General discussion about computer chess...
-
kingliveson
- Posts: 1388
- Joined: Thu Jun 10, 2010 1:22 am
- Real Name: Franklin Titus
- Location: 28°32'1"N 81°22'33"W
Post
by kingliveson » Thu Dec 09, 2010 7:10 pm
ernest wrote:Robert Houdart wrote:...whereas 54.axb5 probably manages to keep a draw (by a whisker!).
Hmm... interesting position!... does it really?...
8/8/2b3p1/1p1k3p/P1N4P/2K3P1/8/8 w - - 0 54
The above position is lost and 54...axb5 leads to a quicker loss to my surprise. I think it might even have been a loss after the following exchanges:
r2r1k2/1p3ppp/3R4/p4b2/4N3/P7/1P3PPP/2R3K1 w - - 3 22
The final standing and games are posted on the tournament sub-forum. Houdini essentially gave Rybka the tournament. Looking forward however to Houdini 1.5 with great anticipation. The estimated released date gives me time to build GTB...now Stockfish guys need to follow with EGTB support.
PAWN : Knight >> Bishop >> Rook >>Queen
-
BB+
- Posts: 1484
- Joined: Thu Jun 10, 2010 4:26 am
Post
by BB+ » Fri Dec 10, 2010 3:30 am
RH: ...whereas 54.axb5 probably manages to keep a draw (by a whisker!).
KLO: The above position is lost and 54...axb5 leads to a quicker loss to my surprise.
With a good choice of parameters, I get a "White is Lost" score within about 30s (mostly I/O) after Bxb5 using IvanHoe with some 6 piece "TripleBases". Black just moves the king in to eat the kingside pawns. After 55. Nd2 Black plays Be2, and after 55. Na3 Black plays Ba6, and will follow up (when possible) with Ke4/f3/xg3/xh4. The bishop is the right colour for the h-pawn. The win after Nb6+ is more delicate (well, at least it keeps more pieces on the board, thus avoiding the TBs), and it seems that the main thing that engines might not understand is how clunky the knight is at handling multiple duties. In some sense, trading off the queenside pawns "looks right" from the standpoint that knights do better with pawns on only one side of the board, but here both pawns fall and White is not able to setup a blockade sufficiently quickly.
-
ernest
- Posts: 247
- Joined: Thu Sep 02, 2010 10:33 am
Post
by ernest » Fri Dec 10, 2010 6:15 pm
Robert Houdart wrote:After 54.Nb6+ the position is lost, whereas 54.axb5 probably manages to keep a draw (by a whisker!).
...I'm certainly going to look into this after the release of Houdini 1.5.
Hi Robert,
I'm afraid that at move 54, it's already too late for Houdini
But at move 53, Houdini can still save the draw with 53. Kd3!
(while what it played, 53. Nc4?, loses)
-
oudheusa
- Posts: 20
- Joined: Tue Jun 15, 2010 9:08 am
Post
by oudheusa » Sat Dec 11, 2010 5:24 pm
Robert Houdart wrote:gaard wrote:That is the approach of some commercial developers. If it's not yours, all the better. I was more interested in release dates, new feature additions, etc. Did that "Crazy Aggressiveness" parameter make the cut?
Houdini 1.5 will be released later this month, stay tuned...
What happened to version 1.4?
-
Robert Houdart
- Posts: 180
- Joined: Thu Jun 10, 2010 4:55 pm
-
Contact:
Post
by Robert Houdart » Sat Dec 11, 2010 5:42 pm
oudheusa wrote:What happened to version 1.4?
There was Houdini 1.0 for which there were the releases 1.01, 1.02 and 1.03.
Now there will be the first release of Houdini 1.5.