually very good at this myself, because I play against a portable computer engine I have at home 3-4 times a day, and I have done a lot of research on anti-computer play. Like you have mentioned, at this point in time turning off the opening book in chess engines would be a disadvantage, so I have learned a lot of rare variations in an attempt to get it out of the opening book. However, you have also said that you could probably beat a chess computer in the opening if its book was off, but this is not the reason I do this. I get out of the opening book to make the computer waste time thinking in the early opening, so that its play in the middle game is not as strong because it is out of time to think. This is the main reason that computers are weak with their opening books turned off: not because they play bad (although they might) but because they run out of time thinking.
You should also note that anti-computer play is assuming regular chess engines, but I have never heard of an anti-computer player who played against AI engines.
|Mar-09-09|| ||FiveofSwords: A funny thing happened today to me in this line which is very relevant to the discussion. I was analyzing various lines in the alapin with my computer, running fritz11 on a very high powered machine, probably would ahve elo of 3100 or so. This was the variation in question: 1 e4 c5 2 c3! d5 3 ed Qxd5 4 d4 Nf6 5 Nf3 Bg4 6 Be2 cd 7 cd e6 8 Nc3!? Qa5 9 Qb3!? Bb4 10 0-0 0-0..of course the same position could be reached via various move orders, and i really dont know what the optimum move order would be for both sides, but either way this is a key position. Now usually white has played a3 here...But i was considering that move a little illogical, because the black bishop does not seem well placed on b4 anyway. So i was looking for options, and considered the idea of 11 Ne5!?, which has the interesting threat of Nc4, winning a piece, and of course hitting the bishop on g4. Also I considered that maybe I could take advantage of not having yet commited the queenside bishop with ideas of swinging the queen to g3 perhaps and Bh6,things like this. So some quick analysis confirmed that black's only response was 11..Bxe2 12 Nxe2 and Nc6! ...This might seem like an obvious move but its not the first one that would come into my mind personally as black, id be more likely to play Nd7, but after some slightly tricky tactics that move simply loses a pawn here. So anyway, 13 Nxc6 bxc6 14 Bf4 Nd5 15 Bg3 (Be5!?)...This position my computer felt was slightly better for black, which I considered to be very confusing, because I saw a clear plan for white to win the endgame, i.e. put pressure on c6, end up trading it for d4, and push the queenside majority, forcing black to commit pieces to blockade the pawns on the queenside, which would leave him vulnerable on the kingside, ergo losing his kingside pawns and the game. A very well know, typical, tried and true endgame strategy that has won me a lot of games. But since my computer insisted that black was slightly better, I decided to test it, and played out the game, not getting any assistence for my moves, just letting the computer play vs me in this position it for some reason felt it was better, which i did not understand. I won the game.|
|Mar-09-09|| ||chessman95: Great job. You beat fritz11 in some Alapin line that you knew was better for white. Maybe consider a stronger engine next time? Also, the analysis of fritz11 is pretty good, but I'm not so sure about how it actually plays. We already talked about anti-computer methods, so did you really expect a computer to beat you in an endgame where you could see twenty moves out that you would be able to demolish the queenside? Not a good place to be asking computers for analysis.|
|Mar-10-09|| ||FiveofSwords: well of course you never really know if there is some bizarre tactical sequence that ocmputer sees that would change the evaluation of the position with computers, those are what you check. When there isnt something like that, computers evaluations recently are suprisingly reasonable, although there are strange exceptions, such as here. I just think this is a cute demonstration of what ive been trying to say the whole time.|
|Mar-26-09|| ||tranquilsimplicity: Firstly, I have to admit that the level of debate on ChessBase is dazzling! and in this case between FiveofSwords and Chessman 95 is absolutely riveting . It is a debate of profound intellectual content befitting good Chess players! Without further ado I arrive at my point. Even though I am a beginner and not a strong player by any standards (around 1800 Elo)nor thoroughly conversant with Computer Programming I have to agree with FiveofSwords and my opinion is this: I believe AI/Computers are notoriously accurate in dry calculations but owing to the "infinite-like" permutations that Chess can throw, it is not uncommon for the computer to evaluate a human move as bad and then after a Computer reply without blundering the human makes the next move and computer evaluates the position as very bad(blunder) and then after the human makes their reply the Computer evaluates the game as won for the human! This can happen because the human has a particular strategy in a certain opening where he/she needs to post pieces on particular squares whilst allowing for a particular temporary weakness in his camp where the computer at first judges the moves to be bad but later in the resulting position evaluates that the human's position is indeed won. I have experienced this phenomenon whilst playing the Kings Indian Attack where whilst i'm busy positioning my Knight at h2 and g4 and then aligning my Queen at d2 with my Bishop at f4 aiming at h6, Fritz 11 doesn't think much about my position. Worse still after I sacrifice my bishop at h6! But after my Queen enters the Kingside at h6 from what Fritz analyses as a lost position, it quickly turns into a Won position for me(White). However usually the game ends a draw! I also have the same experience when i play the Dragon (as Black) in a particular variation based on Bg5 Yugoslav variation. Black quickly seems to have a won position based upon a great deal of GrandMaster theory but after about move 16 or 17 moves Fritz calculates dry variations based on sacrifices and I lose (Black) very quickly! Thus my point is in consensus with FiveofSwords that one cannot trust Computer evaluation 100% as how they analyse and how we humans play is different. Computer do not "see" or is it "forsee" or remember patterns they just throw out calculations! However Computer analysis is indispensable and can only enrich the human game.|
|Mar-26-09|| ||chessman95: <tranquilsimplicity> First of all, I would just like to say that I appreciate you additions to this debate, and I fully respect you opinion on this topic, but I would like to state my views on the points you made. |
<it is not uncommon for the computer to evaluate a human move as bad and then after a Computer reply without blundering the human makes the next move and computer evaluates the position as very bad(blunder) and then after the human makes their reply the Computer evaluates the game as won for the human!>
This is very common. However, the situation you refer to sounds like one of those sacrifices where the computer can't see far enough ahead to see the compensation that will come of it. In this case, the solution is just to give the computer a little more time. Example: once I was playing the Exchange Ruy Lopez against a relatively weak computer under a time limit, and we played down the main line to the point where I pinned his king's knight, he played h3, and I played h5. If you know this line then you will know that for white to capture the bishop is a devestating trap that will lead to checkmate in just a few moves. However, the computer could not see this far ahead and took the bishop, giving itself a won position. Two moves later, it gave me a won position and I mated it the next move. My point is that this was not a problem with the computer, it was a problem with the time. If you have read through the last few pages of this debate then you will know that we were talking particularily about computers that evaluate, NOT computers that play. In that case, there is no time limit, and these short traps will not occur.
<This can happen because the human has a particular strategy in a certain opening where he/she needs to post pieces on particular squares whilst allowing for a particular temporary weakness in his camp where the computer at first judges the moves to be bad but later in the resulting position evaluates that the human's position is indeed won.>
I like the point that you pointed out the Yugoslav Attack for an example. The Yugoslav is an opening that looks like just normal developing moves, but really has a plan for mate if black castles king-side. Another example is the English Attack against the Najdorf Sicilian, which can be deadly if black castles king-side. Humans no from experience to keep their king in the center of the board, but a computer that is not powerful enough under the time limit can sometimes make the mistake of castling and then a few moves later it gives itself a lost position. Now this is interesting because there are two solutions: the first is the obvious one. Give the computer more time. The second is available only for self-learning AI computers that we were talking about. After a few lost games of castling king-side in the English Attack, they will stop making that mistake. Situations like these are what make self-learning computers a hopefull prospect for the future.
<one cannot trust Computer evaluation 100% as how they analyse and how we humans play is different.>
I love how you put that statement, and I agree 100%. However, humans using computer analysis have to be smart enough on their own to know when it is and isn't a good time to use a computer to give advise on a position. As I already said, using a computer to analyze the English Attack would be unwise because it is played with a particular strategy in mind that the computer doesn't know. (that would be Be3, Qd2, 0-0-0, f3, g4, h4, etc.) In fact, computers should avoid highly stratigic positions in general. Apart from that however, computers can be very usefull in evaluating basic or tactical positions that DO NOT involve a particular strategy, because they can add up all the factors (like how many times each square is attacked and defended, and then multiplied by a weighting system) faster than any human could. Remember, humans disign computers to evaluate positions as they would if they just had the brain power and the time. A well-used chess engine is in my view like a speed-boost of how a human would evaluate a position if our brains made it possible. Unfortunately, our brains don't make it possible, so we use computers.
|Mar-27-09|| ||tranquilsimplicity: Hello Chessman 95! I understand your point very well and thanks for your brilliant reply. As I pointed out earlier the debate is really entertaining and engaged in by true gentlemen like yourself Chessman 95.|
|Mar-27-09|| ||chessman95: <tranquilsimplicity> Thanks for the great compliment, and I must say you're a gentlemen yourself! It's nice to finally have a good respectful conversation with someone on this site. I see from your number of kibitzes that you're probably fairly new to this site, so you will most likely soon find that there are many people kibitzing here who have absolutely no respect for other people, which they seem to think are their "opponents". Thank's for not being like one of them! :)|
|Feb-14-10|| ||timhortons: its more than 91 games now for GM Sveshnikov.
he is currently playing in france, i tried to upload this game but it wont pass through.
Sveshnikov,Vladimir - Sandipan,Chanda [B22]
26è Open International de Cappelle la Gr Cappelle la Grande / France (1), 13.02.2010
1.e4 c5 2.c3 Nf6 3.e5 Ng8 4.d4 d6 5.exd6 cxd4 6.dxe7 Bxe7 7.Qxd4 Qxd4 8.cxd4 Nc6 9.Bb5 Bf6 10.Nc3 Bd7 11.Nge2 Nxd4 12.Bxd7+ Kxd7 13.0-0 Nxe2+ 14.Nxe2 Ne7 15.Be3 Rhc8 16.Rad1+ Ke8 17.Bd4 Rc2 18.Bxf6 gxf6 19.Nc3 Rxb2 20.Rd6 f5 21.Re1 Rc8 22.Nd1 Rb1 23.g3 Rd8 24.Rxd8+ Kxd8 0-1
|Jul-06-10|| ||rapidcitychess: When the computers come up, I must point out that, as <Once> put it, hyper-smart gold fish. They carry no balance from the last position.|
|Jul-06-10|| ||MaxxLange: that is a huge "psychological" advantage in man v machine matches. The computer can take a terrible loss and just set 'em up again, good as new. The human of course, has to overcome intense emotions of anger, self-dout, and even grief, and try to continue to play top level chess.|
|Jul-06-10|| ||MaxxLange: ANYWAY: what on Earth do I do against the c3 Sicilian? I've never understood or felt comfortable with the 2...Nf6 lines, and, while I have gotten OK positions from the 2...d5 lines, my results have been poor.|
I looked this all up on Fritz 10's opening book, and it was not encouraging. There were huge piles of variations, and the ones I followed for a long way seemed like kind of tough positional defense for Black.
I'm almost at the point of trying to just play 2....e6, and try to learn how to play the French Advance from Black, plus handling whatever ways White will avoid that transposition. I know that 2...g6 is considered a playable sideline, but Theory gives White the nod there, too.
|Jul-06-10|| ||rapidcitychess: Play 2...d5 if you don't want super-sharp. I remember a lot of gambits in 2...Nf6, mainly Danish-like attacks. |
Normally 2...d5 plays with the pieces in a open game, and 2...Nf6 gives me some kind of a Alekhine.
|Jul-06-10|| ||keypusher: <MaxxLange: ANYWAY: what on Earth do I do against the c3 Sicilian? I've never understood or felt comfortable with the 2...Nf6 lines, and, while I have gotten OK positions from the 2...d5 lines, my results have been poor.>|
Well, if you are getting OK positions out of the opening but bad results, maybe you're working too much on openings and not enough on what comes after. :-)
Seriously, though, I had terrible results with the Alapin as White, and I especially hated facing 2...d5. If you're getting decent positions it's probably just random variation that's causing bad results, and if you keep playing the line your results will change...unless there is some other issue, e.g. you hate playing against the isolated d-pawn or something.
I looked at the databases on gameknot and Black has a quite decent score with 2...d5 and outscores White with 2...Nf6. Not surprising: there's just no way 2.c3 can be White's best move.
|Jul-06-10|| ||MaxxLange: <keypusher> obviously there is something lacking in my game, or I'd not be having this problem. But, in stuff like QGD or KID positions with Black, that I also do not get into that often, I play a better middlegame-to-ending phase, than I do in an unfamiliar Alapin. Much better.|
I'm thinking either I have just built up some psychological "thing" about 2 c3, that I need to get over, and/or I lack confidence, knowledge, or experience in the typical kind of structures that the Alapin tends to bring.
|Jul-06-10|| ||MaxxLange: On reflection, I'm leaning toward the theory that I have a personal problem here. I have not actually played against it many times. The first guy who played it against me was much stronger than me (he was 2200, I was 1400), and he beat me, and all the stronger woodpushers in our circle, like a church bell, with 2 c3|
Also, I have almost no experience playing the French, or the slower ....e6 Sicilian main lines, or the Alekhine, or the QGA, and these Black systems seem to be reference points for Black's main line.
I'm OK with taking on the IQP, and I should be eating up the 2...d5 lines. Maybe I just need to hit the woodshed and learn the theory.
|Jul-06-10|| ||keypusher: I am sorry, I was trying to be funny in my last post, but it just came out mean-spirited. In general I think all of us amateurs tend to think more about the opening than we really should, but of course if you're doing badly against a particular opening you are going to be concerned. |
<MaxxLange: On reflection, I'm leaning toward the theory that I have a personal problem here. I have not actually played against it many times. The first guy who played it against me was much stronger than me (he was 2200, I was 1400), and he beat me, and all the stronger woodpushers in our circle, like a church bell, with 2 c3>
A few years back I played in an online Evans Gambit thematic tournament. There was one guy who was stronger than any of us (and much stronger than me) so he won no matter what he played. His superiority was particularly marked in endings.
So in the tournament, against Lasker's Defense after 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.b4 Bxb4 5.c3 Ba5 6.d4 d6 7.0-0 Bb6, he'd regularly play 8.dxe5 dxe5 9.Qxd8+ Nxd8 10.Nxe5 and go on to kill the rest of us. It was like Kamsky playing against experts in some weekend Swiss. I think the guy was just trying to handicap himself to make the tournament a little more interesting. For me, it was a good lesson in what a really good player can do with nothing much. But if I hadn't had so much first-hand experience with the Evans myself, it probably would have given me an exaggerated opinion of the gambit's value.
|Jul-06-10|| ||MaxxLange: oh, I am not offended. what you said needs to be considered, and you haven't played me or seen my games. no foul.|
I've played casual/training games against people much much stronger than me...IM's putting me in Zugzwang, with all the pieces still on the board, when they got tired of crushing me with sacrifices, and so on. It's both discouraging, or inspiring, to lose to someone who totally outclasses you
|Jul-06-10|| ||MaxxLange: Here is the most recent rated game I played in this line. I got into horrible trouble by playing ...Bg4 too early (?), and I barely escaped, after a scary defense, due to White's later errors:|
1. e4 c5 2. c3 d5 3. exd5 Qxd5 4. d4 cxd4
5. cxd4 Nc6 6. Nf3 Bg4 7. Nc3 Qd7 8. d5 Bxf3 9. Qxf3 Nd4 10. Qd1 e5 11. Be3
O-O-O 12. Bxd4 exd4 13. Qxd4 Nf6 14. Rc1 Kb8 15. Nb5 Re8+ 16. Kd2 Qxd5 17. Qxd5 Nxd5
18. Bc4 Bb4+ 19. Nc3 Rd8 20. Kc2 Nb6 21. Bb3 Bxc3 22. bxc3 Rd7 23. Rcd1 Rhd8
24. Rxd7 Rxd7 25. Rd1 Rxd1 26. Kxd1 f6 27. Kd2 Kc7 28. Kd3 Kd6 29. Ke4 Nd7 30.
Bg8 h6 1/2-1/2
|Jul-06-10|| ||MaxxLange: 7...Qd7? was the lemon, says Fritz. 7...Qf5 is correct|
|Jul-06-10|| ||rapidcitychess: <MaxxLange> I played that in a 120/G tourney and got killed, just saving my self by opposite colored bishops.|
A local master, (from where we were playing) suggested that trading the c pawns is bad and if that is done correctly he was sure it was equal. e.g
1.e4 c5 2.c3 d5 3.exd5 Qxd5 4.d4 Nc6 5.Nf3
|Jul-06-10|| ||MaxxLange: in my game.....14 Rc1?? Kb8 gave me a second chance at life|
14 Qxa7 and
|Dec-27-10|| ||LDJ: I'm planning to play the Sicilian more often in OTB play, but at this moment I only know theory about the open Sicilian lines. I'm looking for a playable variation against the Alapin, which doesn't require too much study time. I know that 2...d5 is a common move and gives me a decent game after 3.exd5 Qxd5, but I hate getting into French advance lines by 3.e5 e6 4.d4. Is there a way to avoid that, or should I play another line such as 2...Nf6, 2...e6, 2...d6, ...?|
< Earlier Kibitzing · PAGE 7 OF 8 ·