4 edits
Originally posted by Northern LadI assume that's this game...
I also played nicohuyboom a number of times. To be honest I never really thought that he was cheating, though it's not always easy to tell. He mostly played pretty strong, basically error-free chess, but a strong correspondence player who has enough time on his hands should be capable of this. In one game against me though, he rather untypically overpre ...[text shortened]... Very much a human frailty that; I very much doubt an engine would have played in that fashion.
Nicohuyboom (white) now plays 23 Qxd5?
Rybka has this move as more than 100 centipawns worse than 23 Nb5, where white is still judged slightly ahead.
1 edit
Originally posted by dottewellRc1 is Fritz's top choice, even after computing for some time. Fritz doesn't rate Bf4 in its top 6 choices. It considers Bf4 to be about 150 centipawns inferior to Rc1.
Here he plays 27 Rc1...Which is also assessed as more than 100 centipawns worse than the alternative 27 Bf4
Originally posted by GatecrasherThat's interesting; I hadn't watched it analyse each move but just left it for an hour or so to look at the game.
Yes, you will get a similar result on most engines. Provided they run long enough. For example, it takes almost a minute before Fritz drops Qxd5 as its top choice.
Just set the position and ran it past a few different engines and see your point; Rybka rejects Qxd5 very quickly. TogaII and Spike II - both very strong engines - are still wedded to it after a minute or more.
Originally posted by GatecrasherFascinating that two of the leading engines can have such a radically different view.
Rc1 is Fritz's top choice, even after computing for some time. Fritz doesn't rate Bf4 in its top 6 choices. It considers Bf4 to be about 150 centipawns inferior to Rc1.
1 edit
Originally posted by dottewellI get a similar result with Shredder. However, if I move 27. Bf4 Qxb1, Shredder likes the outcome. If I then backtrack it uses its hash table, and prefers Bf4 to Rc1. It takes even more moves down the variation for Fritz to realize that Bf4 is a better scoring move.
Fascinating that two of the leading engines can have such a radically different view.
For engines to reach significant depth, they reject many nodes early in their calculations. Here the loss of the rook is something Fritz can't quite come to terms with, and the variation is rejected before sufficient depth is reached.
This is an interesting example because the white moves Bf4, Qe7, Nf7, Ng5 lead to perpetual check, which is why the loss of the rook is considered immaterial.
Whether white would be looking for perpetual check in this position is debatable.
Originally posted by dottewellYes, that's the game. Have to admit I didn't subject it to too much analysis at the time, since most of my moves from that point on were pretty well forced and fortunately ultimately winning!
I assume that's this game...
[fen]rr4k1/3n1pb1/q2N2pp/2pp4/P3P3/4B1PP/1P2RPK1/1R1Q4 w - - id 0-1[/fen]
Nicohuyboom (white) now plays 23 Qxd5?
Rybka has this move as more than 100 centipawns worse than 23 Nb5, where white is still judged slightly ahead.
Originally posted by dottewellDoes Rybka even exist anyway?
I assume that's this game...
[fen]rr4k1/3n1pb1/q2N2pp/2pp4/P3P3/4B1PP/1P2RPK1/1R1Q4 w - - id 0-1[/fen]
Nicohuyboom (white) now plays 23 Qxd5?
Rybka has this move as more than 100 centipawns worse than 23 Nb5, where white is still judged slightly ahead.
Originally posted by dottewellJust looking at this position for a short while; the obvious move seems to be Qxd5. Looking at the position some more Bf4 starts to look more and more interesting.
I assume that's this game...
[fen]rr4k1/3n1pb1/q2N2pp/2pp4/P3P3/4B1PP/1P2RPK1/1R1Q4 w - - id 0-1[/fen]
Nicohuyboom (white) now plays 23 Qxd5?
Rybka has this move as more than 100 centipawns worse than 23 Nb5, where white is still judged slightly ahead.
Nb5 doesn't have any appeal.
So, the fact that anybody played Qxd5 in this position would be no reason to indict them for engine use.
I realize that there is other evidence. I'm just saying Qxd5 in this position is no evidence of engine use even if it is fritz' first choice. It would probably be the first choice of a lot of players and lower rated players in particular.
I think players in the middle band like myself would be more likely to go for Bf4.
I don't know if the higher rated players would be more likely to go for Nb5 or if that is a move that only an engine would pick.
Originally posted by Red NightTrue - but as Gatecrasher has shown, it seems that playing Qxd5 is consistent with engine use (at least use of the main engines). So this move isn't (as was originally supposed) evidence of flawed, human play.
So, the fact that anybody played Qxd5 in this position would be no reason to indict them for engine use.
This does, however, confirm my growing belief that Rybka is some sort of spooky chess genius. Particularly when you consider that I was analysing with the free beta version.
Originally posted by dottewellSo playing natural moves is a capital sin?
True - but as Gatecrasher has shown, it seems that playing Qxd5 is consistent with engine use (at least use of the main engines). So this move isn't (as was originally supposed) evidence of flawed, human play.
This does, however, confirm my growing belief that Rybka is some sort of spooky chess genius. Particularly when you consider that I was analysing with the free beta version.
BTW, IMHO Nb5 is an interesting move.
- J
Originally posted by CrazyLilTingNot at all. I'm not saying it's evidence that he was using an engine. The point is it's not evidence he wasn't using an engine, which is how it was originally presented.
So playing natural moves is a capital sin?
BTW, IMHO Nb5 is an interesting move.
- J