Cluster Rybka vs. Houdini 3 - Game 5 ...

[Lavir]

Houdini decided to not play 22... Nxe5 (that was the move expected by Rybka), probably just because it wants to go for a win.

Btw Nakamura (Sart Wars) was watching the game

[Houdini]

Quite a surprise that Houdini 3 avoided the Nxe5 line!

[Lavir]

Ops, just like in the first game there's a problem with the cluster.

Now if Rybka loses many will surely bring the excuse that it was the reload that caused the loss (as it happened for the first game).

EDIT: Clemens stopped the pondering of Houdini 3 at the same time, so they got the same "disavantage" from the disconnection (probably even more for Houdini because 23 g4 was the move pondered and H3 did reach a lot of depth on it).

[Lavir]

Maybe this game Rybka can hold to a draw. If Rybka plays well it is difficult to make the extra pawn count in this position.

This is my impression on it, naturally.

[Lavir]

There are some moves of the cluster that I really cannot comprehend. 33 Bb5 is one of those (and the aforementioned 19 Ra2, plus a lot other examples in the other games).

Does Rybka 4.1 plays those moves too? I will have to check.

A once defendable game has now turned into a mess, and what's worse is that Rybka seem to don't either see it.

[Houdini]

Indeed Rybka appears utterly clueless.
The current position after 34...Rb8 which Houdini 3 evals at +2.

[d]1r4k1/np3pp1/1r2p2P/nBRpP1B1/2pP3N/2Pb1P2/7P/4R1K1 w - -

[Lavir]

I sincerely don't remember Rybka 4.1 playing so badly from my tests. Or maybe it is just that Houdini 3 on 16 cores is completely another beast than what I can have with my puny 4 cores.

Anyway somebody should throw the sponge for the poor Cluster because it is becoming a massacre.

[Houdini]

The white Rook at c5 doesn't play anymore.
Position after 40...Kg6:

[d]8/n4p2/r1p1p1k1/2RpP3/2pP3B/2Pb1P2/1r4NP/2R3K1 w - -

[Lavir]

Both rooks are out of the game, the bishop is terrible and the knight has no scope.

Houdini completely dominates the board now.

[Thanar]

Yes, the box Houdini realized it could eventually build around Rybka's rook in every tenable continuation must be the reason why Houdini evaluated the earlier position at +2.