| Jun-10-10 | | outsider: i wonder if these silicons could get rated so that we have some comparison. quite a few non-human type moves as well |
 |
| Jun-10-10 | | supertimchan: I think they are about 3200 FIDE. |
|
| Jun-11-10 | | dumbgai: Strange opening by Rybka. |
|
| Jun-11-10 | | zluria: It certainly looks strange, but humans have been known to play it. See for example Carlsen vs Leko, 2009. |
|
| Jun-11-10 | | newton296: look at that !! deep sjeng goes undeafeated 5-0 with a win over rybka !! impressive !! |
|
| Jun-11-10 | | computer chess guy: I think 38. ♖xc5 was inferior to 38. ♗xc5, but Rybka has always been more than willing to sac the exchange. |
|
| Jun-11-10 | | Whitehat1963: So, call me blind, but what's the finish? |
|
| Jun-11-10 | | dumbgai: <Whitehat1963: So, call me blind, but what's the finish?> The operator probably decided to resign because the evaluation was too unfavorable. Various moves leave black with a large advantage. |
|
| Jun-11-10 | | jhoro: i'm guessing rybka was really low on time when it played 38.Rxc5. firebird finds instantly that 38.Bxc5 is better by close to a full point. |
|
| Jun-13-10 | | kingsindian2006: If rybka can lose games than more people should consider this when calling gm's names for not following rybka lines in a chess game. |
|
| Jun-13-10 | | laskersteinitz: Well said! |
|
| Jun-14-10 | | zanshin: <jhoro: i'm guessing rybka was really low on time when it played 38.Rxc5. firebird finds instantly that 38.Bxc5 is better by close to a full point.> Strange, so does my Rybka 3.
<kingsindian2006: If rybka can lose games than more people should consider this when calling gm's names for not following rybka lines in a chess game.> People should not call GMs names for not following any engine - not just Rybka. |
|
Dec-21-10
 | | Domdaniel: Rybka's blunder was seemingly some kind of communication glitch in the huge stack of cores being used -- no normal Rybka makes the same mistake. Rybka still won the event, winning all of its other games. Sjeng finished a point behind, having conceded a few draws to lesser programs. I've seen a pic of the Rybka core stack, and the problem is easy to spot. On the wall behind, instead of a giant fan or a multi-core coolant spray, there's a *radiator*. Heat death. |
|
Dec-21-10
| | Domdaniel: But if this helps to shatter some illusions about the 'perfect' play of the super-engines -- or the total meaninglessness of a 3200 rating in human terms -- then so much the better. |
|
| Jun-30-11 | | Sigmapt: houdine rules and its free http://www.inwoba.de/ |
|
| Jun-30-11 | | msoewulff: rybka has been stripped of its titles... |
|
Kibitzing