|
< Earlier Kibitzing · PAGE 1 OF 2 ·
Later Kibitzing> |
| Aug-30-07 | | psmith: This is really an embarrassing performance by Ehlvest. |
|
| Aug-30-07 | | mrbiggs: This game seems ironic because it contains so many themes that humans are supposed to use against computers. Rybka prevented Elhvest from castling, isolated his king, and traded off any active pieces leaving black with a rook and queen useless in the far corner, while white could easily defeat the isolated king. |
|
| Aug-30-07 | | der623: I am trying to figure out why this was made Game of the Day. I have played over a number of Ehlvest's games, and I do not recall him ever playing as badly as he does here. I don't think that losing to a computer is a good enough reason to call attention to a game. |
|
| Aug-30-07 | | zdigyigy: Ihave to agree with psmith here. Bad play. |
|
| Aug-30-07 | | CapablancaFan: Wow, Rybka plays extremely energetic in the opening. Ehlvest is given pawn odds in which he fails to make the most of. 8...g6 was just simply too slow in terms of development. One alternative (since black's game was cramped and he was badly behind in development) was to return the pawn with 8...Ne4 9.Nxe4 dxe4 10.Qxe4 Nf6 11.Qd3 Be6 with an eye for a future O-O-O. I admit this line is no guarantee but at least it offers some measure of respectable resistance. 12...a4? Here Ehlvest sacs a pawn to open the (a) file for a supposed queenside attack, but this was simply not realistic, especially with his king stuck in the middle (Obviously ...O-O is suicide). In the end, 2 of black's major pieces, queen & rook, sit idle eyeing a pawn it could never attack, while the black king fights for it's life. It was only a matter of time before black's flag fell. |
|
| Aug-30-07 | | epiclurker: to have a rook pawn missing is hardly traumatic to one's game. i mean the rook is instantly in play and under no real threat....i suppose that here it does impede rybka's king side castling -- but the same can be said for ehlvest's. a sort of false odds game to me. |
|
| Aug-30-07 | | psmith: <epiclurker>
well, there is some compensation for the missing pawn, but it is one pawn less. really the game should be lost for White from the initial position here. |
|
| Aug-30-07 | | psmith: And it does say a lot for Rybka that it can beat a 2600+ player with one pawn odds. Amazing, really, even if Ehlvest played poorly. |
|
| Aug-30-07 | | Judah: Hey, the software calls this a Semi-Slav opening. I thought all odds games were listed as 000. Has there been some change? |
|
| Aug-30-07 | | s4life: <zdigyigy: Ihave to agree with psmith here. Bad play.> there is a difference of 400 points in ELO.. Rybka made him look like an amateur though. |
|
| Aug-30-07 | | slowrobot: Does anyone know how to find out what kind of hardware Rybka was running on in this game, the ply of its analysis for each move, or if the logs of its thinking are published anywhere? It'd be interesting to look at, for games like these. |
|
| Aug-30-07 | | aidfarh: I wonder, do computers like Rybka follow opening books in odds games? |
|
| Aug-30-07 | | Fezzik: Wow. This was the sort of game in which we used to see humans beat computers! Rybka was really impressive! |
|
| Aug-30-07 | | kiseiju: In these games it was developed a different book opening for rybka, as I remember I read in a forum. |
|
| Aug-30-07 | | realbrob: Well, Ehlvest was given pawn odds but I actually think that the absence of the White h pawn was an advantage for Rybka, it made easier for it to use the h file to attack the Black king. Rybka's attack starting from move 19 was made possible by pawn odds. Once I read a book by JR Capablanca (I think its name was something like "A primer of chess") in which he said that when you give your opponent pawn odds, you should always take your f pawn off the board. Apart from that, Ehlvest played poorly. It was quite intuitive that Rybka's plan would be to castle queenside and build a kingside attack. He tried to attack on the queenside, but he miscalculated and merely managed to lose a pawn and, what was more important, trap his queen and rook in a place in which they were completely unuseful. I think that humans have some psychological problems playing against engines, because sometimes we think engines are too strong for us and we start playing bad moves. This isn't true, if you don't make mistakes your opponent can be an engine as well - he won't beat you. |
|
| Aug-30-07 | | psmith: 8...g6 seems like a poor move, after Be7, weakening the K-side. Perhaps 8...h6 is a better way to blunt White's h-file pressure. |
|
Aug-30-07
 | | Peligroso Patzer: Curious choice for Game of the Day. In addition to the fact that the game itself seems unremarkable, the pun escapes me. |
|
| Aug-30-07 | | psmith: <Peligroso Patzer> Apparently "Rybka" means "little fish", and there is a Dr. Seuss children's book, "One Fish, Two Fish" -- and Ehlvest is made to look like a fish here. |
|
| Aug-30-07 | | Marmot PFL: I agree with realbrob, not sure moving first and playing without the h pawn is much of a handicap. makes it hard for black to castle kingside if that is what he is accustomed to doing. In effect Rybka has a rook developed on move 1. |
|
Aug-30-07
| | Peligroso Patzer: Thanks for explaining the pun, <psmith>. |
|
| Aug-30-07 | | PAWNTOEFOUR: fischer,karpov and kasparov twisted gm's around like this for years...gm ehlvest has nothing to be ashamed about..this is what you'd expect when your opponent is rated over 3100+ isn't it? |
|
| Aug-30-07 | | amadeus: <kiseiju: In these games it was developed a different book opening for rybka, as I remember I read in a forum.> IIRC, the opening book was minimal. From move 4 on, Rybka was alone. |
|
| Aug-30-07 | | unixfanatic: Does any GM not look like an amateur versus Rybka? |
|
| Aug-30-07 | | Jack Kerouac: Ehlvest has left the building.... |
|
| Aug-30-07 | | psmith: <Marmot PFL>
I am sure that if Rybka had both White and Black pieces and White started with a pawn down, Black would have all the winning chances and if White escaped with a draw that would be interesting.Can RandomVisitor perhaps put this to the test? |
|
 |
|
< Earlier Kibitzing · PAGE 1 OF 2 ·
Later Kibitzing> |
Kibitzing
