Author: scott farrell
Date: 02:20:11 01/15/03
Go up one level in this thread
On January 15, 2003 at 00:08:12, Dann Corbit wrote:
wow, someone has done some reading.
Thanx for that analysis - it looked insightful and accurate, and like you , I am
no GM.
Yeh, my new code, assumes it is a draw when it is drawish, and tries to prove
otherwise. That is why I asked, as to whether someone can prove it is not a
draw, which the GM seem to consider, but ruffian certainly does not.
I might let it think to the .3 billion move mark, or even to the billion move
mark, and see where it gets, at 100Knps that will take some time though.
thanks again for your input
Scott
>On January 14, 2003 at 22:05:34, scott farrell wrote:
>
>>Do we know that if you play Rb4, that white has to take the sac?
>>
>>If white doesnt have to take the Rook sac, then maybe there is no solution to
>>wac.230.
>>
>>BTW: a few more minutes of thinking, and lots more fail lows, and the score has
>>dropped below 1.0. It likes Rb5 and Rh7 equally. I think both moves are equally
>>dumb.
>>
>>ply 10=2.53
>>ply 11=2.35
>>ply 12=0.88
>>ply 13 partial=0.64 (still thinking at this ply)
>>
>>The branching factor goes out the wazzooo as it starts detecting blockades, the
>>branching factor is around 11 now at ply 12, as a good move previously now shows
>>a draw, everything shows a draw, basically it is breaking down
>>null-move,alha-beta, everything, down to a straight brute force look at every
>>move.
>>
>>Thinking more on your ideas on what to do on detecting lots of blockades. I
>>might experiment with turing things off like : nullmove, pruning, reductions, or
>>say moving null move to R=1 (I use R=2 normally).
>
>It is not known if it is accurate. It is known that the key move is the only
>move with winning chances and at worst it draws.
>
>Here is the game it actually occurred in:
>
>[Event "?"]
>[Site "Bad Niendorf"]
>[Date "1927.??.??"]
>[Round "6"]
>[White "Kmoch, Hans"]
>[Black "Nimzowitsch, Aron I"]
>[Result "0-1"]
>[ECO "C15p"]
>
>1.e4 Nc6 2.Nc3 e6 3.d4 Bb4 4.Ne2 d5 5.e5 h5 6.Nf4 g6 7.Be3 Bxc3+ 8.bxc3 Na5 9.
>Bd3 Ne7 10.Nh3 c5 11.Bg5 c4 12.Be2 Nac6 13.Bf6 Rg8 14.O-O Qa5 15.Qd2 Nf5 16.
>Rfd1 Kd7 17.Ng5 Rf8 18.h3 Kc7 19.g4 hxg4 20.hxg4 Nfe7 21.Kg2 Ng8 22.Bg7 Re8 23.
>Rh1 Bd7 24.Rh3 Nd8 25.Rf3 Rc8 26.Rh1 Qxa2 27.Rh7 Kb8 28.Nxf7 Nxf7 29.Rxf7 Bc6
>30.Bf6 a5 31.Rh1 Qb2 32.Bg5 Rf8 33.Rfh7 Rc7 34.Rxc7 Kxc7 35.Qc1 Qxc3 36.Qa1
>Qxa1 37.Rxa1 Ra8 38.Bd2 b6 39.Kg3 Ne7 40.Bd1 Bd7 41.Bb4 Nc6 42.Bd6+ Kb7 43.c3
>b5 44.Rb1 b4 45.Ba4 b3 46.Bxc6+ Kxc6 47.g5 Ra7 48.Rb2 Rb7 49.Kf4 Bc8 50.Kg3
>{
>[D]2b5/1r6/2kBp1p1/p2pP1P1/2pP4/1pP3K1/1R3P2/8 b - -
>}
>.. Rb4
>51.cxb4 a4 52.b5+ Kxb5 53.Ba3 c3 54.Rb1 Kc4 55.f4 Kxd4 56.Kf2 Kc4 57.Ke1 d4 58.
>Ke2 Kd5 59.Kf3 Bb7 60.Re1 Kc4+ 61.Kf2 b2 62.f5 exf5 63.e6 Bc6 0-1
>
>Here is Ruffian's analysis:
>E:\programme\winboard\ruffian>ruffian -h 300M -T WAC230.EPD -t 999 -p 9
>Hash size 314572800
>Ruffian 1.0.1
>feature done=0
>Loading configuration: ruffian.cfg
>Warning: EGTB_HASH_SIZE changed, default=1048576, newvalue=8388608
>Logfile: Tue Jan 14 20:15:53 2003
>Total hash size 300M
>Initializing tablebases: /programme/winboard/nalimov
>Testing WAC230.EPD, 999 seconds, extra ply 9
>Testing: WAC.230 Solution: Rb4
> 2b5/1r6/2kBp1p1/p2pP1P1/2pP4/1pP3K1/1R3P2/8 b - -
> Depth Time Score Best line
> 12/18 1.03 1.98 ... a4 Ba3 Rh7 f3 Rh5 Kf4 Rh1 Bd6 Rc1 Bb4 Bb7 Ke3
> Re1+
> 12/18 1.26 x.xx ... Rh7
> 12--> 1.72 1.98 ... a4 Ba3 Rh7 f3 Rh5 Kf4 Rh1 Bd6 Rc1 Bb4 Bb7 Ke3
> Re1+
> 13/23 2.26 2.00 ... a4 Ba3 Rh7 Rb1 Rh8 f4 Ba6 Kg4 Kd7 Rb2 Bb5 Kg3
> Bc6
> 13--> 2.64 2.00 ... a4 Ba3 Rh7 Rb1 Rh8 f4 Ba6 Kg4 Kd7 Rb2 Bb5 Kg3
> Bc6
> 14/23 2.95 1.96 ... a4 Ba3 Rh7 Rb1 Rh8 f4 Ba6 Kg4 Rh5 Rb2 Rh7 Bd6
> Bb5 Ba3
> 14--> 3.54 1.96 ... a4 Ba3 Rh7 Rb1 Rh8 f4 Ba6 Kg4 Rh5 Rb2 Rh7 Bd6
> Bb5 Ba3
> 15/25 4.63 2.00 ... a4 Ba3 Rh7 Rb1 Rh8 f4 Ba6 Bb2 Bb5 Ba3 Rh7 Rb2
> Rh1 Kf2 Kd7 Ke3
> 15--> 4.96 2.00 ... a4 Ba3 Rh7 Rb1 Rh8 f4 Ba6 Bb2 Bb5 Ba3 Rh7 Rb2
> Rh1 Kf2 Kd7 Ke3
> 16/29 6.62 1.99 ... a4 Ba3 Rh7 Rb1 Rh8 f4 Ba6 Bb2 Rh5 Kg2 Rh7 Ra1
> Bb5 Kg3
> 16--> 9.19 1.99 ... a4 Ba3 Rh7 Rb1 Rh8 f4 Ba6 Bb2 Rh5 Kg2 Rh7 Ra1
> Bb5 Kg3
> 17/29 12.62 2.00 ... a4 Ba3 Rh7 Rb1 Rh8 f3 Ba6 Bb2 Rh5 f4 Kd7 Ba3
> Bb5 Rb2 Rh1 Kg4 Bc6
> 17/32 28.18 x.xx ... Rb4
> 17--> 31.40 2.00 ... a4 Ba3 Rh7 Rb1 Rh8 f3 Ba6 Bb2 Rh5 f4 Kd7 Ba3
> Bb5 Rb2 Rh1 Kg4 Bc6
> 18/33 36.09 2.01 ... a4 Ba3 Rh7 Rb1 Rh8 f3 Ba6 Bb2 Rh5 f4 Rh7 Ra1
> Bb5 Rb1 Kd7 Ra1 Rh8
> 18--> 38.24 2.01 ... a4 Ba3 Rh7 Rb1 Rh8 f3 Ba6 Bb2 Rh5 f4 Rh7 Ra1
> Bb5 Rb1 Kd7 Ra1 Rh8
> 19/34 55.87 2.02 ... a4 Ba3 Rh7 Rb1 Rh8 Bb2 Ba6 Kg4 Kd7 Ba3 Rh5 Bb2
> Bb5 Ba3 Rh2 Kg3 Rh8 Rb2 Bc6 f3 Bb5 Kf4 Rh1 Ke3
> 19--> 1:00 2.02 ... a4 Ba3 Rh7 Rb1 Rh8 Bb2 Ba6 Kg4 Kd7 Ba3 Rh5 Bb2
> Bb5 Ba3 Rh2 Kg3 Rh8 Rb2 Bc6 f3 Bb5 Kf4 Rh1 Ke3
> 20/39 1:22 2.02 ... a4 Ba3 Rh7 Rb1 Rh8 Bb2 Ba6 Kg4 Kd7 Ba3 Bb5 f3
> Rh2 Rb2 Rh1 Bb4 Rc1 Kf4 Bc6 Ke3 Rg1
> 20/40 2:43 x.xx ... Rb4
> 20/40 3:59 ++ ... Rb4
> 20/40 5:40 2.42 ... Rb4 cxb4 a4 b5+ Kxb5 Ba3 c3 Re2 Kc4 f3 Kxd4 Kf4
> Kd3 Rh2 Bd7 Bc1 Bb5 Ba3 c2 Rxc2 Kxc2 Ke3
> 20--> 5:40 2.42 ... Rb4 cxb4 a4 b5+ Kxb5 Ba3 c3 Re2 Kc4 f3 Kxd4 Kf4
> Kd3 Rh2 Bd7 Bc1 Bb5 Ba3 c2 Rxc2 Kxc2 Ke3
> 21/41 16:37 2.61 ... Rb4 cxb4 a4 b5+ Kxb5 Ba3 c3 Rb1 Kc4 f4 Kxd4 Kf3
> Kc4 Ke2 d4 Rb2 cxb2 Bxb2 Kd5 Kf3 Ba6 Kg3 d3
> 21... 16:37 2/16 a4
>
>Time: 16:39, cpu 46.46%, nodes 293396721, nps 293678
>Hash: 128M, used 99%, hits 45%
>EGTB: probs 0, hits 0
>
>Best move: Rb4 (2.61) Solved (1/1)
>
>Summery: 1/1
>
>Compare actual:
>cxb4 a4 b5+ Kxb5 Ba3 c3 Rb1 Kc4 f4 Kxd4 Kf2 Kc4 Ke1 d4 Ke2 ...
>To computed:
>cxb4 a4 b5+ Kxb5 Ba3 c3 Rb1 Kc4 f4 Kxd4 Kf3 Kc4 Ke2 d4 Rb2 ...
>
>And they agree pretty far along the trajectory.
>
>There may be an equally brilliant counter move. This is from Alex Szabo:
>
>"#230 [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 =] 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."
>
>The Re2! is not considered by the chess programs and was not played in the game.
> It may be an equally brilliant refutation to the brilliant move.
>
>Here is the continuation by Ruffian after the counter move 5. Re2! (which is
>clearly the best response):
>
>E:\PROGRA~3\winboard\ruffian>ruffian -h 300M -T foo.epd -t 999 -p 99
>Hash size 314572800
>Ruffian 1.0.1
>feature done=0
>Loading configuration: ruffian.cfg
>Warning: EGTB_HASH_SIZE changed, default=1048576, newvalue=8388608
>Logfile: Tue Jan 14 20:50:56 2003
>Total hash size 300M
>Initializing tablebases: /programme/winboard/nalimov
>Testing foo.epd, 999 seconds, extra ply 99
>Testing: test 1 Solution: Kc4
> 2b5/8/4p1p1/1k1pP1P1/p2P4/Bpp3K1/4RP2/8 b - -
> Depth Time Score Best line
> 10/21 1.20 2.12 ... Kc4 Re1 Kxd4 Rc1 Ba6 Kf4 Bd3 Bb4 c2 Ba3 Kc4
> 10--> 1.21 2.12 ... Kc4 Re1 Kxd4 Rc1 Ba6 Kf4 Bd3 Bb4 c2 Ba3 Kc4
> 11/23 3.29 2.23 ... Kc4 Re1 Kxd4 Rc1 Kd3 Rd1+ Ke2 Ra1 d4 Bc1 Bd7
> Bb2
> 11--> 3.30 2.23 ... Kc4 Re1 Kxd4 Rc1 Kd3 Rd1+ Ke2 Ra1 d4 Bc1 Bd7
> Bb2
> 12/25 8.73 2.25 ... Kc4 Kf4 Kxd4 Re1 Kc4 Rb1 d4 Rc1 Kd5 f3 b2 Bxb2
> cxb2 Rb1
> 12--> 8.75 2.25 ... Kc4 Kf4 Kxd4 Re1 Kc4 Rb1 d4 Rc1 Kd5 f3 b2 Bxb2
> cxb2 Rb1
> 13/27 21.76 2.56 ... Kc4 Kf4 Kxd4 Re1 Kc4 f3 d4 Rh1 Bd7 Rc1 d3 Bb2
> c2 Ke3
> 13--> 21.80 2.56 ... Kc4 Kf4 Kxd4 Re1 Kc4 f3 d4 Rh1 Bd7 Rc1 d3 Bb2
> c2 Ke3
> 14/30 32.82 2.48 ... Kc4 Kf4 Kxd4 Re1 Kc4 Rc1 d4 f3 Ba6 Kg4 d3 Bb2
> d2 Rxc3+ Kd5 Rc1 dxc1=Q Bxc1 Kxe5
> 14--> 32.87 2.48 ... Kc4 Kf4 Kxd4 Re1 Kc4 Rc1 d4 f3 Ba6 Kg4 d3 Bb2
> d2 Rxc3+ Kd5 Rc1 dxc1=Q Bxc1 Kxe5
> 15/35 1:27 2.70 ... Kc4 f3 Kxd4 f4 Ba6 Rh2 Ke4 Rh4 b2 f5+ Kxe5 Rb4
> Kxf5 Rxb2 cxb2 Bxb2 Kxg5 Ba3
> 15--> 1:27 2.70 ... Kc4 f3 Kxd4 f4 Ba6 Rh2 Ke4 Rh4 b2 f5+ Kxe5 Rb4
> Kxf5 Rxb2 cxb2 Bxb2 Kxg5 Ba3
> 16/35 3:16 2.58 ... Kc4 f4 Kxd4 f5 Ba6 Rh2 exf5 e6 Bb5 Rh8 b2 Rh4+
> Ke5 Rb4 Kxe6 Rxb2 cxb2 Bxb2 Bd3
> 16--> 3:16 2.58 ... Kc4 f4 Kxd4 f5 Ba6 Rh2 exf5 e6 Bb5 Rh8 b2 Rh4+
> Ke5 Rb4 Kxe6 Rxb2 cxb2 Bxb2 Bd3
> 17/37 4:37 2.35 ... Kc4 f4 Kxd4 f5 Ba6 Rh2 exf5 e6 Bb5 Rh8 b2 Rh4+
> Ke5 Rb4 Be8 Rxb2 cxb2 Bxb2+ Kxe6 Kf4 Bb5
> 17--> 4:52 2.35 ... Kc4 f4 Kxd4 f5 Ba6 Rh2 exf5 e6 Bb5 Rh8 b2 Rh4+
> Ke5 Rb4 Be8 Rxb2 cxb2 Bxb2+ Kxe6 Kf4 Bb5
> 18/38 6:54 2.27 ... Kc4 f4 Kxd4 f5 exf5 e6 Kd3 e7 Bd7 Rh2 Ke3 Rh8
> b2 Rb8 f4+ Kh2 f3 Bxb2 cxb2 Rxb2 Be8 Kg3
> 18--> 7:07 2.27 ... Kc4 f4 Kxd4 f5 exf5 e6 Kd3 e7 Bd7 Rh2 Ke3 Rh8
> b2 Rb8 f4+ Kh2 f3 Bxb2 cxb2 Rxb2 Be8 Kg3
> 19/42 10:50 2.43 ... Kc4 f4 Kxd4 f5 exf5 e6 Kd3 e7 Bd7 Kf3 Be8 Rh2
> Kd4 Rh4+ Ke5 Rb4 d4 Rb8 Bc6+ Ke2 Ke4 e8=B
> 19--> 11:07 2.43 ... Kc4 f4 Kxd4 f5 exf5 e6 Kd3 e7 Bd7 Kf3 Be8 Rh2
> Kd4 Rh4+ Ke5 Rb4 d4 Rb8 Bc6+ Ke2 Ke4 e8=B
> 20... 11:07 1/10 Kc4
>
>Time: 16:39, cpu 96.75%, nodes 613689959, nps 614298
>Hash: 128M, used 100%, hits 35%
>EGTB: probs 3, hits 3
>
>Best move: Kc4 (2.43) Solved (1/1)
>
>Summery: 1/1
>
>It looks like there is still doubt either way to me, but I'm no GM.
>
>In any case, all other moves don't have any chance to win.
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.