Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Infinite Analysis / Lack of ply information
- - By Monicker (**) Date 2008-08-07 21:05 Edited 2008-08-07 21:25
I had originally brought this up in Aquarium forum, and was told that the issue was not with the gui.  I tried to search the forum for more details but got no results.  It is now my understanding that for some reason Rybka only reports 2 ply of a variation once the score is +5 or more.  An infinite analysis session which I had started last night seems to bear this out, and was the reason why I had originally posted about it.

Is there a chance that this behavior will be altered in the future?  I am currently using Rybka for a correspondence style game on one of the internet chess site, and being able to see more than 2 ply for multiple variations would be desirable.
Parent - - By Apples (*) Date 2008-08-07 21:24
Hmm, you are right, I just tested Rybka3 in Fritz and only shows 2ply when the evaluation is +5.
Parent - By Juggernaut (**) Date 2008-08-07 21:26
Try fixed depth, it will keep on going...
Parent - - By Vasik Rajlich (Silver) Date 2008-08-07 21:37
Yes, we'll clean this up for Rybka 4. It shouldn't really matter much - when you have +5, the exact moves shouldn't be too crucial.

Vas
Parent - - By George Tsavdaris (****) Date 2008-08-07 22:25
This is not exactly right. I will try to find positions about it.

Also i have to ask one question about the famous position(Qxg6 mates in 15, other moves probably lose):

First: Rybka 3 32-bit single CPU can't solve it quickly(<3 min).
1r4r1/3q1npk/2b1pbnp/Rp1p4/1N1P3P/2PQ1pP1/1K3B2/5B1R w - - 0 1


Second: But that's not the strange thing.
The strange thing is that after forcing 1.Qxg6+ Kxg6 2.Bd3+ Kh5
Rybka's behavior in analysis seems too slow. Seems it is analyzing slowly, like having a lag problem or something.
I wonder if it's my impression only.....
1r4r1/3q1np1/2b1pb1p/Rp1p3k/1N1P3P/2PB1pP1/1K3B2/7R w - - 0 1


Analysis by Rybka 3 1-cpu 32-bit :

3.g3-g4+ Kh5xg4
  -+  (-7.11)   Depth: 2   00:00:01
3.g3-g4+ Kh5xg4
  -+  (-6.85)   Depth: 3   00:00:01
3.g3-g4+ Kh5xg4
  -+  (-6.95)   Depth: 4   00:00:01
3.g3-g4+ Kh5xg4
  -+  (-6.97)   Depth: 5   00:00:01
3.g3-g4+ Kh5xg4
  -+  (-7.11)   Depth: 6   00:00:01  14kN
3.g3-g4+ Kh5xg4
  -+  (-7.39)   Depth: 7   00:00:02  46kN
3.g3-g4+ Kh5xg4
  -+  (-7.64)   Depth: 8   00:00:07  203kN
3.g3-g4+ Kh5xg4
  -+  (-6.40)   Depth: 9   00:00:19  624kN
3.g3-g4+ Kh5xg4 4.Rh1-g1+ Kg4-h3 5.Ra5-a1 Rb8-a8 6.Ra1-f1 Qd7-c7 7.Rg1-h1+ Qc7-h2 8.Rh1xh2+ Kh3xh2 9.Nb4xc6 Nf7-d8 10.Nc6xd8 Bf6xd8
  -+  (-4.89)   Depth: 10   00:00:40  1338kN
3.g3-g4+
  -+  (-4.69)   Depth: 11   00:00:43  1441kN
3.g3-g4+
  -+  (-4.49)   Depth: 11   00:00:44  1448kN
3.g3-g4+ Kh5xg4 4.Rh1-g1+ Kg4-h3 5.Ra5-a1 Qd7-d6 6.Nb4-c2 Bf6xh4 7.Rg1-h1+ Qd6-h2 8.Bf2xh4 g7-g5 9.Bd3-f1+ Kh3xh4 10.Rh1xh2+ Kh4-g3 11.Rh2-h3+ Kg3-f4 12.Rh3-h5 g5-g4 13.Nc2-b4 Rb8-a8
  -+  (-4.74)   Depth: 11   00:00:52  1727kN
3.g3-g4+
  -+  (-4.54)   Depth: 12   00:01:02  2046kN
3.g3-g4+
  -+  (-4.34)   Depth: 12   00:01:09  2232kN
3.g3-g4+
  -+  (-3.94)   Depth: 12   00:01:23  2668kN
3.g3-g4+
  -+  (-3.14)   Depth: 12   00:01:24  2695kN
3.g3-g4+
  -+  (-1.54)   Depth: 12   00:01:25  2744kN
3.g3-g4+
  +-  (1.66)   Depth: 12   00:01:31  2939kN
3.g3-g4+
  +-  (8.06)   Depth: 12   00:01:44  3456kN
3.g3-g4+
  +-  (20.86)   Depth: 12   00:01:45  3526kN
3.g3-g4+
  +-  (46.46)   Depth: 12   00:01:45  3564kN
3.g3-g4+
  +-  (97.66)   Depth: 12   00:01:46  3599kN
3.g3-g4+
  +-  (200.06)   Depth: 12   00:01:46  3619kN
3.g3-g4+
  +-  (#14)   Depth: 12   00:01:57  4306kN
3.g3-g4+ Kh5xg4
  +-  (#14)   Depth: 12   00:01:57  4307kN
Parent - - By Vasik Rajlich (Silver) Date 2008-08-08 18:02
Once the scores go above +5, Rybka loses a lot of her power.

Vas
Parent - - By Apples (*) Date 2008-08-08 18:06
Will there be an update with a fix or does one have to patiently wait till the release of version 4?
Parent - - By Vasik Rajlich (Silver) Date 2008-08-08 19:14
If nothing major comes up, this one is a wait-for-Rybka-4 type issue.

Vas
Parent - By Vinvin (***) Date 2008-08-11 01:05
10 littles bugs, that dones't worth a little update ? :-)
We are close now ...
Parent - By Vempele (Silver) Date 2008-08-08 18:40
Like the power of reporting a full mate PV?
Parent - - By Dragon Mist (****) Date 2008-08-09 00:52
What?? Why? For the sake of sustainable development? To spare some power and reduce global warming? :-)
Parent - By Vasik Rajlich (Silver) Date 2008-08-09 16:03
It's complicated .. :)

Vas
Parent - By Monicker (**) Date 2008-08-11 01:51
Seems important to me if you are using Rybka as a tool for analysis.
Parent - - By diskamyl (**) Date 2008-08-08 22:34
I am currently using Rybka for a correspondence style game on one of the internet chess site, and being able to see more than 2 ply for multiple variations would be desirable.

I wonder what that site is?
Parent - - By Monicker (**) Date 2008-08-08 22:47
Check your PMs.

I didn't want to get accused of advertising for another site.  :)
Parent - By Arkansaw (***) Date 2008-08-08 23:00
Tell me about it too :)
Parent - By Leto (***) Date 2008-08-09 01:25
Try using Aquarium's IDeA mode, it is explained in this article: http://chessok.com/?p=518#more-518
Up Topic Rybka Support & Discussion / Rybka Discussion / Infinite Analysis / Lack of ply information

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill