Main
Date: 13 Apr 2006 14:09:08
From: Chesslib
Subject: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
Chesslib's co-admin played this two games against Rybka:

1 QUEEN'S gam. EXCHANGE VARIATON
Vlaardingerbroek Rob - Rybka 1.0 Beta 32-bit

1:0, 4/12/2006.

1. d4 Nf6 2. c4 e6 3. Nc3 d5 4. Bg5 Be7 5. e3 Nbd7 6. Nf3 O-O QUEEN'S GAMBIT
DECLINED, Orthodox Defense. A turn-of-the-century tabiya. It was studied by
such greats as Lasker, Capablanca, Alekhine and occurred in the overwhelming
majority of games in the World Championship match Capablanca - Alekhine,
Buenos Aires1927. Black's play in the Orthodox Defense conforms to the
classical principles; he tries to gradually equalize, achieving a reliable
but slightly constrained position. The outcome of Black's strategy to a
great extent depends on whether he will solve a problem of his light-squared
bishop which is restricted by its own pawns. 7. cxd5 exd5 8. Bd3 Re8 9. O-O
c6 10. Qc2 Nf8 11. Rab1 Ng6 12. b4 Qd6 13. b5N

[13. Rfc1 Bg4 14. Nd2 Rad8 15. Na4 Bc8 16. b5 Ng4 17. Bf4 Nxf4 18. exf4 Qxf4
19. Nf3 Bd6 20. Qd2 Qxd2 21. Nxd2 Bxh2 22. Kf1 Bf4 23. Rc2 Rd6 24. Nf3 Rh6
25. Kg1 cxb5 26. Bxb5 Rd8 27. Re1 Re6 28. Rd1 a6 29. Bd3 g6 30. Nc5 Re7 31.
g3 Bb8 32. Rb1 Bd6 ...1/2-1/2, Benedetto Abel 2171 - Fernandez Arrue Rube
2017 , Villa telli 2000 Ch CAVM]

13... Bd7 14. Rfc1 Rac8 15. bxc6 Bxc6 16. Bf5 Nd7 17. Bxe7 Qxe7?! +1.33
Myrddin

[17... Rxe7!? 18. Nb5 Qb8 19. Qd3 Rce8 +0.89 Myrddin]

18. Qb3 h5 19. a4 a6?! +2.13 Myrddin

[19... Kf8!? 20. Nb5 a6 21. Na7 Rb8 +1.56 Myrddin]

20. Nxd5 Qd6 21. e4 Rcd8 22. Ne3 Nf4 23. Rc2?! +1.56 Myrddin

[23. d5!? Ne2 24. Kh1 Nxc1 25. Rxc1 +2.13 Myrddin]

23... g6 24. Nc4 Qe7?! +2.13 Myrddin

[24... Qf6!? 25. Bxd7 Bxd7 26. e5 Qe6 +1.72 Myrddin]

25. Bxd7 Bxd7 26. g3?! +1.72 Myrddin

[26. Nce5!? Ne6 27. Rd1 b5 28. axb5 +2.13 Myrddin]

26... Nh3?! +2.13 Myrddin

[26... Bc6!? 27. gxf4 Bxe4 28. Nce5 Bxc2 +1.72 Myrddin]

27. Kg2 Bg4 28. Ncd2?! +1.72 Myrddin

[28. Nce5!? Bxf3 29. Nxf3 Ng5 30. Nxg5 +2.13 Myrddin]

28... Rd7 29. Qc3 Rdd8 30. Ne5 Rc8 31. Qb2 Rxc2 32. Qxc2 Rc8 33. Qb3 Qf6?!
+3.75 Myrddin

[33... Rd8!? 34. Qb2 Ng5 35. d5 Bh3 +2.63 Myrddin]

34. f4 Rc7 35. Ndc4 g5 36. f5 Qd8 37. d5 b5 38. axb5 axb5 39. Qxb5 Rc8 40.
d6 Qf8 41. Qd5 Rd8 42. Rb7 Ra8 43. Nxg4 hxg4 44. Ne5 Qg7 45. Rxf7 Qxe5 46.
Qxe5 Ra2 47. Kf1 +156.50 Myrddin [1:0]


--------------------------------------------------------------------------------

2 QUEEN'S INDIAN def.

Vlaardingerbroek Rob - Rybka 1.0 Beta 32-bit

1:0, 4/12/2006.

1. d4 Nf6 2. Nf3=

[2. c4 a6 3. Nf3 e6 4. e3 d5 5. Nc3�/= Eval]

2... e6�/=

[2... d5 3. c4 e6 4. Nc3 dxc4 5. Bg5 a6 6. a4 Nc6 7. e3= Eval]

3. c4=

[3. e3 b6 4. b3 Bb7 5. Bb2 c5 6. c4�/= Eval]

3... b6�/= 4. a3 Ba6 5. Qc2 Bb7 6. Nc3 c5 7. e4 cxd4 8. Nxd4 Bc5 9. Nb3 Nc6
10. Nb5N

[10. Qe2 Nd4 11. Nxd4 Bxd4 12. Nb5 Be5 13. f4 Bb8 14. Nc3 O-O 15. e5 Ne8 16.
Be3 f6 17. exf6 Qxf6 18. O-O-O Bc6 19. Kb1 Bxf4 20. Bd4 Qg6 21. Ka1 Nd6 22.
Qe1 e5 23. Bd3 Qe6 24. Bf2 Bxg2 25. Rg1 Bf3 26. Rb1 e4 27. Bc2 Bxh2 28. Rg5
Bf4 29. Rg1 Be5 ...0-1, Relange Eloi 2488 - Chabanon Jean-Luc 2428 , Cannes
2000 It;

10. Qd1 Be7 11. Bf4 O-O 12. Be2 Rc8 13. O-O d6 14. Rc1 Ne5 15. Nd2 Qc7 16.
Re1 Rfd8 17. Bf1 Qb8 18. h3 Qa8 19. b4 Kh8 20. Qb3 Rg8 21. Rcd1 g5 22. Bxe5
dxe5 23. Nf3 Qb8 24. Qb2 g4 25. hxg4 Rxg4 26. Nb5 a6 27. Nxe5 Rg7 28. Nd7
Qf4 29. Nd6 Bxd6 ...0-1, Toloza Pablo 2327 - Zarnicki Pablo 2488 , Linares
2000 It (open) (active);

10. Nxc5 bxc5 11. Be3 d6 12. Rd1 Qe7 13. Be2 O-O 14. O-O Rfd8 15. Rd2 Rab8
16. Rfd1 a6 17. f3 h6 18. Qa4 e5 19. Nd5 Nxd5 20. cxd5 Nd4 21. b4 f5 22.
bxc5 dxc5 23. Bc4 Kh8 24. Bxd4 cxd4 25. Re2 fxe4 26. Rxe4 Bxd5 27. Bxd5 Rxd5
28. Qxa6 d3 29. Qc4 Qd7 ...1/2-1/2, Nordahl Helge 2359 - Agdestein Simen
2595 , Molde 7/10/2004 Ch Norway;

10. f3 1/2-1/2, Altanoch Genden 2286 - Yu Shaoteng 2518 , Tianjin 2001 Cup
Tan Lian Ann (cat.5);

10. Be2 O-O 11. O-O d6 12. Rd1 Qe7 13. Bf1 Rac8 14. Nxc5 bxc5 15. Ne2 e5 16.
Nc3 Nd4 17. Qd3 a5 18. Rb1 Nd7 19. Be3 Rb8 20. Qd2 Bc6 21. Nd5 Bxd5 22. cxd5
a4 23. f4 Nf6 24. fxe5 dxe5 25. Bg5 h6 26. Bh4 Rb6 27. Bc4 Rfb8 28. Rf1 g5
29. Bxg5 hxg5 ...0-1, Fridman Daniel 2594 - Jaracz Pawel 2539 , Internet
3/16/2003 Dos Hermanas Qualifier (16-A);

10. Bd3 h6 11. O-O d6 12. Bf4 O-O 13. Rad1 e5 14. Bc1 Re8 15. Qb1 a5 16. Nd2
Bd4 17. Nb5 Bc5 18. b3 Nh5 19. Nf3 Bc8 20. Be2 Bg4 21. Qd3 Nf4 22. Bxf4 exf4
23. h3 Bh5 24. Kh1 Qe7 25. Nc3 Bxa3 26. Nd5 Qxe4 27. Qxe4 Rxe4 28. Bd3 Bxf3
29. gxf3 Re5 ...1/2-1/2, Eljanov Pavel 2596 - Alekseev Evgeny 2609 , Geneva
1/27/2004 It (open);

10. Bg5 h6 11. Bh4 Nd4 12. Nxd4 Bxd4 13. Bd3 Be5 14. Bg3 Qb8 15. O-O-O O-O
16. Kb1 Rc8 17. Qe2 a6 18. Na4 Rc6 19. f4 Bxf4 20. e5 Bxg3 21. hxg3 Nh7 22.
Qe3 Ng5 23. Nxb6 Rxb6 24. Qxb6 Bxg2 25. Qxb8 Rxb8 26. Rhe1 a5 27. Bc2 Bc6
28. b3 Kf8 29. Kb2 Ke7 ...1/2-1/2, Shirov Alexei 2736 - Leko Peter 2722 ,
Monte Carlo 3/20/2004 It "Melody Amber" (blindfold);

10. Bf4 O-O 11. Nxc5 bxc5 12. Bd6 Nd4 13. Qd3 Re8 14. b4 e5 15. Rb1 cxb4 16.
axb4 Re6 17. c5 Rc8 18. f3 Ne8 19. Bxe5 Nxf3 20. Qxf3 Rxe5 21. Bd3 Qh4
1/2-1/2, Delchev Aleksander 2624 - Jaracz Pawel 2540 , Nova Gorica 2/ 3/2004
It (open)]

10... Qb8 11. Bg5 a6?! +0.15 Myrddin

[11... Be7!? 12. f3 O-O 13. Bd3 Ne5 -0.29 Myrddin]

12. Nc3 Nd4 13. Nxd4 Bxd4 14. O-O-O?! -0.68 Myrddin

[14. Qd2!? h6 15. Qxd4 hxg5 16. b3 +0.11 Myrddin]

14... Qe5 15. Bxf6 gxf6 16. f3 Rc8 17. Qd2 Bxc3 18. Qxd7?! -1.46 Myrddin

[18. Qxc3!? d5 19. exd5 Bxd5 20. Kd2 -0.52 Myrddin]

18... Kf8 19. bxc3 Qxc3 20. Kb1 Qb3 21. Kc1 Qxa3 22. Kc2 Kg7 23. Qxb7 Rhd8
24. Rxd8 Rxd8 25. Qc7?! -3.25 Myrddin

[25. Qxb6!? Qa2 26. Kc3 Qa1 27. Qb2 -1.46 Myrddin]

25... Qa2 26. Kc3 Qd2 27. Kb3 a5 28. Qxd8 Qxd8 29. Kc2 Qd4 30. Be2 a4 31.
Rd1 Qc5 32. Ra1 b5 33. Rb1 a3 34. Rxb5 Qf2 35. Kb3 Qxe2 36. Kxa3 Qxc4 37.
Rb2 Qc3 38. Ka2 Kg6 39. Kb1 Qd4 40. h3 f5 41. exf5 Kxf5 42. Kc2 Kf4 43. Kb3
Kg3 44. Rc2 f5 45. Rb2 Qxb2 46. Kxb2 Kxg2 47. f4 Kxh3 48. Kc3 h5 49. Kd4 Kg3
50. Ke5 h4 -156.50 Myrddin [...]

[51. Kxe6 Kxf4 52. Kd5 h3 53. Kd4 h2 54. Kd5 Ke3 55. Ke6 f4 56. Kd5 h1=Q 57.
Ke5 -19.22




 
Date: 16 Apr 2006 00:55:21
From: marcosnpesic
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
It=B4s seems to me that "Rybka 1.0 beta" is much better on blitz games
than on long ones...
Rybka kill Fritz/Junior/Sheredder in all games with time control with
less than 1 hour on my computer...
With 1 hour and 10 minutes (+/-) they began to "draw"... And after 1
hour and 30 minutes Rybka almost always loses to Fritz/Junior/Shereder
(last versions)...

It=B4s seems that by playing "positional" the Rybka engine goes deep in
PLYs than other engines much faster... But it backfire after a while,
since Rybka just doens=B4t calculate every "other" move that it doens=B4t
consider positional... And on long games, the "brutal force" of the
other engines overplay Rybka by seeing much deeper...

Usually I got a 16Ply deep game with a full hour with Fritz, and 20
with Rybka... But after that Fritz Ply goes much ahead of Rybka...
So, re-repeating what I am trying to say: Rybka is much good for short
time controls... And not so much for long games...

In all games I used "default" engine parameters... 768MB of Hash for
each engine... no ending game files... and Fritz 9 book...
I also got similar results using "sugested book options for Rybka"
using bouth Fritz9 and "Turk" book (the one that cames along with
Rybka)...
(All games played "inside" Fritz9 GUI)


On my computer:
---
ABIT AV8
AMD 64 "Venice" 3000+ @ 2700Mhz
2 GB PC3200 "Dual Channel" @ 210Mhz
2x 160GB Samsung SATA (7200 RPM 8MB) on Stripe RAID

Running under Windows XP w/SP2 + Anti virus program
---


I also played Rybka in a "Centaur vs Machine" using my own database...
and crafty at some points... And I can say that Fritz9 does much better
playing against me in that mode than Rybka do...
So I also agree with the first post impression !

XD



 
Date: 13 Apr 2006 15:17:53
From: Chesslib
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
OS and hardware used:

OS Name Microsoft Windows 2000 Professional
Version 5.0.2195 Service Pack 4 Build 2195
OS Manufacturer Microsoft Corporation
System Name NYMIA
System Manufacturer MSI
System Model MS-6380E
System Type X86-based PC
Processor x86 Family 6 Model 6 Stepping 2 AuthenticAMD ~1150 Mhz
BIOS Version Version 1.00
Windows Directory F:\WINNT
Locale United States
Time Zone W. Europe Daylight Time
Total Physical Memory 785,904 KB
Available Physical Memory 472,480 KB
Total Virtual Memory 2,707,040 KB
Available Virtual Memory 2,078,408 KB
Page File Space 1,921,136 KB

So what am I doing wrong with the engine Rybka, or is it really just good in
blitz?

Regards,
Rob.




 
Date: 13 Apr 2006 14:10:45
From: David Richerby
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
Chesslib <[email protected] > wrote:
> These games by the way I played as advanced chess games, I did use
> computer assistance from Junior, opening book, database etc. I saw I
> didn't mention that but I am not that good myself ;-)

Human plus computer is usually stronger than even a very strong
computer playing on its own. A recent Chessbase `advanced chess'
competition[1] was won by a couple of guys rated 1685 and 1398 USCF
playing with a couple of reasonably powerful PCs and standard software
(Fritz, Junior and Shredder or something like that). One of the other
competitors was Hydra, running without human assistance, and it didn't
even make the quarter-finals.

Basically, playing with a computer to help you almost eliminates
tactical mistakes. Since it's not hard to understand more about
strategy than a computer, it's not surprising that your strategic
knowledge plus Junior's blunder-avoidance won you the games.


Dave.

[1] http://www.chessbase.com/newsdetail.asp?newsid=2461

--
David Richerby Laptop Toy (TM): it's like a fun
www.chiark.greenend.org.uk/~davidr/ child's toy that you can put on
your lap!


  
Date: 13 Apr 2006 15:38:42
From: Chesslib
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
"David Richerby" <[email protected] > skrev i melding
news:15o*[email protected]...
> Chesslib <[email protected]> wrote:
>> These games by the way I played as advanced chess games, I did use
>> computer assistance from Junior, opening book, database etc. I saw I
>> didn't mention that but I am not that good myself ;-)
>
> Human plus computer is usually stronger than even a very strong
> computer playing on its own

I missed the fact that he was using a engine as playing aid - my
understanding was that only a database and a opening book was used.



Your human + computer thesis is proven to be right :-) a long time ago, so
if this is the matter here, case closed.



-af-




 
Date: 13 Apr 2006 05:32:04
From: TheKooba
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
You're doing advanced chess, that's all. It's easy (but time consuming)
to beat almost any
engine with help of other good engine if you take a closer look to
variations.



  
Date: 13 Apr 2006 14:41:51
From: Chesslib
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?

"TheKooba" <[email protected] > skrev i melding
news:[email protected]...
> engine with help of other good engine if you take a closer look to
> variations

Those lines where added after the games was played.

-af-




  
Date: 13 Apr 2006 14:38:14
From: Chesslib
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
"TheKooba" <[email protected] > skrev i melding
news:[email protected]...
> You're doing advanced chess, that's all. It's easy (but time consuming)
> to beat almost any
> engine with help of other good engine if you take a closer look to
> variations.
>
The only aid used was a opening book and a database - no engines involved..

-af-




   
Date: 13 Apr 2006 14:40:20
From: David Richerby
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
Chesslib <[email protected] > wrote:
> TheKooba <[email protected]> skrev:
>> You're doing advanced chess, that's all. It's easy (but time
>> consuming) to beat almost any engine with help of other good engine
>> if you take a closer look to variations.
>
> The only aid used was a opening book and a database - no engines
> involved..

Ah. Saying, ``I did use computer assistance from Junior, opening
book, database etc. [...] I used a different machine for looking up
things that helped me in the game and Junior,'' made us think that you
were using Junior's analysis to help you find moves, too.


Dave.

--
David Richerby Salted Sword (TM): it's like a
www.chiark.greenend.org.uk/~davidr/ razor-sharp blade but it's covered
in salt!


    
Date: 15 Apr 2006 09:53:16
From: Chesslib
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
"David Richerby" <[email protected] > skrev i melding
news:8yD*[email protected]...
> Chesslib <[email protected]> wrote:
>> TheKooba <[email protected]> skrev:
>>> You're doing advanced chess, that's all. It's easy (but time
>>> consuming) to beat almost any engine with help of other good engine
>>> if you take a closer look to variations.
>>
>> The only aid used was a opening book and a database - no engines
>> involved..
>
> Ah. Saying, ``I did use computer assistance from Junior, opening
> book, database etc. [...] I used a different machine for looking up
> things that helped me in the game and Junior,'' made us think that you
> were using Junior's analysis to help you find moves, too.
>
Hello Dave,

I sent your answere to Rob and here is his reply:

I can only agree fully with Dave here, but that was the main idea, as I
tested Rybka 1.0 beta.

However it is the way Rybka played that amazed me, I see people all over
Playchess shouting that it is such a good engine, and do see that in the
result I had.

Testing engines in 2h/40, 1h/20 and 30 minutes to end games isn't always
that easy as it was to defeat Rybka. Fritz9 is really way harder to deal
with like this.

Pro Deo is very hard to beat like this. And I like to play advanced in order
to see what an engine can do.
Therefor I wondered do I need any special settings for Rybka, I just
attached the UCI engine. So basically, I didn't wonder that I won, but that
it was that easy ;-)

Arild





    
Date: 13 Apr 2006 15:50:41
From: Chesslib
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
"David Richerby" <[email protected] > skrev i melding
news:8yD*[email protected]...
> Chesslib <[email protected]> wrote:
>> TheKooba <[email protected]> skrev:
>>> You're doing advanced chess, that's all. It's easy (but time
>>> consuming) to beat almost any engine with help of other good engine
>>> if you take a closer look to variations.
>>
>> The only aid used was a opening book and a database - no engines
>> involved..
>
> Ah. Saying, ``I did use computer assistance from Junior, opening
> book, database etc. [...] I used a different machine for looking up
> things that helped me in the game and Junior,'' made us think that you
> were using Junior's analysis to help you find moves, too.
>
>
> Dave.
>
> --
> David Richerby Salted Sword (TM): it's like a
> www.chiark.greenend.org.uk/~davidr/ razor-sharp blade but it's
> covered
> in salt!

Dave

I missed ``I did use computer assistance from Junior...." sorry about
that.

-af-




     
Date: 20 Apr 2006 10:26:56
From: Henri Arsenault
Subject: Re: What am I doing wrong with the engine Rybka, or is it really just good in blitz?
I don<t know the answer to your questions, but it does say on the Rybka
site that the demo is quite a bit weaker than the commercial version, and
it mentions some specific weaknesses.

Henri