Page 1 of 11

Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Wed Jun 23, 2010 8:23 pm
by Martin Thoresen
Information

Rybka 4 x64 4CPU vs Houdini 1.02 x64 4CPU
Both engines use SSE4.2 builds.

Time Control:
120 Minutes For The First 40 Moves
60 Minutes For The Next 20 Moves
15 Minutes For The Rest
+30 Seconds Added Per Move From Move 61

Misc:
Opening Database: AH-48_Basic_Positions.pgn (24 Positions Used)
Tablebases: 3-4-5 Men - Nalimov & Gaviota
CPU: Intel Core i7 920 @ 4050 MHz
OS: Windows 7 Professional x64
Match Length: 48 Games
GUI: Deep Fritz 12
Hash: 1024 MB
Ponder: Off


Live Broadcast

Click Here - Updated Each ~3 Minutes

A full PGN will be available after the match has ended,
meanwhile you can view the games in the PGN-viewer on the
website as the match progresses.

Re: Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Wed Jun 23, 2010 9:30 pm
by S.Taylor
I don't REALLY call that long time control. Perhaps it is tournament time control.

Re: Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Wed Jun 23, 2010 9:37 pm
by S.Taylor
If 18...Rg8 is NOT a book move, AND Houdini does not lose (and fully equalizes), then I will say that Houdini is great!

(actually, on further examination, i'm wondering if 18...Rg8 was actually a regretable concession, and the mistakes were earlier).

Re: Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Wed Jun 23, 2010 10:06 pm
by Albert Silver
S.Taylor wrote:I don't REALLY call that long time control. Perhaps it is tournament time control.
The average game will probably be upwards of 70 moves, which is about 7 hours long. 7 hours is a pretty long game IMHO.

Re: Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Wed Jun 23, 2010 10:07 pm
by Albert Silver
S.Taylor wrote:If 18...Rg8 is NOT a book move, AND Houdini does not lose (and fully equalizes), then I will say that Houdini is great!

(actually, on further examination, i'm wondering if 18...Rg8 was actually a regretable concession, and the mistakes were earlier).
The game score shows they were thinking after move 4.

Re: Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Wed Jun 23, 2010 10:30 pm
by S.Taylor
I deleted my last comment, as only now, I see that a feew more moves were made (only after pressing the refresh button).
I find this game quite interesting, as I personally never knew how to survive as black, in such positions. Maybe he WILL.

(I do not have a program analyzing in front of me, and i am finding this game amazing. I never expected Black to get a counter attack).

Re: Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Thu Jun 24, 2010 12:11 am
by S.Taylor
scary! (at 41.nc6)

Re: Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Thu Jun 24, 2010 1:08 am
by Albert Silver
Martin Thoresen wrote:Information

Rybka 4 x64 4CPU vs Houdini 1.02 x64 4CPU
Both engines use SSE4.2 builds.

Time Control:
120 Minutes For The First 40 Moves
60 Minutes For The Next 20 Moves
15 Minutes For The Rest
+30 Seconds Added Per Move From Move 61

Misc:
Opening Database: AH-48_Basic_Positions.pgn (24 Positions Used)
Tablebases: 3-4-5 Men - Nalimov & Gaviota
CPU: Intel Core i7 920 @ 4050 MHz
OS: Windows 7 Professional x64
Match Length: 48 Games
GUI: Deep Fritz 12
Hash: 1024 MB
Ponder: Off
Without meaning to detract from the high level of the match, isn't 1GB hash a bit low? We are talking about 4GHz x 4, running for 5 minutes or more at the very least. Of course, if that is all there is, no problem, but if not, it is something to consider.

Re: Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Thu Jun 24, 2010 1:51 am
by gerold
Thanks Martin.
This match will be very interesting for me.
With these two being so closely matched
should be a lot of draws and a close match.

Best,
Gerold.

Re: Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Posted: Thu Jun 24, 2010 8:33 am
by S.Taylor
It's not so often recently, that i follow games live (or at all), but now with these, which are easier since my computer supports it without complication, plus, it is of special interest to me and many of us, I get the feeling that these games are almost like human world championship level games.
It is manuevering and counter manuevering, with many visions appearing and then dissapearing, and the feeling that if one of us mortals would play one of the sides, we would be smashed to pieces very soon after the first sub-optimal move, especialy in positions like some of these.
It might STILL be slightly off, from the level of a few human world champion level games. However, I'm not sure, because even if it didn't play the same combination or plan as some great human masterpieces (which i have in mind), it would probably find a different one from the same positions.