Author: Drexel,Michael
Date: 11:40:48 01/28/04
Go up one level in this thread
On January 27, 2004 at 16:52:59, Mike S. wrote:
>On January 27, 2004 at 03:07:02, Reinhard Scharnagl wrote:
>
>>a) an intelligent engine, which could navigate itself through any opening phase
>
>The Rogozenko match has shown that nowadays engines (even if not especially
>adapted for FRC or Shuffle) are most probably stronger than a "normal" GM at
>this. Rogozenko lost that match against Tiger 15, even although he was allowed
>to use a computer for tactical assistance!
GM Rogozenko might not be able to win against computers in FRC with tactical
assistance.
You conclude from one match (specific player vs. specific program) that
a normal GM is "most probably" weaker.
The same person would cry out loud if someone would draw the conclusion from one
short match Ruffian 2-Shredder 8 that ended clearly in favour of Ruffian 2:
Ruffian 2 is most probably stronger than Shredder 8.
>
>You always argue like engines would be complete unable to play openings
>themselves. This is just plain wrong and provides false information for new
>computerchess fans. It's just that the level of play may be somewhat lower
>*sometimes* (in a minority of cases IMO), maybe 2300 instead of 2600 in extreme
>cases, i.e. in difficult long range gambits. Man needed *decades* of opening
>theory and practise to explore such gambits. Often, much of the old analysis is
>wrong, refutations are found again and again. So why expect from chess engines,
>that they find all this correctly in 3 minutes?? :-))
>
>In general, engines will be better than IMs and GMs anyway, when "normal"
>(normal for computers means very deep) tactical things have to be calculated in
>the opening. Im not talking about Kasparov, but "normal" GMs. Most engines know
>the common opening principles quite well (different quality of engines
>undisputed).
Man, you have obviously no clue about what you are talking here.
Two of the three best programs according to SSDF (Fritz and Junior) know nothing
about common opening principles.
Junior moves its Qeen around in the opening like someone who hasn't learned yet
how to move with knights and bishops and Deep Fritz plays sometimes even worse:
http://talkchess.com/forums/1/message.html?345123
I would call a human who plays these moves a Novice. Other chessplayers call
them Patzers (or even Idiots).
Even Shredder 8 is not able to play a decent opening completely without book
moves. However with a small book it will relatively often come up with something
reasonable.
[Event "Blitz:15'"]
[Site ""]
[Date "2004.01.28"]
[Round "?"]
[White "Drexel"]
[Black "Shredder 8"]
[Result "*"]
[ECO "D06"]
[Annotator ""]
[PlyCount "18"]
[TimeControl "900"]
{128MB, Empty.ctg, ATHLONXP} 1. d4 {0} d5 {0.20/15 32} 2. c4 {(Bf4) 2} Nf6 {
0.18/14 21} 3. cxd5 {(Nf3) 2} e6 {0.28/14 27} 4. dxe6 {2} Bxe6 {0.37/14 18} 5.
Nf3 {3} c5 {0.42/13 27} 6. Nc3 {(e4) 36} Nc6 {0.39/14 22} 7. e3 {9} Be7 {
0.40/13 4} 8. Be2 {(Bd3) 20} O-O {0.47/13 16} 9. O-O {7} Re8 {0.50/12 10} *
2...Nf6?! is a bad opening move. 3...e6?? is an absolute scream!
Shredder got no compensation for the pawn at all. I just won an important
central pawn.
The rest is just a matter of technique as the GMs say.
Of course most strong human players blunder at times in the opening phase of a
game. They seldom play moves like 6.Bb5?! in the next example however:
[D] r1bqkb1r/pp2pppp/2np1n2/8/3NP3/2N5/PPP2PPP/R1BQKB1R w KQkq - 0 6
Engines which suggest such inferior moves are only suitable for blundercheck in
the opening.
Michael
[Event "Blitz:5'"]
[Site ""]
[Date "????.??.??"]
[Round "?"]
[White "Olithink 4.12X "]
[Black "Deep Fritz 8"]
[Result "*"]
[PlyCount "117"]
[TimeControl "900"]
{64MB, Empty.ctg, ATHLONXP} 1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 d5 5. Nxc6
bxc6 6. exd5 e6 7. dxc6 Qxd1+ 8. Kxd1 Ne7 9. Bb5 f6 10. Ke2 a6 11. Ba4 Nd5 12.
Rd1 Bb4 13. c4 Ne7 14. a3 Ba5 15. Be3 Bc7 16. h3 Ng6 17. Nc3 Ne5 18. b3 O-O 19.
Bc5 Rf7 20. Ne4 h6 21. Nd6 Rf8 22. f4 Nf7 23. Nxc8 Rfxc8 24. Kf3 g6 25. b4 e5
26. g3 exf4 27. gxf4 g5 28. fxg5 Nxg5+ 29. Kg4 Ne4 30. Bd4 Kf7 31. Kf3 Ng5+ 32.
Kg2 Rg8 33. Bb2 Ne4+ 34. Kf1 Rae8 35. Rd7+ Re7 36. Rxe7+ Kxe7 37. Rd1 Ke6 38.
c5 Rg3 39. Rd7 Rf3+ 40. Ke2 Rxh3 41. Rxc7 Rh2+ 42. Kd3 Nf2+ 43. Kd4 Nd1 44.
Bxd1 Rxb2 45. Bg4+ f5 46. Bxf5+ Kxf5 47. Rf7+ Ke6 48. Rf8 Rd2+ 49. Kc4 Ke7 50.
Rb8 Rd7 51. Rb7 Rxb7 52. cxb7 Kf6 53. b8=Q Kf7 54. c6 a5 55. c7 Kf6 56. c8=Q a4
57. Qf4+ Kg6 58. Qe8+ Kh7 59. Qff7# *
>
>Again, take a look at the Rogozenko match. See i.e.
>
>http://www.romanianchess.org/rogozenko/tiger15.html
>http://ccn.correspondencechess.com/pdf/ccnews91.pdf
>
>A report was in CSS 3/2003:
>
>http://www.computerschach.de/303web.htm
>
>I'd like to hear a comment about that from you. It seems to contradict to what
>you always tell about engine's weaknesses without opening book. This is just not
>an up-to-date view.
>
>Opening books are IMO not for lifting a chess program from an opening level of
>i.e. 1800 to 2600, but I'd say as a rough guess, to lift it from an average of
>~2400 to 2650, nowadays, for events where every bit of performance is crucial.
>So the question would be if there is really much improvement necessary regarding
>this.
>
>Regards,
>Mike Scheidl
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.