Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / GM Joel Benjamin vs. Rybka: Game 6
- - By Felix Kling (Gold) Date 2008-01-05 23:25 Edited 2008-01-06 09:02
[White "Benjamin"]
[Black "Rybka"]
[WhiteElo "2575"]
[BlackElo "3100"]
[Result "0-1"]
[GameID "41"]
[UniqID "109298"]
[WhiteClock "0:02:47"]
[BlackClock "1:07:28"]
[Stamp "1150"]
[LastMoves "21.axb5 Ra7 -2.40"]

1.d4 Nf6 2.Nf3 e6 3.Bg5 h6 4.Bxf6 Qxf6 5.Nbd2 d6 6.c3 Nd7 7.e4 g6 8.Bd3 Bg7 9.Nc4 {--out of book--} O-O { N} 10.O-O e5 11.dxe5 dxe5 12.b4 Nb6 13.Na5 Qe7 14.Qe2 c5 15.b5 Rd8 16.Bc2 Qc7 17.Bb3 Bd7 18.a4 a6 19.Qa2 Be8 20.Rfb1 axb5 21.axb5 Ra7 0-1

Current standing: 4-2 for Rybka!
Parent - - By Debaser (***) Date 2008-01-05 23:28 Edited 2008-01-05 23:33
This was fast!

What is happening, is GM Benjamin starting to show some fatigue. Or have been the openings?

GM Benjamin is always bad on time and Rybka has plenty of it in the end. Is this a trick not letting him think on Rybka time? Or is it that time management has not been improved and Rybka still wins with much of the time unused? In the future this will give you some elo points, at least on short time controls ;)
Parent - - By lkaufman (*****) Date 2008-01-06 00:56
     Time management has not (yet) been changed, so this is not a deliberate trick, just the old Rybka ways.
Parent - - By Debaser (***) Date 2008-01-06 13:48
With that yet, do you mean that you will work on this feature for Rybka 3?

Or for later versions, if the competence gets closer
Parent - By Vasik Rajlich (Silver) Date 2008-01-06 14:57
If there is time, we'll take a look at this.

Parent - By Permanent Brain (*****) Date 2008-01-05 23:36 Edited 2008-01-05 23:40
I was late and started to kibitz after the 19th move. It didn't look that bad for White I think. Suddenly, a shock... :-)

Appearantly, the problem, or surprise was 23...Nc8 in the variation, which discovers a 3rd attack at the pinned a5 knight. It is easy for us to see now, after the engines have shown it. Although I guess, in 9 from 10 cases (if not more often), a GM will see such a move while considering 20.Rfb1. But nobody is perfect.

After the possible continuation 22.c4 Rda8 23.Qd2 Nc8:

r1n1b1k1/rpq2pb1/6pp/NPp1p3/2P1P3/1B3N2/3Q1PPP/RR4K1 w - -
Parent - - By George Tsavdaris (****) Date 2008-01-06 00:08
What is 22.Z0?
I can't follow the game :-(
Parent - By Permanent Brain (*****) Date 2008-01-06 00:16
The last move was 21...Ra7, the rest is commentary / engine analysis.

[White "Benjamin"]
[Black "Rybka"]
[WhiteElo "2575"]
[BlackElo "3100"]
[Result "0-1"]

1.d4 Nf6 2.Nf3 e6 3.Bg5 h6 4.Bxf6 Qxf6 5.Nbd2 d6 6.c3 Nd7 7.e4 g6 8.Bd3 Bg7 9.Nc4 {--out of book--} O-O { N} 10.O-O e5 11.dxe5 dxe5 12.b4 Nb6 13.Na5 Qe7 14.Qe2 c5 15.b5 Rd8 16.Bc2 Qc7 17.Bb3 Bd7 18.a4 a6 19.Qa2 Be8 20.Rfb1 axb5 21.axb5 Ra7 { 0-1}
Parent - - By Nelson Hernandez (Gold) Date 2008-01-06 00:27
Actually, while the was novelty move was as indicated, the first new position came on the 13th move: 

[Event "Partida evaluada, 3m + 2s"]
[Site "Sala de máquinas"]
[Date "2002.06.07"]
[White "Tartessos, Fritz 7"]
[Black "Oberberger, Fritz 7"]
[Result "0-1"]
[ECO "A45"]
[WhiteElo "2551"]
[BlackElo "2525"]

1. d4 Nf6 2. c3 e6 3. Bg5 h6 4. Bxf6 Qxf6 5. e4 d6 6. Bd3 g6 7. Nf3 e5 8. Nbd2
Bg7 9. dxe5 dxe5 10. Nc4 Nd7 11. b4 O-O 12. O-O Nb6 13. Ne3 Rd8 14. Qc2 Bd7 15.
b5 a6 16. Rfb1 axb5 17. Bxb5 Be6 18. c4 c6 19. Bxc6 Nxc4 20. Bxb7 Nxe3 21. fxe3
Ra3 22. Rb2 Qe7 23. Rab1 Rxe3 24. a4 Ra3 25. Kh1 Kh7 26. Rf1 Bd7 27. Qc7 Re8
28. a5 Be6 29. Qb6 Rb8 30. Rc1 f5 31. exf5 Bxf5 32. h3 e4 33. Rc7 Qf8 34. Nh4
Rd8 35. Rf2 e3 36. Nxg6 Qe8 37. Rxf5 e2 38. Nf8+ Kh8 39. Ng6+ Qxg6 40. Qf2 Qxf5
41. Rxg7 Rd1+ 42. Kh2 Rxh3+ 43. gxh3 Qxf2+ 44. Bg2 Kxg7 45. h4 Qxh4+ 46. Bh3
Rh1+ 47. Kxh1 e1=Q+ 48. Kg2 Qhf2# 0-1

Funny how two Fritz 7s got to this position!
Parent - - By turbojuice1122 (Gold) Date 2008-01-06 00:49
Interesting that you have a computer game with this--I was only able to find four human games, in which the novelty was 11...dxe5, and no computer games...
Parent - - By miles (**) Date 2008-01-06 02:25
" A knight on the rim is very dim." 13.Na5? is a dubious move, Ne3 seems more to the point. I would not have played 12.b4 either.
Parent - - By utzinger (*) Date 2008-01-06 13:08
Your comments are almost the same as mine
Parent - - By Josef (**) Date 2008-01-06 17:42
Hi Kurt,

in your game list I see one game where you achieved remis with Black against Rybka 2.3.2a on a Quad. Was this a fair game?

Parent - - By utzinger (*) Date 2008-01-09 12:15
Hi Josef
This was a (very) fair game. You can replay some further games I have
played vs strong computer programs, most of them under the suspicious
eyes of my chessfriend Rolf Bühler as operator.
From this experience I can (still) hardly believe why GM's seem to be
unable to achieve more draws. The best try was done by Kramnik, who
unfortunately did at some moments not hesitate to play for a win.
Kind regards
Parent - - By Uri Blass (*****) Date 2008-01-09 13:43 Edited 2008-01-09 13:47
Hi Kurt,

Are you the same player who got only 1 out of 4 against humans with average fide rating of 2140 based on the following table?

If yes then I do not understand how you can lose against these humans.
Did you try to play against them in the same way that you play against chess programs?

Parent - By utzinger (*) Date 2008-01-09 21:21
Hi Uri

You are right: it's hard to understand at first glance. But I have never played
against humans in the same way that I do against chess programs. When
fighting vs computer programs I know from the very beginning that I have
in principle no chance at all with may normal playing style and this makes
a very big difference in so far as I do not take the slightest risk. And there
is a further important thing to mention: against humans I need usually
much more thinking time and I often get in serious time trouble with then
the corresponding negative consequences. This happens relatively seldom
in games vs computer programs. I have no satisfactory answer why this
is so - it must be something psychological. By the way: some other games
of me vs computer programs can be found under the following address:
Kind regards
Parent - - By SR (****) Date 2008-01-09 15:18 Edited 2008-01-09 15:20
I have a quite good score against my Rybka (including many wins) however there are three issues here.  Number of CPUs and their speed, opening book and contempt. Despite my successes I must admit that on a faster machine and with unfamiliar openings I would have a VERY hard time. I would be surprised if you could draw as easily as you suggest if you play against Rybka on a fast quadro with a hostile opening book and possible some moderate contempt. 

Parent - - By turbojuice1122 (Gold) Date 2008-01-09 15:32
Wow, that sounds quite good; what is your elo or rough playing strength?

EDIT: Never mind--I've found it--2300 :-)
Parent - - By SR (****) Date 2008-01-09 17:33
When I played I had an elo of 2300, but this was ages ago, and when I look at the games I played then, I am shocked about the low quality of many of the games. I think I learned a lot from playing strong programs especially from playing programs like Rybka, though I am not sure how well I would do against humans these days.
Parent - By turbojuice1122 (Gold) Date 2008-01-09 20:17
Haha, I'm quite the opposite--when I look back at my games from a time when I was around 2200 level or so, I'm shocked that I actually played such games.  I'm very sure that I can't now.
Parent - - By Uri Blass (*****) Date 2008-01-09 23:44
Who knows?

Maybe you can be the world champion if you play against humans.
You cannot know if you do not try.

Parent - By SR (****) Date 2008-01-10 17:58
Oh no... I have played enough games agaist humans to know that any idea of a serious chess career would not bring enough bread on the table.
Parent - By billyraybar (***) Date 2008-01-09 15:42
Do you have any games against more current engines?
Up Topic Rybka Support & Discussion / Rybka Discussion / GM Joel Benjamin vs. Rybka: Game 6

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill