Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka @ Leiden 8th ICT
- - By Vasik Rajlich (Silver) Date 2008-06-01 17:40 Edited 2008-06-02 20:02
The 8th International CSVN Tournament in Leiden has finished and Rybka has won with 8.5/9, ahead of Deep Sjeng, Diep and Hiarcs.

The full Rybka team was:

Operation: Hans van der Zijden
Hardware: Lukas Cimiotti
Opening book: Jeroen Noomen
Engine: Vasik Rajlich, Larry Kaufman

I'd like to thank the organizers for once again pulling the tournament together, and best wishes to Theo van der Storm, who continues to recover from his illness. Kudos also the the Sjeng team for nearly pulling off one of the bigger surprises in recent computer chess.

Rybka's play on the whole seemed ok. There were a couple of issues here and there which we will look into. Larry's latest 12+ months of evaluation work were on display. Jeroen has started his six-month sabbatical and has been putting in extra time preparing for his upcoming release. Rybka enjoyed nine good book positions. Lukas has upgraded his hardware to a Skulltrail running at a cool 4 GHz.

Jeroen's updates from the tournament site can be found here.

All of the games from the event can be found here.

Jeroen's comments about Rybka's games can be found here.

A brief rundown of the games:

Rd 1, Rybka - Hermann 1-0: A long Spanish book line put Rybka into a middlegame position with interesting attacking prospects. They materialized very quickly - black underestimated the danger and Rybka's third move out of book was already decisive.

Rd 2, Crash Test Dummy - Rybka 0-1: White avoided a big theoretical discussion with the quiet 4. Nbd2 and 5. e3, but this game shows the drawback of this type of approach. White never found a good plan for the awkwardly developed queenside - the knight went to f1 to make way for the queen bishop, and by the time it moved again white was already barely holding on. The game was eventually decided in an interesting rook and bishop endgame.

Rd 3, Rybka - Deep Sjeng 1-0: This turned out to be the decisive game of the tournament. Sjeng swept through the rest of his games and would have taken clear first had this game been drawn. The Sjeng team avoided a theoretical debate by going for an an offbeat Spanish line. From a worse but playable queenless middlegame, black played the provocative 18. .. Be7, inviting a long forcing variation which led to a complex rook endgame. It's not clear if this endgame can be drawn with best play - this is the sort of position which a good correspondence player would analyze for several dozen hours. In the game, white seemed to win fairly smoothly, but no doubt there are all sorts of subtleties which both engines overlooked.

Rd 4, Hiarcs - Rybka 1/2-1/2: The Hiarcs team boldly played into the heart of Jeroen's opening preparation and lived to tell about it. Black is probably a tiny bit better in the early middlegame, but Rybka overlooked a deep variation until it was too late (see details), leaving the position quite balanced. White played with energy and purpose after that and never seemed to be in danger of losing the thread.

Rd 5, Rybka - Diep 1-0: The opening seemed very promising for white - better pawns, better knight, and black didn't even have time to set up a dark-squared blockade with .. Bd6. The black position was on the verge of collapse throughout the middlegame, but he hung on and Rybka eventually settled for another debatable rook endgame, this one probably being closer to drawn than to winning. I am not too happy with Rybka's play in this game, but anyway black took the wrong approach and white pocketed the point.

Rd 6, The King - Rybka 0-1: White's opening was not ambitious and gave black easy equality, if not slightly more. After a few inaccuracies by white, black seized control of the d-file and simplified into yet another rook endgame. Unlike the previous two, this one seems fairly clear-cut, and white was duly ground down.

Rd 7, Rybka - Hansdamf 1-0: This is Gerd Isenberg's latest program, replacing IsiChess. It played very ambitiously, wildly sacrificing a pawn in the center for some initiative. Rybka defused it with a positional exchange sacrifice, leaving her with a strong bishop pair and a big pawn center. The queens were traded and black could not stop the march of white's pawns.

Rd 8, Tzjezz - Rybka 0-1: White must have been playing with some sort of search bug(s) and the result in this game was not pretty.

Rd 9, Joker - Rybka 0-1: A long Sicilian book line left Rybka in an interesting queenless middlegame where white has an extra pawn while black has some serious positional advantages. It's going to be hard to pinpoint white's mistakes, but he needs to be accurate in a position like this. His position gradually went downhill as black's pieces took control.
Parent - By Nick (*****) Date 2008-06-01 17:46
Brilliant!  Hearty congratulations on an exhibition tournament.  Well done the mighty team Rybka!
Parent - - By Laszlo (***) Date 2008-06-01 19:10
Hi Vas. Bravo! One question about the interface... Fritz? or Aquarium? (I choose Fritz gui:)
Parent - - By Felix Kling (Gold) Date 2008-06-01 19:43
Because of the book I would guess Fritz. Aquarium is still at beta stage and such a tournament is no good place for testing :)
Parent - By Laszlo (***) Date 2008-06-01 23:26
one question about Aquarium... are you in everyday, interactiv relationship with th ChessAssistant team? Don't need some Aquarium-testers? (I have  a lot of free time nowadays, and I have three machine for testing.
Parent - By FWCC (***) Date 2008-06-01 23:44
Congrats to Rybka team.Hey! Vas you can release Rybka3.0  now  hehehe.
Parent - - By Venator (Silver) Date 2008-06-01 19:56
Yes, it was the Fritz GUI. My book is in CTG format.
Parent - - By Laszlo (***) Date 2008-06-01 21:32
I see. Of course...
By the way, did you enjoy Budapest?
Parent - By Venator (Silver) Date 2008-06-02 05:40
Absolutely. Budapest is a very nice city, with great buildings, a lot of interesting things to do and beautiful women. I even thought about emigrating :-).
Parent - - By Lukas Cimiotti (Bronze) Date 2008-06-01 19:23
This is the hardware I used for Leiden:
The first computer is an Intel Skulltrail with water cooling running at 4 GHz. 2 days before the event a new BIOS came out, that allowed me to overclock that computer to 4.2 GHz. That speed looked stable to me, but cpus got really hot. So I decided to leave it at 4.0 GHz, as this was a tested and very stable setting.
These are the hardware components currently in use:
2x QX9775, Intel D5400XS mainboard, 4x 2 GB RAM PC 6400 by Hynix, Mora II radiator with 8 fans, Laing DDC ultra, 2x Swiftech Apogee GTX cpu coolers, WD Raptor 150 GB for system, Mtron 64 GB and Samsung 64 GB SSDs for tablebases (~120 Gb  total), 1 TB Hitachi HD for data.

I also had a backup computer ready - it played  ~ 1 1/3 games.
Here are the hardware components of it:
Intel Skulltrail Board (D5400XS), 2x Xeon X5460 overclocked to 3.8 GHz, 4x2 GB PC6400 RAM by Hynix, cooling by 2x Zalman CNPS9700 LED (will be replaced by Zalman CNPS 9700 NT as that mainboard isn't capable of regulating 3 pin fans :(  ), Adaptec 19160, Fujitsu SCSI MAU 3147 for tablebases, SATA 300 GB Maxtor for system.
That computer had to play the game against Tzjezz as Jeroen didn't get a connection to my Skulltrail. I am quite shure that Jeroen's laptop would have been sufficient to win against that ??? engine. Hm... trading a queen for a bishop is very strange, indeed.
Anyway - I'm really happy with the result of that weekend :)

Regards,
Lukas

P.S. Hynix PC6400 FB-DIMMs aren't available for normal computers - only for Mac Pros. So I got the Mac Pro versions and exchanged the heatspreaders as it is really hard to find reasonable PC 6400 FB-DIMMs for Skulltrail boards.
Parent - - By Nick (*****) Date 2008-06-01 19:35
Thanks Lukas and well done for being the hardware "engine" behind this legendary win.   May I ask what mhz and timings you get on that RAM?  Also on the water setup machine are you using chilled water?

Best regards,
Nick
Parent - - By Lukas Cimiotti (Bronze) Date 2008-06-01 20:05
Hi Nick,
the RAMs are running at 800 MHz 5-5-5-18. I've tested them at 4-4-4-15 and everything looked OK, but for a tournament I prefer absolute stability.
My water cooling is a Mora II radiator with 8 12 cm fans and a Laig pump. I used to use chilled water, but my chiller (Aquamedic Titan 500) isn't strong enough to cool that Skulltrail comp - it's just fine for one cpu, but 2 cpus are too much for it.
Regards,
Lukas
Parent - - By Nick (*****) Date 2008-06-01 20:12
Thanks again.
Parent - - By Lukas Cimiotti (Bronze) Date 2008-06-01 20:22
My pleasure.

In case you need any hardware advice - please feel free to ask me.

Regards,
Lukas
Parent - - By M ANSARI (*****) Date 2008-06-01 21:06
I am setting up my system now and will try on air first.  I got some Mushkin PC6400's with special heat spreaders for Skulltrail.  What was it in the new BIOS that allowed you to improve on the speed ... was it extra voltage settings?  Also do you have active cooling on the memory?  I hear the memory gets really hot and that could be one thing that can be worked on to get good memory latencies.  Cooling the Quad processors will be an interesting test.  Personally I always believed that chilled water does a better job than Vapor due to better delta curves at high wattage loads ... but that is only with normal off the shelf Vapor.  Vapor can be designed with high wattage in mind that should be able to do better than chilled water.  I guess this will be an interesting experiment.
Parent - By Lukas Cimiotti (Bronze) Date 2008-06-02 07:42
The new BIOS allows a multiplier of 10.5 which wasn't available with the old one. FB-DIMMs need active cooling, so there is an extra fan for that job.
Parent - - By Dhanish (***) Date 2008-06-02 07:03
Hi Lukas,

What will be the space requirements for such a system? Will all components go inside the cabinet? Is there a photo available anywhere?

Regards,
Dhanish
Parent - - By Lukas Cimiotti (Bronze) Date 2008-06-02 09:46
The mainboard requires an EATX case.
Here's a link to a photo of the backup computer: http://img134.imageshack.us/my.php?image=p1010191gf4.jpg

Regards,
Lukas
Parent - By Dhanish (***) Date 2008-06-06 06:36
Hi Lukas,

Thankyou for the photo!

Regards,
Dhanish
Parent - By NATIONAL12 (Gold) Date 2008-06-01 21:34
Kingston have got them.buy from orcalogic.co.uk 2x2 KVR 800Mz DDR2 ECC FBDimm £164.63 inc VAT
Parent - - By pawformation (*) Date 2008-06-01 21:53
"Larry's latest 12+ months of evaluation work were on display"

What about your recent work? Was that not on display?
Parent - By Vasik Rajlich (Silver) Date 2008-06-02 22:02
Some of it, but not yet all. There is still a little bit of work left.

Vas
Parent - - By Gaмßito (****) Date 2008-06-02 02:14
Congratulations Vas, and also to all Rybka team by this victory.  
 
Did you take the risk to play in Leiden with the last Rybka version, or you preferred to play more safe with a previous but more tested version? 
 
By the way, can you please tell me what happened with the rook and pawn module, have you finally worked on this or definitively it will be made for  the following versions? 
 
Regards,
Gaмßito.
Parent - - By Samson101 Date 2008-06-02 06:26
Hey Vasik,
             I am in the process of playing an online tournament on www.chesscube.com   and would like you to help me in using Rybka 2.3.2a MP(64-Bit) in playing my moves. Please let me know if this is possible for you. In return I shall send you some nice chess e-books that can help you to develop the next version of Rybka(Version 3).
  
Parent - - By Vasik Rajlich (Silver) Date 2008-06-02 22:06
Sorry, I'm really busy atm. Good luck :)

Vas
Parent - By NATIONAL12 (Gold) Date 2008-06-02 22:10
you should have been a diplomat.
Parent - By Vasik Rajlich (Silver) Date 2008-06-02 22:05

> Did you take the risk to play in Leiden with the last Rybka version, or you preferred to play more safe with a previous but more tested version?


We went ultra-safe. All of Larry's stuff is in, but not all of the search stuff.

> By the way, can you please tell me what happened with the rook and pawn module, have you finally worked on this or definitively it will be made for  the following versions?


There is no decision yet. We're going to try to make a few small tweaks before Rybka 3, to address problems like we saw in the game vs Diep. Time is very limited so any bigger work will have to wait for Rybka 4.

Vas
Parent - By lonesometheblue (*) Date 2008-06-02 07:26
Congratulations to Vasik Rajlich and Rybka team !..

Sait
Parent - - By Venator (Silver) Date 2008-06-02 10:48
Some comments on the opening:

Rd 1, Rybka - Hermann 1-0: The Ruy Lopez Zaitsev, now where have I seen that before :-). I played with Rybka3.ctg in this game and theory took us to 20.a3, where 20... Nb6 was expected. A bit later Rybka uncorked 22.Nf5! after a long think, I would be very interested to know if other engines are capable to find this clever move as well. Hermann was happy to take the pawn on e4, but that was already the decisive mistake. With an exchange sac white collected many pawns and black could have safely resigned on move 38.

Rd 2, Crash Test Dummy - Rybka 0-1: White played a seldom line with 4.Nbd2 against the Slav, which fails to put any pressure on black. Already after 9 moves black has the more comfortable game and at move 26 a white isolated pawn was lost. The question remains if 28... Rd8 was the right idea, as after that move white was able to force opposite coloured bishops. Still, there was no easy draw and after 35... Rb2 white has problems with his b3 pawn. He lost this pawn a few moves later and after that the game was in the bag.

Rd 3, Rybka - Deep Sjeng 1-0: Sjeng did very well in this tournament and elected the super solid Berlin Wall in the Ruy, to avoid any Sicilian debate. IMO Rybka positioned her pieces excellently, after the modern treatment with h3! and a bishop placed on b2 (from book, of course). Sjeng made the mistake 18... Be7?, where it should have blockaded the K-side instead. Now Rybka was able to exchange into a very pleasant ending with a nice tactical line and remained with the dream ending in the Berlin Wall: extra pawn on the K-side. Without tablebases black was unable to hold this, in the final position a mate in 21 was announced.

Rd 4, Hiarcs - Rybka 1/2-1/2: I was happy to see the line team Hiarcs selected: murky position and black a pawn up. This so called 'Moscow gambit' is very popular amongst the top GM's as well, it is hard to evaluate who is better. In any case, it is a 'one mistake and you are dead' kind of position. Hiarcs played it well, Rybka looked a little better but white had enough recources and active pieces to compensate for the pawn deficit. Maybe 22.a4!? was a better try (maybe somebody can have a long analysis on that move?), after 22.b3 the position became a clear draw rather soon.

Rd 5, Rybka - Diep 1-0: A Petroff was selected by Diep, it has been ages ago since he tried the Najdorf against me :-). 10... Be6 was a surprise, it is very clear that 10... Bf5 is the better move here. Book ended with the strong move 16.Ne1! after which white has a clear edge. Somehow Rybka didn't play it optimally and with b5 and a5-a4 black managed to exchange a lot of pawns and a rook ending that should have been drawn with correct play. Rook endings are still difficult for programs, though, and without TB access Diep went for the white a pawn and lost instantly.

Rd 6, The King - Rybka 0-1: The tradition that Rybka has black against The King was not broken and for the third time they played an English defence. Aready on move 5 white made a slight mistake (5.a3! is correct), after which the surprising 5... d5! equalised instantly. With 14.Re1?! white spoiled the chance for 14.d5! with a completely equal position and 15.Be4 and 16.Bg2? were just bad. Rybka took the d-file, forced her way into another rook ending being a pawn up and white never saw a chance to recover.

Rd 7, Rybka - Hansdamf 1-0: Having enough of all the Ruys I tried 1.Nf3 against the former Isichess, now called Hansdampf. This immediately led to success, as 5... e5 and 6... a6 are certainly not the way to play this. Black went for the exchange with Bf5 and Nc2, but white has huge compensation and it didn't take Rybka long to dominate the black position. A nice game and finally no rook ending :-).

Rd 8, Tzjezz - Rybka 0-1: The white program is still young and logically full of bugs. It played some strange moves and lost quickly.

Rd 9, Joker - Rybka 0-1: Joker played with a book supplied by another chess programmer, yet I was very surprised to see it play into a very topical English attack line of the Najdorf. Joker didn't know 14... Be7!, but found the theoretical moves by itself. Rybka left book after 21... Rbg8 and has nice compensation for the pawn invested. Playchess adherents probably know this position well, it has been played many times before and results show that black has excellent compensation due to the weaknesses on f3 and h2. Rybka quickly grasped the initiative, but Joker put up a good fight and it lasted until move 45 when Joker finally made a big mistake with 45.Rd1? After 45.Nc5 I am not sure if black can win this! I am also not sure 44... Kc4 was the right move, Rybka played this rather quickly at a low ply depth, even when the score just dropped 0.30 pawns. Maybe someone can have a look.

So 8.5 points from 9 games and the 5th Rybka victory in 5 Leiden tournaments.
Parent - - By farseer (*) Date 2008-06-03 00:24
Analysis at move 22 against Hiarcs

Rybka 2.3.2a mp x64
Intel Xeon E5440 @ 2.83Ghz (Quad)
2GB of Hash

  5.00   0:00   -0.21   22.Nd6 Nb6 (8.887) 29
  6.00   0:00   -0.20   22.Nd6 Nb6 23.Bh5 (25.003) 62
  6.00   0:00   -0.09   22.Rh1 Nb6 23.Rh7 (65.658) 122
  7.00   0:00   -0.07   22.Rh1 Nb6 23.Bf3 Nd5 (76.988) 132
  8.00   0:00   -0.03   22.Rh1 Nb6 23.Bf3 Nd5 24.Nd6 (101.269) 150
  9.00   0:00   -0.05   22.Rh1 Nb6 23.Bf3 Nd5 24.Nd6 a6 (141.254) 177
10.00   0:02   -0.05   22.Rh1 f5 23.exf6 Bxf6 24.Bd6 Qg7 25.Nxf6 Qxf6 26.Qc7 (339.255) 156
11.00   0:02   -0.10   22.Rh1 f5 23.exf6 Bxf6 24.Bd6 Qg7 25.Nxf6 Qxf6 26.Qc7 Qf5+ 27.Kc1 (517.636) 183
11.06   0:04   +0.03   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.Qxb5 Bd5 26.Kc2 Qc7 (1.114.689) 240
12.01   0:05   -0.07   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.Qxb5 Bd5 26.Bh5 f5 27.exf6 (1.436.265) 258
13.01   0:08    0.00   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.Qxb5 Bd5 26.Bh5 Rf8 27.Kc2 Kb8 (2.590.932) 305
14.01   0:12   -0.03   22.a4 Nb6 23.Nc5 Rb8 24.axb5 cxb5 25.Qxb5 Bxg2 26.Kc2 Qc7 27.Qa5 Rc8 28.Rg1 Qc6 (3.955.260) 337
15.01   0:14    0.00   22.a4 Nb6 23.Nc5 Rb8 24.axb5 cxb5 25.Qxb5 Bxg2 26.Kc2 Qc7 27.Qa5 Rc8 28.Rg1 Qc6 (4.865.098) 344
16.01   0:24   +0.03   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Rh1 Rh8 26.Rxh8+ Bxh8 27.Nxa4 Nd5 28.Nc5 f5 (8.278.671) 350
17.01   0:36   +0.09   22.a4 Nb6 23.Nc5 Rb8 24.Rc1 Nd5 25.Rh1 Bc8 26.Rh7 Qf8 27.axb5 cxb5 28.Ne4 Bd7 (13.179.667) 366
18.01   0:59   +0.06   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Rh1 Rh8 26.Rxh8+ Bxh8 27.Nxa4 Qd8 28.Nxb6+ Qxb6 (17.788.515) 304
19.01   1:24   +0.09   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Nxa4 Nxa4 26.Qxa4 g4 27.Bxg4 c3 28.Rd3 c5 (26.023.790) 315
20.01   1:59   +0.08   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Nxa4 Nxa4 26.Qxa4 g4 27.Bxg4 c3 28.Rd3 c5 (37.478.444) 322
21.01   4:07   +0.05   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Rh1 Rh8 26.Rxh8+ Bxh8 27.Nxa4 Qd8 28.Nxb6+ Qxb6 (86.080.859) 356
22.01   8:04   +0.05   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Nxa4 Nxa4 26.Qxa4 g4 27.Bxg4 c3 28.Rd3 c5 (177.784.619) 375
23.01  20:32   +0.16   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 gxf3 27.Qc6+ Kb8 28.Bxf3 Bxf3 (445.215.534) 369
24.01  31:18   +0.16   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 gxf3 27.Qc6+ Kb8 28.Bxf3 Bxf3 (675.266.248) 368
25.01  56:39   +0.12   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 gxf3 27.Bxc4 Nxc4 28.Qc6+ Kb8 (1.257.220.201) 378
26.01  166:40   +0.22   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 Bxf3 27.Bxf3+ gxf3 28.Rf1 Qd7 (3.571.435.477) 365

Total analysis time: 206min
Parent - - By farseer (*) Date 2008-06-05 07:50
Analysis update:

  5.00   0:00   -0.21   22.Nd6 Nb6 (8.887) 29
  6.00   0:00   -0.20   22.Nd6 Nb6 23.Bh5 (25.003) 62
  6.00   0:00   -0.09   22.Rh1 Nb6 23.Rh7 (65.658) 122
  7.00   0:00   -0.07   22.Rh1 Nb6 23.Bf3 Nd5 (76.988) 132
  8.00   0:00   -0.03   22.Rh1 Nb6 23.Bf3 Nd5 24.Nd6 (101.269) 150
  9.00   0:00   -0.05   22.Rh1 Nb6 23.Bf3 Nd5 24.Nd6 a6 (141.254) 177
10.00   0:02   -0.05   22.Rh1 f5 23.exf6 Bxf6 24.Bd6 Qg7 25.Nxf6 Qxf6 26.Qc7 (339.255) 156
11.00   0:02   -0.10   22.Rh1 f5 23.exf6 Bxf6 24.Bd6 Qg7 25.Nxf6 Qxf6 26.Qc7 Qf5+ 27.Kc1 (517.636) 183
11.06   0:04   +0.03   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.Qxb5 Bd5 26.Kc2 Qc7 (1.114.689) 240
12.01   0:05   -0.07   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.Qxb5 Bd5 26.Bh5 f5 27.exf6 (1.436.265) 258
13.01   0:08    0.00   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.Qxb5 Bd5 26.Bh5 Rf8 27.Kc2 Kb8 (2.590.932) 305
14.01   0:12   -0.03   22.a4 Nb6 23.Nc5 Rb8 24.axb5 cxb5 25.Qxb5 Bxg2 26.Kc2 Qc7 27.Qa5 Rc8 28.Rg1 Qc6 (3.955.260) 337
15.01   0:14    0.00   22.a4 Nb6 23.Nc5 Rb8 24.axb5 cxb5 25.Qxb5 Bxg2 26.Kc2 Qc7 27.Qa5 Rc8 28.Rg1 Qc6 (4.865.098) 344
16.01   0:24   +0.03   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Rh1 Rh8 26.Rxh8+ Bxh8 27.Nxa4 Nd5 28.Nc5 f5 (8.278.671) 350
17.01   0:36   +0.09   22.a4 Nb6 23.Nc5 Rb8 24.Rc1 Nd5 25.Rh1 Bc8 26.Rh7 Qf8 27.axb5 cxb5 28.Ne4 Bd7 (13.179.667) 366
18.01   0:59   +0.06   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Rh1 Rh8 26.Rxh8+ Bxh8 27.Nxa4 Qd8 28.Nxb6+ Qxb6 (17.788.515) 304
19.01   1:24   +0.09   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Nxa4 Nxa4 26.Qxa4 g4 27.Bxg4 c3 28.Rd3 c5 (26.023.790) 315
20.01   1:59   +0.08   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Nxa4 Nxa4 26.Qxa4 g4 27.Bxg4 c3 28.Rd3 c5 (37.478.444) 322
21.01   4:07   +0.05   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Rh1 Rh8 26.Rxh8+ Bxh8 27.Nxa4 Qd8 28.Nxb6+ Qxb6 (86.080.859) 356
22.01   8:04   +0.05   22.a4 Nb6 23.Nc5 Rb8 24.Bf3 bxa4 25.Nxa4 Nxa4 26.Qxa4 g4 27.Bxg4 c3 28.Rd3 c5 (177.784.619) 375
23.01  20:32   +0.16   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 gxf3 27.Qc6+ Kb8 28.Bxf3 Bxf3 (445.215.534) 369
24.01  31:18   +0.16   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 gxf3 27.Qc6+ Kb8 28.Bxf3 Bxf3 (675.266.248) 368
25.01  56:39   +0.12   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 gxf3 27.Bxc4 Nxc4 28.Qc6+ Kb8 (1.257.220.201) 378
26.01  166:40   +0.22   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 Bxf3 27.Bxf3+ gxf3 28.Rf1 Qd7 (3.571.435.477) 365
27.01  259:14   +0.15   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 Bxf3 27.Bxf3+ gxf3 28.Rf1 Qd7 (5.586.714.846) 367
28.01  468:07   +0.14   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.Qxb5 Bxf3 27.Bxf3+ gxf3 28.Rf1 Qd7 (10.279.842.200) 374
29.01  2337:17 +0.23   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Bxg2 25.f3 g4 26.fxg4 Qd7 27.Rg1 Bc6 28.Rf1 Rf8 (49.253.165.125) 359  TB:1

3530 minutes
Parent - - By Mark (****) Date 2008-06-05 13:26
The best thing about this analysis is that it finally had a tablebase hit after about 60 hours!
Parent - - By farseer (*) Date 2008-06-05 23:01
Yes, I found that amuzing also.  The line has been updated again.

Shallower ply's removed to save space:

30.01  3844:48   +0.22   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Qd7 25.Nxb5 Bc6 26.Nd6 Bxg2 27.f3 g4 28.fxg4 Rb8 (83.366.169.066) 370  TB:1
Parent - By farseer (*) Date 2008-06-06 22:22
Final update:

.
.
.
31.01  5690:30   +0.22   22.a4 Nb6 23.axb5 cxb5 24.Nd6 Qd7 25.Nxb5 Bc6 26.Nd6 Bxg2 27.f3 g4 28.fxg4 Rb8 (123.006.523.601) 368  TB:3
Parent - - By Ricardo_dk (*) Date 2008-06-05 14:47
Good for Rybka team.
The question is, why did the "famous" engines" dont play in the tournament?
I read the engines which played in the tournament, but I can not see: Junior, Schreder, Fritz engines ....
I dont say that these engines are bad .... but could be better if Rybka plays with stronger engines ....?
If I knew about The 8th International CSVN Tournament in Leiden  i could be sure that rybka could win it(100%!!)
What is the propose to the rybka team in playing with engines which are not recognize??
Thanks
Parent - By NATIONAL12 (Gold) Date 2008-06-06 22:32
politics, they cannot be beaten if they dont enter,only Hiarcs has the balls,so well done Mark,Harvey and their team.
Parent - - By Felix Kling (Gold) Date 2008-06-07 08:29
Well, Shredder is playing in Mainz, so Stefan is also not hiding :)
Parent - - By Lukas Cimiotti (Bronze) Date 2008-06-07 10:46
When will that event in Mainz be?
Parent - - By Felix Kling (Gold) Date 2008-06-07 11:34
30.07, 31.07 and 01.08
Parent - - By Harvey Williamson (*****) Date 2008-06-07 11:40
"30.07, 31.07 and 01.08"

For those that need to qualify for Mainz there is a tournament on ICC:

Playing Schedule Computer Section

There will be one qualifier on Friday, June 20, 1pm. The finals will take
place on Saturday, June 21, 3pm.

Also the same weekend on ICC is the Computer World Rapidplay Championship the dates of which were announced 1st. I guess authors will have to chose which one to enter. We have entered the World Rapidplay.

Best Wishes,
Harvey
Parent - - By Banned for Life (Gold) Date 2008-06-07 18:08
It is unfortunate that there are so many events on the computer chess tournament calender that they must be scheduled on top of each other.
Parent - - By Harvey Williamson (*****) Date 2008-06-07 18:11
But you would think ICC would spot the clash if not the people organising the tournaments.
Parent - By Banned for Life (Gold) Date 2008-06-07 18:14
Yes, it would be nice if the organizers could keep track of the half dozen or so events held every year to prevent conflicts.
Parent - By Roland Rösler (****) Date 2008-06-07 11:40
After Rybka 3 release (I hope)! :-) See here (30.07.-01.08.2008).
Parent - - By Banned for Life (Gold) Date 2008-06-07 18:06
He is hiding from Jeroen rather than Rybka. :-)

Regards,
Alan
Parent - - By Vasik Rajlich (Silver) Date 2008-06-08 07:03
Just wait until we unveil our 960 opening book :)

Vas
Parent - By Sesse (****) Date 2008-06-09 04:16
Perhaps we will see the "Noomen opening" in certain Chess960 positions? ;-)

/* Steinar */
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka @ Leiden 8th ICT

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill