Not logged inRybka Chess Community Forum
Forum RybkaChess Help Search Register Login
Up Topic Rybka Support & Discussion / Rybka Discussion / funny position
- - By termodox [de] Date 2007-04-12 14:39
hello friends.

my Rybka 2.3.1 plays bxa6 always...
Also if she analys that position very long...

3b3r/4k3/rp1p1p1p/pPpPpPpP/P1P1P1P1/3K4/8/8 w - - 0 1


but Fritz moves with King and do Remis

(bxa6 leads to mate in 16)

does anybody has good CPU and can analys this position ?
Parent - - By Uri Blass (*****) [il] Date 2007-04-12 15:32
Unfortunately I find it not funny that people do not read old threads when a similiar position was discussed

http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?tid=711

5k1r/8/p3p1b1/PprpPp1p/1PpP1PpP/2P3P1/3K4/8 w - - 0 1


Here is another thread and again Fritz solve it

http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?tid=752

Here is a similiar thread not in this forum

http://216.25.93.108/forum/viewtopic.php?t=12804

I would like to know if Fritz can find that white win or avoid bxa6 in the following position
r1b5/1p1p1ppp/rp1Pk3/pP6/P1P5/8/5PPP/K2R4 w - - 0 1


Uri
Parent - By RobL (*) [nl] Date 2007-04-12 16:53
Why are you so upset if some doesn't know an old thread. Do you really think that your posts are so important that no one can forget them.
If you haven't given the link I had not know either and I am a lurker:)
Best regards
RobL
Parent - By Gaмßito (****) [cr] Date 2007-04-13 00:28
Hi,

Your modified position is a bit harder now.  With a powerful hardware, it is possible that Deep Fritz 10 never thinks to take in a6.

r1b5/1p1p1ppp/rp1Pk3/pP6/P1P5/8/5PPP/K2R4 w - - 0 1


Analysis by Deep Fritz 10:

1.bxa6 bxa6 2.f3 Rb8 3.Rd4 Bb7 4.Rg4 Rg8 5.Rh4 h6
  -+  (-4.89)   Depth: 9/21   00:00:00  98kN
1.bxa6 bxa6 2.f3 Rb8 3.Rd4 Bb7 4.Rg4 Rg8 5.Rh4 h6
  -+  (-4.89)   Depth: 10/15   00:00:00  123kN
1.bxa6 bxa6 2.f3 Rb8 3.Rd4 Bb7 4.Rg4 Rg8 5.Rh4 h6 6.Rf4
  -+  (-4.97)   Depth: 11/23   00:00:00  269kN
1.bxa6 bxa6
  -+  (-5.25)   Depth: 12/26   00:00:01  852kN
1.bxa6 bxa6 2.Ka2 Bb7 3.Rb1 Kxd6
  -+  (-5.32)   Depth: 12/26   00:00:02  1501kN
1.bxa6 bxa6 2.Ka2 Bb7 3.Rb1 Kxd6 4.Rxb6+ Bc6 5.Rb3 Bxg2 6.Rg3 Bf1 7.c5+ Kxc5 8.Rxg7 Bc4+
  -+  (-5.51)   Depth: 13/29   00:00:02  2069kN
1.bxa6 bxa6
  -+  (-5.79)   Depth: 14/31   00:00:04  3346kN
1.bxa6 bxa6 2.Ka2
  -+  (-5.79)   Depth: 14/31   00:00:04  3524kN
1.bxa6 bxa6 2.Ka2 Bb7 3.Rb1 Kxd6 4.Rxb6+ Bc6 5.Rb3 Re8 6.Rg3 g6 7.Kb3 Re2 8.f3 Re3+ 9.Kb2 Bxa4 10.Rg5 f5 11.Ka1
  -+  (-5.99)   Depth: 15/32   00:00:08  7417kN
1.bxa6 bxa6 2.Ka2 Bb7 3.Rb1 Kxd6 4.Rxb6+ Bc6 5.Rb3 Re8 6.Ka3 Re2 7.Rb6 Rxf2 8.Rxa6 Rxg2 9.Rxa5 Rxh2 10.Rg5 Rg2 11.Rh5 h6 12.a5
  -+  (-6.16)   Depth: 16/34   00:00:17  16471kN
1.bxa6 bxa6 2.Ka2 Bb7 3.Rb1 Kxd6 4.Rxb6+ Bc6 5.Rb3 Re8 6.Ka3 Re2 7.Rb6 Rxf2 8.Rxa6 Rxg2 9.Rxa5 Rxh2 10.Rg5 Rg2 11.Rh5 h6 12.a5
  -+  (-6.16)   Depth: 17/35   00:00:24  24438kN
1.bxa6 bxa6 2.Ka2 Bb7 3.Rb1 Bxg2 4.Rxb6 Bc6 5.Rb3 Kxd6 6.Rh3 h6 7.Rh5 Re8 8.Rxa5 Re2+ 9.Kb3 Rxf2 10.Rxa6 Rxh2
  -+  (-6.12)   Depth: 18/36   00:00:46  46121kN
1.h4
  -+  (-6.12)   Depth: 18/36   00:01:09  69587kN
1.h4 R6a7 2.g4 f5 3.gxf5+ Kxf5 4.h5 Ke5 5.Rd5+ Kf4 6.Rd1 Kg5 7.Rd5+
  -+  (-5.98)   Depth: 18/36   00:01:11  71697kN
1.h4 R6a7 2.g4 f5 3.gxf5+ Kxf5 4.h5 Ke5 5.Rd5+ Kf4 6.Rd1 Kg5 7.Rd5+
  -+  (-5.98)   Depth: 19/38   00:01:58  118mN
1.h4 f5 2.bxa6 bxa6 3.Kb2 Bb7 4.Rd3 Rc8 5.Kb3 g6 6.h5 g5 7.g3 Be4 8.Rd2 Rc6 9.g4 Rxd6 10.gxf5+ Bxf5 11.Re2+ Kf6 12.h6 Rd3+ 13.Kb2
  -+  (-6.16)   Depth: 20/41   00:03:58  240mN
1.h3
  -+  (-6.16)   Depth: 20/41   00:06:04  370mN
1.h3
  -+  (-6.00)   Depth: 20/41   00:06:33  398mN
1.h3 R6a7 2.f4 Kf5 3.g3 h6 4.Rd4 Ke6 5.g4 Rb8 6.h4
  -+  (-5.73)   Depth: 20/41   00:08:42  530mN
1.g4
  -+  (-5.73)   Depth: 20/41   00:09:37  586mN
1.g4 R6a7 2.f4 h6 3.Rd5 Rb8 4.h4
  -+  (-5.70)   Depth: 20/41   00:09:48  596mN
1.g4
  -+  (-5.42)   Depth: 21/43   00:14:20  866mN
1.g4 g5 2.bxa6 bxa6 3.h4 gxh4 4.Rh1 Bb7 5.Rxh4 f6 6.Kb2 Rh8 7.Rh6 Rg8 8.Rh4 Kxd6 9.Kb3 Be4 10.g5 Ke5
  -+  (-5.59)   Depth: 22/48   00:24:45  1487mN
1.g4 g5 2.bxa6 bxa6 3.Rb1 Ra7 4.Ka2 Kxd6 5.Rxb6+ Ke7 6.Rh6 Rc7 7.Kb3 d6 8.h4 gxh4 9.Rxh4 Be6 10.g5 Rc5 11.Rxh7 Bxc4+ 12.Kb2 Rxg5 13.Kc3
  -+  (-5.70)   Depth: 23/43   00:45:46  2768mN
1.g4 g5 2.bxa6 bxa6 3.Rb1 Ra7 4.h4 gxh4 5.Rh1 Kxd6 6.Rxh4 Kc5 7.Rxh7 d6 8.g5 Rb7 9.Rh5 Kxc4
  -+  (-5.75)   Depth: 24/47   01:40:38  6138mN
1.h3
  -+  (-5.75)   Depth: 24/56   02:02:42  7478mN
1.h3 R6a7 2.f4 Kf5 3.g3 h6 4.Rd4 Ke6 5.g4 f5 6.Rd5
  -+  (-5.70)   Depth: 24/56   02:29:07  9010mN

(, AMD 12.04.2007)

Regards,
Gambitto.
Parent - - By Vasik Rajlich (Silver) [hu] Date 2007-04-13 13:10
Rybka would need to try every combination of piece placements for black at least once in the search to report a draw score. There are 4 pieces and 24 squares. This is light years away without help from the evaluation.

Vas
Parent - - By vroger007 (**) [be] Date 2007-04-16 18:41
hi Vas,

but wouldn't it be rather easy for the evaluation function to find out that the position is completely closed and that the side with the material advantage cannot capture anything?  In that case the side with material disadvantage should just play waiting moves...

best regards,
Roger
Parent - - By theoak (**) [us] Date 2007-04-16 22:11
Actually the first position was constructed by Roger Penrose (the famous Cambridge physicist and mathematician) and his brother who is an IM and CC World Champion. This was used in Roger's book Shadow's of Mind , a book in which he stated his objections against the so called Strong A.I. The possibility that a computer engine can actually "solve" positions like these and understand them strategically requires a "GIANT" leap in the artificial Intelligence field and it is far distant to what nowadays chess engines can do!

a)Luckily those are very special cases that do not appear often in practise
b)Vasik once said that he does not continue research in an idea that appears to be futile after one month. Believe me...Positions like these are not an one month plan!!!
c)I don't think we have to go that far!!!Why don't you ask evaluation of sacrifices like the one in the Kasparov vs Topalov game in 1999 in Wijk Ann See. Computer programms still need too much work in order to "understand" strategy and tactics.
Parent - By Uri Blass (*****) [il] Date 2007-04-16 23:34
1)The sacrifice of kasparov against Topalov is not winning and topalov could reject it.
This is one of the reasons that many program do not play the sacrifice.

2)I also do not think that humans can see the win in case of kasparov-topalov after accepting the sacrifice.
The win is not trivial unlike the draw in the first position that is posted in this thread.

3)for b I certainly believe you that position like the diagram of fortress draw are not one month plan and it is easy to evaluate them correctly faster.
The main problem is not to evaluate them correctly but to evaluate them correctly in a way that is productive for the playing strength of the program.

Uri
Parent - - By Roland Rösler (****) [de] Date 2007-04-17 04:21
to b) Okay, you have seen my short disput with Vas abaout AI. And nevertheless you come with this bullshit! I will make this suggest: Roger should solve this little issue and when he is ready, he give it to the community of chess programmers. If he have questions, please tell him, that he have to ask his collegue Edward but not Vas! When he is ready, Vas will have an interface and build it in his engine.
PS: But believe me, Roger (and Edward even more) is a smart person. He will never do such crazy things! Super string theory is not quite easy, but strong AI is a dump.
Parent - - By theoak (**) [us] Date 2007-04-17 05:44
Actually I did not have the chance to go over past topics in this forum but I find the idea of creating a thematic category for A.I very tempting. I did not yet understand your opinion in the matter of Strong A.I? The reason that I use Rybka and so are many chess fans in the world (from not-rated to world champion class) is simply because it works!!! I believe that any strong chess engine nowadays passes the so called Turing test!!! I cannot say though, that they are intelligent (at least in the human sense). I also believe that people who share Strong views For or Against A.I should be the first to try to answer this simple (phenomenically!) question. What is Intelligence???
I just pose a question! There is no need for such heated debate!

p.s:The Kasparov -Topalov rook sacrifice was probably as drawish as the simple Nc6 which is what Rybka gives!

   Yours
Parent - - By Roland Rösler (****) [de] Date 2007-04-17 06:45
My opinion abaut AI see here: 1.04.2004 "Why no tablebases?
Oh, I´m not very pleased when reading this answer. The history of AI is the history of desasters only. They have a good story but no successful project. Chess is also none, although it was their darling for a long time. In their dispair, they nowadays share between strong and weak AI and praises their success in weak-AI-projects. But nobody of normal people see here the Intelligence in the weak AI. The best way to ruin the career of the most brilliant brains is to work in AI-projects. Only my opinion.
There is no heated debate! There is some bullshit from a person, who take his bath always luke.
Parent - - By theoak (**) [us] Date 2007-04-17 08:53
But my friend, I was clever enough not to state an opinion on that matter. I initially made a comment on a position , its history e.t.c. I somehow shared the same views with you in the past. Now ,the only thing that I can state is that we might be unable to see something that is not a matter of 20-30 years like many experts believed but in the very distant future. Yet I must admit I am very skeptical in a matter like this. I think Von Neumann's auphorism better suits my beliefs "Whatever A.I means we are many Nobel prizes away". You did not answer my question! What is Intelligence?

    Sincerely
Parent - By Vasik Rajlich (Silver) [hu] Date 2007-04-17 11:26
Intelligence (or articifial intelligence, if you prefer) is whatever you currently don't have an algorithm for :)

Vas
Parent - By Vasik Rajlich (Silver) [hu] Date 2007-04-17 11:27
For a 100% closed position, it might not be that difficult. At some point, I will look into it.

Vas
Up Topic Rybka Support & Discussion / Rybka Discussion / funny position

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill