WAC.230 revisited yet again

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Dann Corbit
Posts: 9901
Joined: Wed Mar 08, 2006 7:57 pm
Location: Redmond, WA USA
Contact:

WAC.230 revisited yet again

Post by Dann Corbit » Sat Jul 11, 2009 4:27 am

I was wondering about WAC.230, now that we have superior analysis tools at our disposal. Alex Szabo showed a refutation some time ago, and I was trying to confirm his refutation from the root position of WAC.230. Crafty gets his line right up to the passed pawn creation move of 5. Re2! but the score is nowhere near drawlike. This crafty is the last version you can compile with DETECTDRAW (version 20.0).

Code: Select all

Analysis from C:\test\WAC230.EPD   
7/10/2009 8:06:22 PM Level: 3600 Seconds
Analyzing engine: CraftySMP

1) Rb4; id "WAC.230";   
    Searching move: Rb7-b4
    Best move (CraftySMP): Rb7-b4
    identical moves! Found in: 00:00
     12	00&#58;00	   1.309.562	9.354.014	 0.00	1. ... Rh7 2. Rb1 Kb5 3. Ba3 Rh5 4. f4 a4 5. Bb2 Bd7 6. Ba1 a3 7. Kg4 <HT>
     12	00&#58;00	   1.324.010	3.677.805	-1.70	1. ... Rb4 2. cxb4 a4 3. Rb1 c3 4. Rc1 b2 5. Rxc3+ Kd7 6. Rc7+ Kd8 7. Ra7 b1=Q 8. Rxa4
     12	00&#58;00	   3.899.320	10.831.444	-1.70	1. ... Rb4 2. cxb4 a4 3. Rb1 c3 4. Rc1 b2 5. Rxc3+ Kd7 6. Rc7+ Kd8 7. Ra7 b1=Q 8. Rxa4
     13	00&#58;00	   6.216.277	11.100.494	-2.00	1. ... Rb4 2. cxb4 a4 3. Rb1 Kb5 4. Bc5 c3 5. Rc1 b2 6. Rxc3 b1=Q 7. Ra3 Kc4 8. Rxa4
     13	00&#58;00	   7.079.876	11.062.306	-2.00	1. ... Rb4 2. cxb4 a4 3. Rb1 Kb5 4. Bc5 c3 5. Rc1 b2 6. Rxc3 b1=Q 7. Ra3 Kc4 8. Rxa4
     14	00&#58;00	   9.500.690	11.047.313	-1.76	1. ... Rb4 2. cxb4 a4 3. Rb1 Kb5 4. Ra1 c3 5. Rc1 c2 6. Kf4 a3 7. f3 a2 8. Rxc2 a1=Q 9. Rxc8 Qxd4+
     14	00&#58;01	  12.749.074	11.696.398	-1.76	1. ... Rb4 2. cxb4 a4 3. Rb1 Kb5 4. Ra1 c3 5. Rc1 c2 6. Kf4 a3 7. f3 a2 8. Rxc2 a1=Q 9. Rxc8 Qxd4+
     15	00&#58;01	  18.900.899	11.887.357	-2.16	1. ... Rb4!!
     15	00&#58;02	  25.623.315	12.811.657	-2.32	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Rc2 bxc2 5. Ba3 c3 6. Kf4 Kc4 7. Ke3 c1=Q+ 8. Bxc1 Kb3 9. Kf4 a3
     15	00&#58;02	  28.875.793	14.437.896	-2.32	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Rc2 bxc2 5. Ba3 c3 6. Kf4 Kc4 7. Ke3 c1=Q+ 8. Bxc1 Kb3 9. Kf4 a3
     16	00&#58;03	  37.873.722	12.624.574	-2.27	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Rc2 bxc2 5. Ba3 c3 6. Kf4 Kc4 7. Ke3 c1=Q+ 8. Bxc1 Kb3 9. Kf4 a3 10. Be3
     16	00&#58;04	  51.677.782	12.919.445	-2.27	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Rc2 bxc2 5. Ba3 c3 6. Kf4 Kc4 7. Ke3 c1=Q+ 8. Bxc1 Kb3 9. Kf4 a3 10. Be3
     17	00&#58;05	  75.484.013	12.580.668	-2.34	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Rc2 bxc2 5. Ba3 c3 6. Kf4 Kc4 7. Ke3 c1=Q+ 8. Bxc1 Kb3 9. Kf4 a3 10. Be3 Bd7
     17	00&#58;07	  93.196.456	13.313.779	-2.34	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Rc2 bxc2 5. Ba3 c3 6. Kf4 Kc4 7. Ke3 c1=Q+ 8. Bxc1 Kb3 9. Kf4 a3 10. Be3 Bd7
     18	00&#58;21	 263.576.580	12.551.265	-2.74	1. ... Rb4!!
     18	00&#58;34	 422.594.741	12.429.257	-2.71	1. ... Rb4 2. Rb1 Ra4 3. Rc1 Kb5 4. Kf4 Ra2 5. Ke3 a4 6. f4 Bd7 7. Rh1 Rc2 8. Bb4 b2 9. Rb1 Bc6 <HT>
     18	00&#58;38	 478.311.197	12.587.136	-2.71	1. ... Rb4 2. Rb1 Ra4 3. Rc1 Kb5 4. Kf4 Ra2 5. Ke3 a4 6. f4 Bd7 7. Rh1 Rc2 8. Bb4 b2 9. Rb1 Bc6 <HT>
     19	00&#58;51	 634.009.656	12.431.561	-3.11	1. ... Rb4!!
     19	01&#58;23	1.034.971.693	12.469.538	-3.27	1. ... Rb4 2. Rb1 Ra4 3. Rc1 Kb5 4. Kf3 Ra2 5. Ke3 a4 6. f3 a3 7. Bb4 Ka4 8. Rb1 Rc2 9. Ra1 a2 10. Kf4 Kb5
     19	01&#58;31	1.135.277.436	12.475.576	-3.27	1. ... Rb4 2. Rb1 Ra4 3. Rc1 Kb5 4. Kf3 Ra2 5. Ke3 a4 6. f3 a3 7. Bb4 Ka4 8. Rb1 Rc2 9. Ra1 a2 10. Kf4 Kb5
     20	02&#58;14	1.662.136.918	12.404.006	-3.13	1. ... Rb4 2. Rb1 Ra4 3. Rc1 Kb5 4. Kf3 Ra2 5. Ke3 a4 6. Bb4 a3 7. Rh1 Ka4 8. Rh8 Ba6 9. Ra8 Kb5 10. Rb8+ Kc6 11. Ra8 Kb6
     20	02&#58;50	2.116.991.815	12.452.893	-3.13	1. ... Rb4 2. Rb1 Ra4 3. Rc1 Kb5 4. Kf3 Ra2 5. Ke3 a4 6. Bb4 a3 7. Rh1 Ka4 8. Rh8 Ba6 9. Ra8 Kb5 10. Rb8+ Kc6 11. Ra8 Kb6
     21	04&#58;34	3.346.270,630	12.189	-3.53	1. ... Rb4!!
     21	09&#58;40	6.866.280,070	11.815	-4.26	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Ba3 c3 5. Rb1 Kc4 6. Kf4 Kxd4 7. Rh1 Kd3 8. Rh7 b2 9. Rh1 <HT>
     21	10&#58;53	7.803.996,586	11.933	-4.26	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Ba3 c3 5. Rb1 Kc4 6. Kf4 Kxd4 7. Rh1 Kd3 8. Rh7 b2 9. Rh1 <HT>
     22	17&#58;26	12.414.409,879	11.858	-4.50	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Ba3 c3 5. Re2 Kc4 6. f4 Kxd4 7. f5 exf5 8. e6 Kc4 9. Re5 b2 10. e7 Bd7 11. Re1 d4 12. Bxb2 cxb2
     22	21&#58;52	15.801.615,065	12.038	-4.50	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Ba3 c3 5. Re2 Kc4 6. f4 Kxd4 7. f5 exf5 8. e6 Kc4 9. Re5 b2 10. e7 Bd7 11. Re1 d4 12. Bxb2 cxb2
     23	36&#58;36	26.407.851,901	12.025	-4.30	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Ba3 c3 5. Re2 Kc4 6. f4 Kxd4 7. f5 exf5 8. e6 Kc4 9. Re5 b2 10. e7 Bd7 11. Re1 d4 12. Bxb2 cxb2 13. Kf4
     23	56&#58;28	41.449.515,645	12.234	-4.30	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Ba3 c3 5. Re2 Kc4 6. f4 Kxd4 7. f5 exf5 8. e6 Kc4 9. Re5 b2 10. e7 Bd7 11. Re1 d4 12. Bxb2 cxb2 13. Kf4
Which nearly matches Alex Szabo's line:

Code: Select all

                                              .. Rb4!? 2. cxb4 a4 3. b5+ Kxb5 4. Ba3 c3 5. Re2! Kc4 6. f4 Kxd4 7. f5 exf5 8. e6 Kd3 9. e7 Bd7 10. Kf3 d4 11. Rh2 Kc4 12. Rh8 b2 13. Rb8 d3 14. Bxb2 cxb2 15. Rxb2 a3 16. Rb7 Be8 17. Ra7 Kb3 18. Ke3 a2 19. Kxd3 Kb2 20. Rb7+ Kc1 21. Ra7
     

Question, can any engine other than Crafty 20.10 find this line and if so, can any engine also come up with a draw score due to white's passed pawn (I do see that Crafty's score has just stared to drop).

For reference, here is some previous WAC.230 analysis form a CCC of yesteryear:
================================================================================

Code: Select all

Author&#58; Alex Szabo

Date&#58; 20&#58;37&#58;36 07/08/02

Go up one level in this thread


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

On July 08, 2002 at 13&#58;58&#58;27, Miguel A. Ballicora wrote&#58;

>On July 07, 2002 at 15&#58;39&#58;04, Steve Maughan wrote&#58;
>
>>Mike,
>>
>>Is WAC 230 correct?  If I remember corretly there was a discussion here a few
>>months back that said that Rb4 was not such a forced win e.g.
>>
>>http&#58;//www.it.ro/ccc_search/ccc.php?art_id=163138
>>
>>Regards,
>>
>>Steve
>
>I still think that WAC230 is a win for black. I did not have time to do an
>analysis and _prove_ it but my intuition tells me so.
>I reproduce the post and answer later.
>
>Subject &#58; WAC 230
>
>----------------------------------------------------------------------------
>Posted by &#58; Bruce Moreland on April 13, 2001 at 04&#58;14&#58;13
>
>This is from a post by Alex Szabo.
>
>#230 &#91;Rb4 does not win. The main line is, 1... Rb4!? 2. cxb4 a4
>3. b5+ Kxb5 4. Ba3 c3 5. Re2! Kc4 6. f4 Kxd4 7. f5 exf5 8. e6 Kd3
>9. e7 Bd7 10. Kf3 d4 11. Rh2 Kc4 12. Rh8 b2 13. Rb8 d3
>14. Bxb2 cxb2 15. Rxb2 a3 16. Rb7 Be8 17. Ra7 Kb3 18. Ke3 a2
>19. Kxd3 Kb2 20. Rb7+ Kc1 21. Ra7 =&#93;
>
>Ra7, Rb6, Rb5, Rd7, Rf7, Rg7, Rh7, Bd7, Kd7, Kb6, Kb5, a4, and Rc7
>are just as good as the book solution Rb4 -- they all hold the game.
>
>&#91;D&#93;2b5/1r6/2kBp1p1/p2pP1P1/2pP4/1pP3K1/1R3P2/8 b - - 0 1
>
>I think this is worth thinking about. After 1. ... Rb4 2. cxb4 a5 3. b5+ Kxb5
>4. Ba3 c3 5. Re2 Kc4 6. f4, we get this&#58;
>
>&#91;D&#93;2b5/8/4p1p1/3pP1P1/p1kP1P2/Bpp3K1/4R3/8 b - - 0 1
>
>6. ... Bd7 is a possibility, but still seems like a draw.
>
>There is some fascinating stuff going on in the main line.
>
>WAC attributes this to Nimzovich, but doesn't give a game reference.
>
>I doubt that the position is solvable by a computer, and if anyone "finds" this,
>they've got some lucky eval terms.
>
>bruce
>-------------------------------------------------------------------------------
>
>The position is from the book "my system", so it is well known from there, not
>from WAC.
>
>IIRC when I looked to this analysis, 6... Bd7 wins. I am "remembering"
>blindfold, so beware. For instance, you can answer 7.f5 with Be8. There is no
>need to give counterchances to white on the king side. Why doing it? White is
>lost in the queen side. One thing is certain, Szabo's move 5. Re2! I is much
>stronger than the "classical" PV &#40;that came from the Nimzovich game&#41;.
>
>Regards,
>Miguel
>
>
>> ...

I don't see a win for Black after 6... Bd7.  The best line I can find is&#58; 6...
Bd7  7. f5 Be8!?  8. fxe6 Kxd4  9. e7 Kc4  10. Kf4 d4  11. Rh2 d3  12. Rh8 b2
13. Rxe8 b1=Q  14. Rf8 Rc8+ =.  Please let me know if Black's play can be
improved!

Alex.

User avatar
AdminX
Posts: 5115
Joined: Mon Mar 13, 2006 1:34 pm
Location: Acworth, GA
Contact:

Re: WAC.230 revisited yet again

Post by AdminX » Sat Jul 11, 2009 4:57 am

[d]2b5/1r6/2kBp1p1/p2pP1P1/2pP4/1pP3K1/1R3P2/8 b - - 0 1

Stockfish will follow Alex Szabo's line up to 7. f5, however gives 7. ... Kd3 as best after 6 minutes of thought.


Analysis by Stockfish 1.4 JA 64bit:

1...Rf7 2.f3 a4 3.Kf2 Kb5 4.Ke3 Rh7 5.Bc5 Bb7 6.f4 Rh3+ 7.Kd2
-+ (-2.94) Depth: 6 00:00:02 3kN
-+ (-2.74) Depth: 10 00:00:02 35kN

1...a4 2.Kf4 Rh7 3.f3 Rh5 4.Rg2 Kb5 5.Ke3 Bd7 6.Be7 Be8 7.Rb2 Bc6 8.Bf6
-+ (-2.84) Depth: 10 00:00:02 74kN
-+ (-3.01) Depth: 14 00:00:03 368kN

1...Rh7 2.Rb1 Kb5 3.Ba3 Rh5 4.f4 Bd7 5.Rg1 a4 6.Kg2 Rh4 7.Kg3 Rh8 8.Kg2 Ka5 9.Rh1 Rb8 10.Bb2 Rf8 11.Kf3 Bb5 12.Ba3 Rb8 13.Bb4+ Kb6 14.Ke3
-+ (-3.07) Depth: 14 00:00:03 690kN
-+ (-2.48) Depth: 25 00:01:19 360mN

1...Rb4 2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3 5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2 9.g7 b2 10.g8Q b1Q 11.Qxc8 Qd3+ 12.Kh4 Qe4+ 13.Kh5 Qh7+ 14.Kg4 d4 15.Qc4+ d3 16.Bc1 Qf5+ 17.Kh5 c2 18.Qb4 Kf3 19.Kh6 Qxe5 20.Qxa4 Qh2+ 21.Kg6 Qe5 22.Kh5 Qh8+ 23.Kg6 Qg8+ 24.Kh6 Qf8+ 25.Kh5 e5 26.Qg4+ Kf2
-+ (-3.03) Depth: 25 00:01:55 523mN
-+ (-3.13) Depth: 29 00:24:06 6298mN
"Good decisions come from experience, and experience comes from bad decisions."
__________________________________________________________________
Ted Summers

User avatar
Eelco de Groot
Posts: 4136
Joined: Sun Mar 12, 2006 1:40 am
Location: Groningen

Re: WAC.230 revisited yet again

Post by Eelco de Groot » Sat Jul 11, 2009 11:35 am

Dann Corbit wrote:

Code: Select all

     23	56&#58;28	41.449.515,645	12.234	-4.30	1. ... Rb4 2. cxb4 a4 3. b5+ Kxb5 4. Ba3 c3 5. Re2 Kc4 6. f4 Kxd4 7. f5 exf5 8. e6 Kc4 9. Re5 b2 10. e7 Bd7 11. Re1 d4 12. Bxb2 cxb2 13. Kf4
>
>Ra7, Rb6, Rb5, Rd7, Rf7, Rg7, Rh7, Bd7, Kd7, Kb6, Kb5, a4, and Rc7
>are just as good as the book solution Rb4 -- they all hold the game.
>
When endgame experts like Crafty still give such large evals I did not expect to see interesting lines from Ancalagon, did not even hope for that because I was worried that if you should see that there is no progress for Black, is there not a danger that your program picks one of those other 13 continuations? Much worse than not seeing a draw! But to my relief that has not happened.

Ancalagon is based more on Stockfish 1.3 so I think it should have diverged a little from Stockfish 1.4 in the evaluation function by now, but for the endgame this can be ignored I think. 99% Glaurung code there! So it should be not a big surprise that Ancalagon also picks Ted's 7... Kd3 after White's 7. f5

Most pleased I am that the eval did not go up, but that may be just a search fluctuation, or an artifact of the search combined with the fact that 5. Re2 was not found until the PVs with a lower eval.

This is still build 171, as all later versions are probably worse. I was experimenting if I could get nullmove killers, moves that did well refuting the null move but were not connected to the previous move from the same colour, and move those up a ply in the SearchStack. But this does not seem to work so well. It is a pity but that happens 8-)

I had made some new functions like this:

Code: Select all

 
 // update_nullmoveKillers&#40;) add a good "repair"-move that made the nullmove fail low,
  // but was not connected to the previous move, to the killer moves of that previous ply.

  void update_nullmoveKillers&#40;Move m, SearchStack& ss&#41; &#123;

    if &#40;m == ss.nullmoveKillers&#91;0&#93; /*|| m == ss.killers&#91;0&#93;*/)
        return;

    for &#40;int i = NULLMOVEKILLER_MAX - 1; i > 0; i--)
        ss.nullmoveKillers&#91;i&#93; = ss.nullmoveKillers&#91;i - 1&#93;;

    ss.nullmoveKillers&#91;0&#93; = m;
  &#125;
and

Code: Select all


  // move_is_nullmoveKiller&#40;) checks if the given move is among the
  // killer moves of that ply.

  bool move_is_nullmoveKiller&#40;Move m, const SearchStack& ss&#41; &#123;

      const Move* k = ss.nullmoveKillers;
      for &#40;int i = 0; i < NULLMOVEKILLER_MAX; i++, k++)
          if (*k == m&#41;
              return true;

      return false;
  &#125;
but unfortunately the idea does not seem to fly as yet :(

[D]2b5/1r6/2kBp1p1/p2pP1P1/2pP4/1pP3K1/1R3P2/8 b - -

Engine: Ancalagon 1.3 WS180 Build 171 (Athlon 2009 MHz, 256 MB)
by Romstad, Costalba, Kiiski, de Groot

2.00 0:00 +2.62 1...a4 2.Kf4 Kb5 (2.515) 8

2.00 0:00 +2.63 1...Rf7 (3.942) 12

3.00 0:00 +2.94 1...Rf7 2.Rb1 a4 (16.828) 51

3.00 0:00 +2.94 1...a4 (18.161) 55

4.00 0:00 +2.86 1...a4 2.Kf4 Ra7 3.Ba3 Rh7 (33.530) 97

4.00 0:00 +2.86 1...Rf7 (34.605) 100

5.00 0:00 +3.00 1...Rf7 2.f4 Rh7 3.Bc5 a4 (44.643) 124

6.00 0:00 +2.74 1...Rf7 2.f3 a4 3.Kf2 Rf5 4.Ke3 Rxg5 (116.879) 276

7.00 0:00 +2.74 1...Rf7 2.f3 a4 3.Kf2 Rf5 4.Ke3 Rxg5 (238.954) 436

8.00 0:00 +2.82 1...Rf7 2.f3 a4 3.Rb1 Kb5 4.Kf2 Rf5
5.Ke3 Rxg5 (701.358) 735

9.01 0:01 +2.76 1...Rf7 2.Rb1 a4 3.f4 Ra7 4.Ba3 Kb5
5.Kf3 Rh7 (1.311.418) 865

10.01 0:03 +2.58 1...Rf7 2.Rb1 a4 3.f3 Rf5 4.f4 Rf7
5.Kf3 Rh7 6.Bb4 (3.510.131) 1021

10.02 0:05 +2.62 1...Kb5 2.Rb1 Rh7 3.f3 a4 4.Bc5 Rh5
5.Be7 Rh8 6.Bc5 Bb7 7.Kf4 (6.503.012) 1172

10.04 0:06 +2.63 1...Rh7 (7.493.481) 1187

11.01 0:08 +2.66 1...Rh7 2.Rb1 a4 3.Ba3 Kb5 4.f3 Rh8
5.Kf4 Rh5 6.Bc5 Rh2 (10.103.026) 1208

12.01 0:19 +2.74 1...Rh7 2.Rb1 a4 3.Bc5 Rf7 4.Ba3 Kb5
5.Rh1 Bb7 6.f4 Ba8 7.Kf3 (25.198.760) 1263

13.01 0:36 +2.74 1...Rh7 2.Rb1 a4 3.Bc5 Rh5 4.Kg4 Rh2
5.Kg3 Rh8 6.Kg4 Kb5 7.Kf4 Rh4+ 8.Kg3 Re4 (47.918.654) 1305

14.01 0:48 +2.88 1...Rh7 2.Rb1 a4 3.Kg4 Kb5 4.Kg3 Bd7
5.f4 Be8 6.Bb4 Rh8 7.Kf3 Rh3+ 8.Ke2 Rd3
9.Rd1 (64.687.385) 1345

15.01 1:17 +2.88 1...Rh7 2.Rb1 a4 3.Kg4 Kb5 4.Ba3 Rh2
5.Kg3 Rh5 6.f4 Bd7 7.Bb4 Bc8 8.Kf2 Rh3 (105.346.947) 1358

15.09 8:28 +3.15 1...Rb4 (586.455.516) 1152

16.01 8:55 +3.15 1...Rb4 2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3
5.Rb1 Kc4 6.Kf3 Kxd4 7.Ke2 Ba6+
8.Kd1 c2+ 9.Kc1 cxb1Q+ 10.Kxb1 Kxe5
11.Kb2 d4 12.Bc5 Kd5 13.Ba3 e5 (621.363.531) 1160

17.01 10:30 +3.15 1...Rb4 2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3
5.Rb1 Kc4 6.Kf3 Kxd4 7.Ke2 Ba6+
8.Kd1 c2+ 9.Kc1 cxb1Q+ 10.Kxb1 Kxe5
11.Kb2 d4 12.Bc5 Kd5 13.Ba3 e5
14.Bb4 (738.043.479) 1170

18.01 17:33 +3.17 1...Rb4 2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3
5.Rb1 Kc4 6.Kf3 Kxd4 7.Ke2 Ba6+
8.Kd1 c2+ 9.Kc1 cxb1Q+ 10.Kxb1 Kxe5
11.Kb2 d4 12.Bc5 Kd5 13.Bb4 Kc4
14.Bd2 Bb7 (1.184.542.407) 1123

19.01 32:32 +2.98 1...Rb4 2.cxb4 (2.071.516.777) 1061

20.01 68:18 +2.84 1...Rb4 2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3
5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2
9.g7 b2 10.g8Q b1Q 11.Qxc8 Qd3+
12.Kg4 c2 13.Qf8 d4 14.Bc1 Kd1 (4.154.951.935) 1013


Eelco
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan

Vinvin
Posts: 4341
Joined: Thu Mar 09, 2006 8:40 am
Full name: Vincent Lejeune

Re: WAC.230 revisited yet again

Post by Vinvin » Sat Jul 11, 2009 12:16 pm

Do you mean 1..Rb4 is a forced draw ?

Dann Corbit wrote:I was wondering about WAC.230, now that we have superior analysis tools at our disposal. Alex Szabo showed a refutation some time ago, and I was trying to confirm his refutation from the root position of WAC.230.

User avatar
Eelco de Groot
Posts: 4136
Joined: Sun Mar 12, 2006 1:40 am
Location: Groningen

Re: WAC.230 revisited yet again

Post by Eelco de Groot » Sat Jul 11, 2009 2:01 pm

Ply 21 follow-up:
Eelco de Groot wrote:
When endgame experts like Crafty still give such large evals I did not expect to see interesting lines from Ancalagon, did not even hope for that because I was worried that if you should see that there is no progress for Black, is there not a danger that your program picks one of those other 13 continuations? Much worse than not seeing a draw! But to my relief that has not happened.
This is what I meant, it can happen when the score falls lower. Of course if it is really all drawn, for every one of these thirteen moves, and taking into account this is correspondence level play so your opponent may also find a draw with long analysis, it is slightly less bad but you would prefer the engine to stick to 1..Rb4 (everything else being equal and assuming all moves draw):


21.01 215:46 +2.33 1...Rb4 2.cxb4 (12.149.450.974) 938

21.02 240:55 +2.33 1...Rh7 (13.624.632.955) 942

Eelco
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan

bob
Posts: 20408
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: WAC.230 revisited yet again

Post by bob » Sat Jul 11, 2009 6:15 pm

Vinvin wrote:Do you mean 1..Rb4 is a forced draw ?

Dann Corbit wrote:I was wondering about WAC.230, now that we have superior analysis tools at our disposal. Alex Szabo showed a refutation some time ago, and I was trying to confirm his refutation from the root position of WAC.230.

Code: Select all

[/quote]

I believe that during all the analysis that went on, we came to the conclusion that Rb4 is the best move to play, as it is the only move that gives winning chances, but that it likely leads to a draw, the same as several other moves.  I do not believe we ever had program analysis from the root position to prove this however, and "stepping down thru the game" leaves too many holes in the analysis that a full-width search from the root does not have.

User avatar
beachknight
Posts: 3533
Joined: Tue Jan 09, 2007 7:33 pm
Location: Antalya, Turkey
Contact:

Re: WAC.230 revisited yet again

Post by beachknight » Mon Jul 13, 2009 4:30 pm

A different line:

1562: 2b5/1r6/2kBp1p1/p2pP1P1/2pP4/1
2b5/1r6/2kBp1p1/p2pP1P1/2pP4/1pP3K1/1R3P2/8 b - - 0 1

Analysis by Deep Fritz 11:

1...a4 2.Ba3 Rh7 3.Rb1 Bd7 4.Kg2 Rh4 5.f3 Rh5 6.f4 Rh4 7.Rf1 Bc8 8.Bc1 Rh7 9.Rh1 Ra7 10.Ba3 Rb7 11.Bb2 Rf7 12.Kf3 Bb7 13.Ba3 Kb5 14.Rh8 Rg7 15.Ke3 Rd7 16.Rh6
-+ (-2.00) Depth: 25/41 00:00:17 145mN
1...a4 2.Ba3 Rh7 3.Rb1 Bd7 4.Kg2 Rh4 5.f3 Rh5 6.f4 Rh4 7.Kg3 Rh7 8.Kg2 Bc8 9.Rh1 Rb7 10.Bb2 Ra7 11.Ba3 Rb7
-+ (-2.00) Depth: 26/38 00:00:23 200mN
1...a4 2.Ba3 Rh7 3.Rb1 Kb5 4.f4 Rh5 5.Kg2 Rh4 6.Kg3 Rh7 7.Kg2 Rg7 8.Kg3 Ra7 9.Rb2 Bd7 10.Kf3 Rb7 11.Ke3 Bc6 12.Rh2 Rf7 13.Rh1 Bb7 14.Rh6 Rg7 15.Rh1 Bc6 16.Kd2
-+ (-1.93) Depth: 27/43 00:00:35 293mN
1...a4 2.Ba3 Rh7 3.Rb1 Kb5 4.f4 Rh5 5.Kg2 Rh4 6.Kg3 Rh7 7.Kg2 Kc6 8.Rh1 Rb7 9.Bb2 Ra7 10.Ba3 Bd7 11.Kf3 Rb7 12.Rh2 Rb6 13.Ke3 Rb8 14.Kf3 Rb6
-+ (-1.93) Depth: 28/41 00:01:02 522mN
1...a4 2.Ba3 Rh7 3.Rb1 Rh5 4.f4 Rh7 5.Kg2 Rf7 6.Kg3 Rb7 7.Kg2 Bd7 8.Bb2 Ra7 9.Ba3 Be8 10.Rh1 Rb7 11.Bb2 Ra7 12.Ba3
-+ (-1.93) Depth: 29/47 00:01:24 705mN
1...a4 2.Ba3 Rh7 3.Rb1 Rh5 4.f4 Rh7 5.Kg2 Rf7 6.Kg3 Rb7 7.Kg2 Bd7 8.Bb2 Ra7 9.Ba3 Be8 10.Rh1 Rb7 11.Bb2 Ra7
-+ (-1.93) Depth: 30/44 00:01:51 926mN
1...a4 2.Ba3 Rh7 3.Rb1 Rh5 4.f4 Rh7 5.Kg2 Rf7 6.Kg3 Rb7 7.Kg2 Kc7 8.Kg3 Rb8 9.Bd6+ Kc6 10.Ba3 Ba6 11.Bb2 Rh8 12.Rg1 Bb7 13.Ba3 Kb6 14.Kg2 Rh4 15.Bc1 Kb5 16.Rh1
-+ (-1.93) Depth: 31/58 00:02:33 1269mN, tb=1
1...a4 2.Ba3 Rh7 3.Rb1 Rh5 4.f4 Ba6 5.Kg2 Rh4 6.Kg3 Rh7 7.Kg2 Rf7 8.Kg3 Rb7 9.Kf3 Rb8 10.Kg2 Bc8 11.Bb2 Kb5 12.Kf3 Ra8 13.Ba3 Ra6 14.Ke3 Ra7 15.Rh1 Rc7 16.Kd2
-+ (-1.93) Depth: 32/51 00:04:04 2033mN, tb=2
1...a4 2.Ba3 Rh7 3.Rb1 Rh5 4.f4 Ba6 5.Kg2 Rh4 6.Kg3 Rh7 7.Kg2 Rb7 8.Kf3 Rb8 9.Kg2 Kc7 10.Kf3 Kd7 11.Kg2 Kc6 12.Bb2 Rf8 13.Kg3 Rh8 14.Ba3 Bb5 15.Rg1 Rh7 16.Kg2
-+ (-1.93) Depth: 33/48 00:05:37 2806mN, tb=8
1...a4 2.Ba3 Rh7 3.Rb1 Rh5 4.f4 Ba6 5.Kg2 Rh4 6.Kg3 Rh8 7.Kg2 Rb8 8.Bb2 Rf8 9.Kg3 Rh8 10.Ba3 Bb5 11.Rg1 Rh7 12.Kg2 Rf7 13.Kf3 Kb6 14.Rh1 Ba6 15.Ke3 Bb7 16.Rh6
-+ (-1.93) Depth: 34/52 00:08:24 4209mN, tb=8
1...a4 2.Ba3 Rh7 3.Rb1 Rh5 4.f4 Bd7 5.Kg2 Rh4 6.Kg3 Rh8 7.Kg2 Rb8 8.Kf3 Kc7 9.Bd6+ Kb6 10.Ba3 Kc6 11.Kg2 Kc7 12.Bd6+ Kb6 13.Rh1 b2 14.f5 exf5
-+ (-1.93) Depth: 35/55 00:28:21 14408mN, tb=233
1...a4 2.Ba3 Rh7 3.Rb1 Kb5 4.Kg2 Rb7 5.Rh1 Kc6 6.Rb1 Bd7 7.Bb2 Be8 8.Rh1 Ra7 9.Ba3 Bd7 10.Kf3 Rb7 11.Bb2 Ra7 12.Ba3
-+ (-1.68) Depth: 36/54 00:56:05 28366mN, tb=319
1...Kb5 2.Rb1 Ka6 3.Rb2 a4 4.Ba3 Rh7 5.Rb1 Bb7 6.Bb4 Bc6 7.Ba3 Rh5 8.f4 Bb7 9.Rf1 Kb5 10.Bc1 Bc6 11.Kg2 Rh8 12.Ba3 Rb8 13.Rh1 Rg8 14.Kf3 Rg7 15.Ke3 Rb7 16.Rh6
-+ (-1.93) Depth: 36/55 01:01:07 30863mN, tb=389

(, chessengines20 13.07.2009)
hi, merhaba, hallo HT

PauloSoare
Posts: 1335
Joined: Thu Mar 09, 2006 4:30 am
Location: Cabo Frio, Brasil

Re: WAC.230 revisited yet again

Post by PauloSoare » Mon Jul 13, 2009 4:54 pm

I tried for several hours, improving the line of Alex Szabo, but not
see a win for blacks.
Maybe someone else can try.

User avatar
beachknight
Posts: 3533
Joined: Tue Jan 09, 2007 7:33 pm
Location: Antalya, Turkey
Contact:

Re: WAC.230 revisited yet again

Post by beachknight » Mon Jul 13, 2009 5:02 pm

Analysis after 1...Rb4:

1562: 2b5/1r6/2kBp1p1/p2pP1P1/2pP4/1
2b5/8/2kBp1p1/p2pP1P1/1rpP4/1pP3K1/1R3P2/8 w - - 0 1

Analysis by Deep Fritz 11:

2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3 5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2 9.g7 b2 10.g8Q b1Q
-+ (-1.90) Depth: 28/50 00:00:30 260mN
2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3 5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2 9.g7 b2 10.g8Q b1Q
-/+ (-1.21) Depth: 29/54 00:00:55 478mN
2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3 5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2 9.g7 b2 10.g8Q b1Q 11.Qxc8 Qd3+ 12.Kh4
-/+ (-1.01) Depth: 30/57 00:01:34 807mN, tb=12
2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3 5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2 9.g7 b2 10.g8Q b1Q 11.Qxc8 Qd3+ 12.Kh4 c2
-/+ (-0.99) Depth: 31/58 00:02:44 1420mN, tb=68
2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3 5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2 9.g7 b2 10.g8Q b1Q 11.Qxc8 Qd3+ 12.Kh4 c2
-/+ (-1.24) Depth: 32/53 00:05:30 2888mN, tb=149
2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3 5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2 9.g7 b2 10.g8Q b1Q 11.Qxc8 Qd3+ 12.Kh4 c2 13.Qc6 Qh7+ 14.Kg4 Qe4+
-/+ (-0.86) Depth: 33/62 00:09:01 4734mN, tb=674
2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3 5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2 9.g7 b2 10.g8Q b1Q 11.Qxc8 Qd3+ 12.Kh4 c2 13.Qc6 Qh7+ 14.Kg4 Qe4+
-/+ (-0.98) Depth: 34/65 00:16:26 8692mN, tb=1892

(, chessengines20 13.07.2009)
hi, merhaba, hallo HT

Dann Corbit
Posts: 9901
Joined: Wed Mar 08, 2006 7:57 pm
Location: Redmond, WA USA
Contact:

Re: WAC.230 revisited yet again

Post by Dann Corbit » Mon Jul 13, 2009 6:57 pm

{snip}
Eelco de Groot wrote: 20.01 68:18 +2.84 1...Rb4 2.cxb4 a4 3.b5+ Kxb5 4.Ba3 c3
5.Re2 Kc4 6.f4 Kxd4 7.f5 Kd3 8.fxg6 Kxe2
9.g7 b2 10.g8Q b1Q 11.Qxc8 Qd3+
12.Kg4 c2 13.Qf8 d4 14.Bc1 Kd1 (4.154.951.935) 1013 [/color]

Eelco
Actually, after only 68 minutes, the eval is falling quite nicely. This is a promising result.

Post Reply