|
< Earlier Kibitzing · PAGE 6 OF 6 ·
Later Kibitzing> |
| Jan-29-12 | | Ulhumbrus: <<Ulhumbrus> Nxd5 Bf7, Nxf6+> 30...Qxf6 31. Bb2 and White has an edge :-)> In fact Black is in trouble because White threatens Qc3. This suggests that in the sequence above instead of 29..Bf7, 29...Ne4 may be better. |
|
| Jan-29-12 | | capanegra: It is a draw. All WV has to do us put his♔ in a3 and shuffle his ♗ through the b8-f4 diagonal. There is no way Black can improve. |
|
| Jan-29-12 | | SoundWave: This looks drawn now. |
|
| Jan-29-12 | | frogbert: capanegra, it's worse actually. :o) all white needs to do, is put his king on a1. he doesn't even need to defend his g-pawn - he can simply sacrifice the bishop for black's f-pawn and hold the draw against B and two pawns. |
|
| Jan-29-12 | | Kinghunt: White could even sac his bishop for the black f pawn and still draw. |
|
| Jan-29-12 | | whiteshark: Without kingside pawns white even needs his ♗, ♔a1-b1 secures the draw. |
|
| Jan-29-12 | | timhortons: magnus tried to squeeze blood out of a turnip, nice try but its not hard to see its a draw. |
|
Jan-29-12
 | | Penguincw: A draw it is! Geez. I wouldn't really say short games (3 hours) but more like they're in a rush or something. We still have Kamsky-Topalov and Caruana-Gelfand. I perfer Kamsky-Topaov. |
|
| Jan-29-12 | | capanegra: <frogbert> So true. :) |
|
| Jan-29-12 | | whiteshark: Caruna is still playing against Gelfand. A win will bring him shared 2.-4. place. |
|
| Jan-29-12 | | kia0708: oh noooo, one more draw today |
|
Jan-29-12
| | Penguincw: In Group B, Bruzon compromised L'Ami chances of winning. |
|
| Jan-29-12 | | hoodrobin: Vanitas vanitatum, et omnia vanitas. |
|
Jan-29-12
| | Penguincw: Carlsen, in the final position, up two pawns, but pawn structure matters. Double a-pawns aren't helping. |
|
| Jan-29-12 | | waustad: Wrong color queening square to even consider playing on. |
|
| Jan-29-12 | | hoodrobin: Caruana emulating MC? |
|
Jan-29-12
 | | paulalbert: Interesting how inept computers are in these kind of endgames unless they get to a position which can use a built in table base. I plugged this final position into Crafty on ICC. It concludes black is winning by -1.61. The computers also will exhaust every variation at the end to avoid three fold repetition and then advance P h6 to h7 until finally the 50 move rule mandates a draw. Knowledgeable humans of course see the wrong colored bishop issue immediately and avoid 100 moves of useless maneuvering. |
|
| Jan-29-12 | | hoodrobin: <paulalbert> Unfortunately you have to reach the endgame to beat the computer (speaking of myself of course). |
|
Jan-29-12
| | paulalbert: <hoodrobin> Probably even more true of me at my age than you, especially if you try to take on a computer at its higher level of play. At lower level settings if you get through the middle game without a tactical blunder, you can benefit sometimes from computer endgame ineptitude. Of course that's not really a meaningful result, but who likes to lose every time, even if it's supposed to be instructive. |
|
| Jan-29-12 | | hoodrobin: <paulalbert> I agree, thanks. I appreciate your profile, it's very instructive (but I'm not in my teens).
:) |
|
Jan-29-12
| | Penguincw: Well, if you're not going to do any more live games, then thanks for the broadcast. :) |
|
Jan-29-12
 | | AylerKupp: <paulalbert> Don’t be so harsh on the engines. Their evaluation depends on both the engine and the position. With my version of Crafty 23.4 it's evaluation was [-1.33] at d=37. Here is what other engines came up with after about 30 secs of calculation (I'm a chess engine junkie): Critter 1.4: [-0.61], d=28
Houdini 1.5c: [-0.33], d=27
Komodo 3.0: [-0.38], d=28
Protector 1.4: [-0.04], d=32
Rybka 4.1: [-0.96], d=28
Spark 1.0: [-0.71], d=29
Spike 1.4: [-0.02], d=28
Stockfish 2.2.2: [-1.37], d=36
Toga II 2.0: [-1.03], d=26
So at least Protector 1.4 and Spike 1.4 (maybe it's that 1.4?) got it right. But that's this position. With different positions I suspect that those 2 engines would get it wrong and other engines would get it right. That's why I believe that when analyzing a position you need to try several engines, and then review their suggestions carefully before coming to any conclusions. |
|
Jan-29-12
| | paulalbert: The disparity of evaluation is interesting. Actually I have no computer engine at all, so going to Crafty on ICC or playing computer on Chessgames here from positions is all I can do. |
|
Jan-29-12
| | AylerKupp: Yes, all engines give somewhat different evaluations since they all have different evaluation functions. And, because chess engines are non-deterministic, particularly multi-processing engines, the same engine on different days will give different evaluations of the same position (although not as different as the evaluation from a different engine). That's why I always try to run 3 or 4 different engines on the analysis of each position and see how well they agree. I'm very fortunate that I have 2 computers that I can use for this. You should consider downloading one or two engines and play around with them. Many of the best ones are free, as is the Arena GUI. |
|
| Jan-30-12 | | timhortons: < I'm very fortunate that I have 2 computers that I can use for this.> was there a point where magnus or van wely make a huge advantage in this game? its vain to call players names just because the game are drawn. |
|
 |
|
|
< Earlier Kibitzing · PAGE 6 OF 6 ·
Later Kibitzing> |
Kibitzing
