Hard talkchess 2016 beta 1

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

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Paloma
Posts: 926
Joined: Thu Dec 25, 2008 8:07 pm

Re: Hard talkchess 2016 beta 1

Post by Paloma » Sat Apr 30, 2016 2:58 pm

> ...but the evaluation is worsening

of course because black loses.
at starting position find best white move. (eg 1.Nf5)

User avatar
Eelco de Groot
Posts: 4243
Joined: Sun Mar 12, 2006 1:40 am
Location: Groningen

Re: Hard talkchess 2016 beta 1

Post by Eelco de Groot » Sun May 01, 2016 12:59 am

Paloma wrote:> ...but the evaluation is worsening

of course because black loses.
at starting position find best white move. (eg 1.Nf5)
That would be interesting for sure, but Vincent already tried to find Nf5. Still with 20 threads Lazy SMP has a certain added luck factor. Maybe Louis would succeed here, but it looks like a hard problem for Stockfish? It would be interesting to know if the luck of Lazy SMP helps in a position with a lot of quiet moves at the end or that instead the old SMP with a more methodical search would be better. But that is just hard to find out without many tries, and you would have to go back to an old version of Stockfish to try, let's forget that. Also, trying the very latest Stockfish may make a difference just for the added Elo, but there are also two patches on the request list that might help if you want to try Multi PV. But those are only recently added to the Pull request list.
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

zullil
Posts: 6150
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: Hard talkchess 2016 beta 1

Post by zullil » Sun May 01, 2016 1:14 am

Eelco de Groot wrote:
Paloma wrote:> ...but the evaluation is worsening

of course because black loses.
at starting position find best white move. (eg 1.Nf5)
That would be interesting for sure, but Vincent already tried to find Nf5. Still with 20 threads Lazy SMP has a certain added luck factor. Maybe Louis would succeed here, but it looks like a hard problem for Stockfish? It would be interesting to know if the luck of Lazy SMP helps in a position with a lot of quiet moves at the end or that instead the old SMP with a more methodical search would be better. But that is just hard to find out without many tries, and you would have to go back to an old version of Stockfish to try, let's forget that. Also, trying the very latest Stockfish may make a difference just for the added Elo, but there are also two patches on the request list that might help if you want to try Multi PV. But those are only recently added to the Pull request list.
Not finding Nf5:

Code: Select all

info depth 55 seldepth 85 multipv 1 score cp 58 nodes 560905429666 nps 38410098 hashfull 999 tbhits 59167378 time 14603072 pv h2h4 g7g6 g2d2 c5d7 g1f1 d7e5 d2f4 e5d7 h4h5 e6e5 f4d2 e5d4 d1e1 d7e5 e1e5 e7e5 f1e1 e5e1 d2e1 e8d7 e1e4 g6h5 g4h5 d7c7 e4d4 c7b7 d4f6 h8d8 c2c3 d8d7 f6h6 f7f5 h6e6 d7h7 b1a2 c8c6 e6f5 c6h6 f5d5 b7a7 d5d4 a7b8 d4e5 b8b7 e5e4 b7b6 e4d4 b6c6 b2b4 h6h5 d4e4 c6b6 a3a4 b6a7 e4d4 a7a8 a2b3 h7h6 d4g7 a8b8 g7f8 b8b7 f8f3 b7a7 f3f2 a7b8 f2e3 b8c8 b3a3 c8d7 e3e4 d7c8 e4d4 c8b8 d4f4 b8c8 f4f7 c8b8

Vinvin
Posts: 4617
Joined: Thu Mar 09, 2006 8:40 am
Full name: Vincent Lejeune

Re: Hard talkchess 2016 beta 1

Post by Vinvin » Sun May 01, 2016 1:20 pm

Vinvin wrote:
Eelco de Groot wrote:...

But the analysis on Immortalchess got some more, I hope you can find it:

[Event "Анализ"]
[Site "?"]
[Date "2012.03.25"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[SetUp "1"]
[FEN "2r1k2r/4qpp1/p3p2p/1pn1P3/3N2P1/P7/1PP3QP/1K1R2R1 w k - 0 1"]
[PlyCount "35"]

1. Nf5 exf5 2. gxf5 f6 (2... Rg8 3. Rd5 $5 (3. e6 $6 fxe6 4. fxe6 g5 $5 (4...
g6 5. Rgf1 a5 6. Rde1 b4 7. Qg3 Nxe6 8. Rxe6 Qxe6 9. Re1 Kd7 10. Rxe6 Kxe6 11.
axb4 axb4 12. Qg4+ Kf7 13. Qxb4 $16) 5. Rge1 g4 $13) 3... g5 (3... Qc7 4. Re1
g6 5. Re3 Rd8 6. Rxd8+ Kxd8 7. Qa8+ $1 (7. Qd5+ Kc8 8. Rc3 Rd8 9. Qxc5 Qxc5 10.
Rxc5+ Kb7 $16) 7... Qc8 8. Qd5+ Ke8 9. f6 $18) 4. Qh3 Na4 5. Qxh6 Qc7 6. Rc1
Nb6 7. Rd4 Qc5 8. c3 Nc4 9. Qh7 Rf8 10. e6 Qe5 11. f6 Qxe6 12. Rcd1 Qxf6 13.
Re4+ Ne5 14. Qh3 $18) 3. exf6 Qxf6 4. Qd5 Kf8 5. Rg6 Qf7 6. Qd6+ Kg8 7. Rdg1
Rh7 8. f6 Ne6 9. Qxa6 Qd7 10. Qb6 Re8 11. h4 Kh8 12. Qf2 Rf8 (12... gxf6 13.
Qf5 Qd8 14. Rxf6 Ng7 15. Qf3 Qe7 16. Rf1 Rg8 17. Qc3 $18) 13. Qf5 Qf7 14. fxg7+
Nxg7 15. Qxb5 Nf5 16. a4 Rg7 17. Rxg7 Nxg7 18. Qc6 $16]
OK, thanks !
I'll let SF thinks for 20 hours after 1. Nf5 exf5 2. gxf5 with 4 best moves to have a final opinion.
After 14 hours (depth 49/100), SF thinks Rg8 is not completely losing (+0.55) but when I played Rg8 (with filled HT), SF sees immediately Rd5 with a >+1 score.
It's probably a new example of "good move pruned".

Code: Select all

FEN: 2r1k2r/4qpp1/p6p/1pn1PP2/8/P7/1PP3QP/1K1R2R1 b k - 0 2

Stockfish_16042402_x64_modern:
 49/100	9:09:02	401 569 132k	12 190k	+1,03	2. ... Qg5 3.Qf3 Qe7 4.Rd5 Nd7 5.Qg4 Rc5 6.Rxc5 Nxc5 7.Qxg7 Rf8 8.Qxh6 Qxe5 9.Qc6+ Nd7 10.f6 Kd8 11.Rd1 Qc7 12.Qd5 Re8 13.Qxf7 Qc6 14.h4 Re2 15.Qh7 Kc8 16.f7 Nf8 17.Qf5+ Re6 18.b3 Qc3 19.Qd5 Rc6 20.Qd8+ Kb7 21.Rd7+ Nxd7 22.Qxd7+ Rc7 23.Qxc7+ Kxc7 24.f8Q Qe1+ 25.Kb2 Qxh4 26.Qb4 Qg5 27.Qd4 Kb7 28.Qd7+ Kb6 29.Qd6+ Kb7 30.a4 bxa4 31.bxa4 a5 32.Kb3 Qf5 33.Qd3 Qe6+ 34.Qc4 Qe1 35.Qb5+ Ka7 36.Qc5+ Ka8 37.Qd5+ Ka7 38.Qd4+ Ka6 39.Qd6+ Ka7 40.Kb2 Qe8 41.Qc5+ Ka6 42.Ka3 Qb8 43.Qc4+ Ka7
 49/100	9:09:02	401 569 132k	12 190k	+1,02	2. ... f6 3.exf6 Qxf6 4.Qd5 Kf8 5.Rg6 Qf7 6.Qd6+ Kg8 7.Rdg1 Rh7 8.f6 Ne6 9.Qxa6 Qd7 10.Qb6 Kf7 11.fxg7 Rxg7 12.Rxg7+ Nxg7 13.Qxh6 Qd4 14.Qg6+ Kf8 15.c3 Qe5 16.Rg5 Rc6 17.Rxe5 Rxg6 18.Rxb5 Ra6 19.Rb8+ Ne8 20.Rb4 Rh6 21.h4 Ke7 22.a4 Ra6 23.Kc2 Nd6 24.Kb3 Nf5 25.Rf4 Ke6 26.Kb4 Ke5 27.Rg4 Rb6+ 28.Ka3 Rh6 29.a5 Nxh4 30.b4 Nf3 31.Ka4 Kd6 32.Rg7 Rh1 33.a6 Kc6 34.b5+ Kc5 35.Rc7+ Kd6 36.Rc6+ Kd5 37.c4+ Kd4 38.Rc7 Ra1+ 39.Kb4 Ne5 40.a7 Nd3+ 41.Kb3 Ne5
 49/100	9:09:02	401 569 132k	12 190k	+0,55	2. ... Rg8 3.e6 fxe6 4.fxe6 Nxe6 5.Qg6+ Qf7 6.Qe4 g5 7.Rde1 Rg6 8.Rgf1 Qg8 9.Qb7 Rd8 10.Rf5 g4 11.Qxa6 Ke7 12.Rxb5 Rf6 13.a4 Qg6 14.a5 g3 15.hxg3 Qxg3 16.Rb7+ Rd7 17.Rxd7+ Kxd7 18.Rd1+ Ke7 19.Qb5 Qc7 20.a6 Nc5 21.Re1+ Re6 22.Rg1 Nd7 23.Rg7+ Kd8 24.Qa4 Qa7 25.b3 Rxa6 26.Rxd7+ Qxd7 27.Qxa6 Qd1+ 28.Kb2 h5 29.Qf6+ Kc7 30.Qc3+ Kd7 31.Qg7+ Kd8 32.Qg5+ Kd7 33.Qf5+ Kd8 34.Qa5+ Kd7 35.Qa7+ Kd8 36.Qb8+ Kd7 37.Qb7+ Kd8 38.Qb6+ Kd7 39.b4 Qg4 40.Qb5+ Kc7 41.Qc5+ Kb7 42.Kb1 h4 43.Qd5+ Ka7 44.Qa5+ Kb7 45.Qb5+ Ka7 46.Qa4+ Kb8 47.Qe8+ Kc7

Vinvin
Posts: 4617
Joined: Thu Mar 09, 2006 8:40 am
Full name: Vincent Lejeune

Re: Hard talkchess 2016 beta 1

Post by Vinvin » Mon May 09, 2016 10:44 pm

Results for a subset of 27 positions from the Evgeniy's Set.
2 runs -> 2 columns
Conditions : Latest SF, 30 min by move, 8 GB hash, 6 pcs Syzygy
Notes :
1) position #18 still unsure, I don't see why 1.Qf4 is stronger than 1.N3d4.
2) position #22 : not sure 1.Rxc8 is stronger than 1.Qe2. It seems to be a transposition.
3) position #25 : probably a mistake, I see no point in 1...Bxe4. 1...Ne6 is always better.

Code: Select all

Analyzing engine: Stockfish_16050620_x64_modern
   Use seconds per move      : 1800 Seconds per move
   Analysis Lines            : Minimum search depth=12
   Activate abort analysis   : True. Minimal analysis depth 22. Abort, if the correct move is held 3 additional ply
 
 1) Nd4-f5             h2-h4                         h2-h4                    
 2) Kb7-c8             a5-a6                         a5-a6                    
 3) Be2-h5             Nh3xf2                        Nh3xf2                   
 4) Ne2-f4             Nb3-c5                        Nb3-c5                   
 5) Rf2-f6             Rf2-f6     * 260 Seconds      Rf2-f6     * 31 Seconds  
 6) Qc6-f3             Qc6-f3     * 144 Seconds      Qc6-f3     * 114 Seconds 
 7) Nf3-g5             Nf3-g5     * 61 Seconds       Nf3-g5     * 47 Seconds  
 8) c7-c8N             Kf3-g2                        Kf3-g2     
 9) Bd3xh7             Bd3xh7     * 436 Seconds      Bd3xh7     * 115 Seconds 
10) Nc3-d5             Qd2-e3                        Qd2-e3                   
11) Qh4-d4             Qh4xe7                        Qh4xe7                   
12) Nf3-h2             Be3-d2                        Be3-d2                   
13) .. Nc5-e4          d5-d4                         d5-d4                    
14) Ne5-g6             Ne5-d3                        Ne5-d3                   
15) .. Rh6xh2          a7-a6                         a7-a6                    
16) Nd4-c6             g7-g8Q                        g7-g8Q                   
17) Nf3-h4             Qd1xe1                        Qd1xe1                   
18) Qc1-f4             Nf3-d4                        Nf3-d4                   
19) Bb6-c7             Bb6xd8                        Bb6xd8                   
20) f4-f5              f4-f5      * 27 Seconds       f4-f5      * 61 Seconds  
21) .. Rb8xb4          Rb8xb4     * 74 Seconds       Rb8xb4     * 57 Seconds  
22) Rc1xc8             Qd1-e2                        Qd1-e2                   
23) e4-e5              e4-e5      * 21 Seconds       e4-e5      * 37 Seconds  
24) f5xe6              f5xe6      * 42 Seconds       f5xe6      * 1051 Seconds
25) .. Ba8xe4          Nc5-e6                        Nc5-e6                   
26) Rh1-h6             Bf7-e8                        Bf7-e8                   
27) e5-e6              e5-e6      * 228 Seconds      e5-e6      * 138 Seconds 
Set used :

Code: Select all

2r1k2r/4qpp1/p3p2p/1pn1P3/3N2P1/P7/1PP3QP/1K1R2R1 w k - id ?; bm Nf5;
4k1br/1K1p1n1r/2p2pN1/P2p1N2/2P3pP/5B2/P2P4/8 w - - id ?; bm Kc8;
8/2N4r/1p3pkp/8/5K1p/2P4N/P3Bn2/8 w - - id ?; bm Bh5;
1rb2r1k/3p1pp1/p3p2p/1p2P3/4q3/1N2B1R1/PPP1N1P1/2KR1B2 w - - id ?; bm Nf4;
2b2r2/2r2pbk/4n1pp/pp1pP3/2pP3q/P1P1B2P/1P1QNRPK/1B3R2 w - - id ?; bm Rf6;
2r1r1k1/2P2bp1/p1Q2pq1/2R3b1/1P6/4p1Bp/4B2P/3R3K w - - id ?; bm Qf3;
2r2r1k/1bqnbp1p/p3p1p1/npppP3/3P1PN1/1PPBBN2/P2Q2PP/R4RK1 w - - id ?; bm Ng5;
5nr1/2Pp2pk/3Pp1p1/4P1P1/6P1/5K2/8/7n w - - id ?; bm c8N;
r4rk1/4bppp/bq2p3/npnpP3/p2B1P2/P1PB1N2/1P2QNPP/R4RK1 w - - id ?; bm Bxh7+;
r5rk/1bq1bpp1/1p1ppn1p/pP5P/2P1P3/P1N2P2/1B1Q1P2/1K1R1BR1 w - - id ?; bm Nd5;
q1r5/k3ppp1/2R5/PP1KP2r/2p4Q/2p1p3/5pP1/3R4 w - - id ?; bm Qd4+;
1nrqr3/1b4bk/pp3pp1/3pPp1p/P1pP1P1P/2P1BN2/1PB2QP1/R4RK1 w - - id ?; bm Nh2 g4;
2q1r1k1/b1r2pp1/p4np1/2np4/1Q6/4PP1P/P1B1N1P1/1RBR1K2 b - - id ?; bm Nce4;
2q2r1k/p1p1b1pp/br1p1n2/2p1N1BP/Q1P1P3/8/PP1N1PP1/2KR3R w - - id ?; bm Ng6+;
r5k1/p1pb2bp/3p3r/P1pPp1p1/2B1Pq2/1R2QPp1/1P4PP/5RBK b - - id ?; bm Rxh2;
8/bp4Pk/4KB2/R6p/2pNp1P1/p7/3p1Pnq/2r5 w - - id ?; bm Nc6;
r2qrbk1/1b3p2/3p2pB/1p1P1N2/1np1P3/5NRP/1P3PP1/1B1Qn1K1 w - - id ?; bm N3h4;
1r2r1kb/2qn1p1p/3p2pB/pbpP1N2/1n2P3/R4N1P/1P3PP1/1BQ1R1K1 w - - id ?; bm Qf4;
1k1n4/1p1b1p2/1Bp1pPp1/1PP1P3/8/6P1/8/1K6 w - - id ?; bm Bc7+;
1r1b4/n2b3k/1p1Np1p1/p2pP1Pp/P2P1P2/1PBN4/6K1/2R5 w - - id ?; bm f5;
1r2kb1r/2q2ppp/p2pp3/6PB/1N2bP1P/2P1B2Q/PK6/3R3R b - - id ?; bm Rb4;
2r2rk1/p3qbb1/Pp2p3/1N1p3p/3PnNpP/3B2P1/1P3P2/2RQ1RK1 w - - id ?; bm Rxc8;
2rb3r/R5pp/p2pk3/2n2p2/2P1PP2/B7/P2N3P/1R4K1 w - - id ?; bm e5;
4kb1r/2q2ppb/p2pp2p/5P2/NrPB2P1/1P3Q2/1K1R3P/7R w k - id ?; bm fxe6;
brnq1rk1/5pp1/p2p4/Npn1p1RP/4P3/P4P2/1PPQB3/R2NK3 b - - id ?; bm Bxe4;
kr1b4/p1p2B2/2p5/2p1p3/KpP1P3/1P4N1/8/7R w - - id ?; bm Rh6;
rn2kbnr/pp2ppp1/2p5/3pP1Bp/3P3P/3Q4/PqP2PP1/RN2K1NR w KQkq - id ?; bm e6;

Vinvin
Posts: 4617
Joined: Thu Mar 09, 2006 8:40 am
Full name: Vincent Lejeune

Re: Hard talkchess 2016 beta 1

Post by Vinvin » Tue May 10, 2016 7:44 am

I forgot to mention that the CPU is an Intel 6*4GHz.
Vinvin wrote:Results for a subset of 27 positions from the Evgeniy's Set.
2 runs -> 2 columns
Conditions : Latest SF, 30 min by move, 8 GB hash, 6 pcs Syzygy

Dann Corbit
Posts: 11045
Joined: Wed Mar 08, 2006 7:57 pm
Location: Redmond, WA USA
Contact:

Re: Hard talkchess 2016 beta 1

Post by Dann Corbit » Tue May 10, 2016 6:15 pm

Here are some of the analysis records from that set for the more difficult problems, both with the position itself and also the position after the key move(s) have been made.
q1r5/k3ppp1/2R5/PP1KP2r/2pQ4/2p1p3/5pP1/3R4 b - - acd 51/11; acs 2126 ; acn 9057449124; ce 0 ; bm Kb8; pv Kb8 Qb6+ Qb7 Qxb7+ Kxb7 Rb6+ Ka7 Ra6+ Kb7 Rb6+;
q1r5/k3ppp1/2R5/PP1KP2r/2p4Q/2p1p3/5pP1/3R4 w - - acd 48/42; acs 3206 ; acn 28382353193; ce 0 ; bm Qxe7+; pv Qxe7+ Qb7 Qxb7+ Kxb7 Rb6+ Ka7 Ra6+ Kb7 Rb6+;
q1r5/k3ppp1/2R5/PP1KP2r/2p4Q/2p1p3/5pP1/3R4 w - - acd 46/13; acs 2034 ; acn 9095466001; ce 0 ; bm Qxe7+; pv Qxe7+ Qb7 Qxb7+ Kxb7 Rb6+ Ka7 Ra6+ Kb8 Rb6+ Ka8 Ra6+ Kb8;
brnq1rk1/5pp1/p2p4/Npn1p1RP/4P3/P4P2/1PPQB3/R2NK3 b - - acd 35/64; acs 3349 ; acn 21497582576; ce 217 ; bm Ne6; pv Ne6 Rg4 Ne7 h6 g6 Ne3 Kh7 Rd1 Qb6 Nb3 Rbc8 a4 Rcd8 axb5 axb5 Bd3 f5 exf5 Nxf5 Nxf5 Rxf5 Bxf5 gxf5 Rg7+ Nxg7 Qg5 Nh5 Qxh5 Qe3+ Kf1 Qxf3+ Qxf3 Bxf3 Rd2 Rc8 Kf2 Ba8 c3 Rc6 Nc1 Kxh6 Nd3 Rb6 Rd1 Kg5 Rg1+ Kh4 Nb4 Be4 Rg7;
brnq1rk1/5pp1/p2p4/Npn1p1RP/4P3/P4P2/1PPQB3/R2NK3 b - - acd 35/60; acs 2077 ; acn 6926817072; ce 212 ; bm Ne6; pv Ne6 Rg4 Ne7 h6 g6 Ne3 Kh7 Rd1 Qb6 Nb3 Rbd8 Bd3 f5 exf5 Nxf5 Nxf5 Rxf5 Bxf5 gxf5 Rg7+ Nxg7 Qg5 Nh5 Qxh5 Qe3+ Kf1 Qxf3+ Qxf3 Bxf3 Rd2 Rc8 Kf2 Bh5 Rxd6 Rxc2+ Kg3 f4+ Kh4 Bg6 Nd2 Kxh6 b3 a5 a4 b4 Kg4 Rc3 Nc4 Rg3+ Kh4 Rxb3 Nxe5;
2r1k2r/4qpp1/p3p2p/1pn1PN2/6P1/P7/1PP3QP/1K1R2R1 b k - acd 40/70; acs 1826 ; acn 6926641642; ce -30 ; bm exf5; pv exf5 gxf5 f6 e6 0-0 b4 Nxe6 fxe6 Rfd8 Qg4 Rxd1+ Rxd1 Rd8 Re1 Re8 Qh5 Rf8 Qd5 f5 Qc6 f4 Qxa6 f3 Rf1 f2 Qc6 Qf6 Qd6 Qf4 Qd5 Qf3 Qd3 Qxd3 cxd3 Rf6 Kc2 Kf8 Kd2 Ke7 Ke3 Rxe6+ Kxf2 Rf6+ Kg2 Rd6 Re1+ Kf6 Re3 Rd8;
2r1k2r/4qpp1/p3p2p/1pn1P3/3N2P1/P7/1PP3QP/1K1R2R1 w k - acd 38/64; acs 3527 ; acn 23555580981; ce 34 ; bm h4; pv h4 g6 Qd2 Nd7 Qe1 0-0 Qe3 Kg7 Rgf1 Nc5 Qf2 Kg8 Qf4 Kh7 Rd2 Qb7 Rdf2 Qd5 Qe3 Kg7 Re1 Qb7 Nf3 Qe7 h5 g5 Ref1 a5 b3 Nd7 Nd4 b4 Qe4 bxa3 Nc6 Rxc6 Qxc6 Nxe5 Qb6 Kh8 Rd1 a2+ Kxa2;
2r1k2r/4qpp1/p3p2p/1pn1P3/3N2P1/P7/1PP3QP/1K1R2R1 w k - acd 37/58; acs 2186 ; acn 7388131489; ce 48 ; bm h4; pv h4 g6 Rgf1 0-0 Qf2 Qb7 h5 b4 axb4 Qxb4 c3 Qb7 hxg6 fxg6 Qh2 Ne4 Kc2 Ng5 Rfe1 Kg7 Kc1 Rf7 Rh1 Rh8 Qe2 Re8 Rh2 Qe4 Qxe4 Nxe4 Kc2 Nf2 Ra1 Nxg4 Re2 a5 Rxa5 h5 Ra6 h4 Nxe6+ Kg8 Nd4 Nxe5 Rxg6+ Kh7;
2q1r1k1/b1r2pp1/p4np1/3p4/1Q2n3/4PP1P/P1B1N1P1/1RBR1K2 w - - acd 37/68; acs 2169 ; acn 7495362811; ce -112 ; bm Ba4; pv Ba4 Ree7 fxe4 Rxe4 Nd4 Bxd4 exd4 Qf5+ Kg1 Re2 Be3 Rxe3 Rf1 Qe6 Qb6 Qe7 Qb4 Rc4 Qxe7 Rxe7 Bb3 Rxd4 Rfd1 Rxd1+ Rxd1 Rd7 Kf2 Kh7 Ke3 Kh6 g3 Kg5 Rd3 Rd8 Kf3 Kf5 Re3 Ne4 Bc2 g5 g4+ Kf6 Bxe4 dxe4+ Kxe4 Re8+ Kf3 Rxe3+ Kxe3 Ke5 Kf3 Kd4 a3 f6 a4;
2q1r1k1/b1r2pp1/p4np1/2np4/1Q6/4PP1P/P1B1N1P1/1RBR1K2 b - - acd 41/69; acs 2399 ; acn 17090964647; ce 80 ; bm d4; pv d4 exd4 Rce7 Nc3 Ne6 Be3 Rc7 Ba4 Rxc3 Bxe8 Nxe8 Rbc1 Bxd4 Rxc3 Bxc3 Rc1 Bxb4 Rxc8 Kf8 Rb8 Bc5 Bxc5+ Nxc5 Ra8 Ke7 Ra7+ Ke6 h4 Nd6 Rc7 Kd5 Kf2 a5 Ra7 a4 g4 Kc4 Re7 Kb4 Rc7 Kb5 Ke3 Kb6 Re7 Ne6 Kf2 Kc5 Ra7 Kb5 Rd7 Kc6;
2q1r1k1/b1r2pp1/p4np1/2np4/1Q6/4PP1P/P1B1N1P1/1RBR1K2 b - - acd 40/68; acs 1898 ; acn 6618541727; ce 91 ; bm d4; pv d4 exd4 Rce7 Nc3 Ne6 Be3 Rc7 Ba4 Rxc3 Bxe8 Nxe8 Rbc1 Bxd4 Rxc3 Bxc3 Rc1 Bxb4 Rxc8 Kf8 Rb8 Bc5 Bxc5+ Nxc5 h4 Ke7 Kf2 Nc7 Ke3 N5e6 Ke4 Kd6 g3 Nd5 Ra8 Nc5+ Kd4 Nb6 Rd8+ Kc6 Rh8 Ne6+ Ke4 Nc4 Rc8+ Nc7 Rg8 Nd6+ Ke3 Ne6 Kf2 Nb5 g4 f6 Rc8+ Kd5 Rb8 Kc5 Ke3 Nd6 Kf2 Kc6 Ra8 Kb6;
1rnq1rk1/5pp1/p2p4/Npn1p1RP/4b3/P4P2/1PPQB3/R2NK3 w - - acd 39/65; acs 1551 ; acn 5358917610; ce -62 ; bm h6; pv h6 g6 fxe4 Nxe4 Nc6 Nxd2 Nxd8 Ne4 Nxf7 Kxf7 Rg2 Rh8 Bd3 Nf6 Bxg6+ Ke6 Rh2 Ne7 Bd3 e4 Bf1 Neg8 h7 Nxh7 Ne3 Ne7 a4 Ng5 Rxh8 Rxh8 axb5 axb5 Bxb5 Nf5 Nxf5 Rh1+ Bf1 Kxf5 Kd2 Kf4 Rd1 e3+ Kc1 Ne4 Bd3 Rh2 Bxe4 Kxe4 Rxd6 Kf5;
1rb2r1k/3p1pp1/p3p2p/1p2P3/4qN2/1N2B1R1/PPP3P1/2KR1B2 b - - acd 38/63; acs 2228 ; acn 7543453201; ce -106 ; bm Qxe5; pv Qxe5 Bd3 Bb7 Rh1 Qd6 Nh5 Rg8 Nc5 Rbc8 Bf4 Qxf4+ Nxf4 Rxc5 b4 Rc7 Kb2 d6 Nh5 Re7 Nf6 Rd8 c3 Rc7 Bc2 Re7 Nh5 Rg8 Rd1 d5 Nf4 Rc8 Nd3 Rec7 Nc5 a5 a3 axb4 axb4 Ra8 Nb3 Bc6 Rf1 Kg8 Nd4 Bd7 Bd3 Rca7 Nxb5 Ra2+ Kb1;
1rb2r1k/3p1pp1/p3p2p/1p2P3/4q3/1N2B1R1/PPP1N1P1/2KR1B2 w - - acd 38/65; acs 1812 ; acn 5839095030; ce 39 ; bm Bc5; pv Bc5 Rg8 Bd6 Rb6 Kb1 Qf5 Rf3 Qh5 Ng3 Rxd6 exd6 Qe5 Na5 f5 Ne2 g5 Rh3 Qf6 Rc3 f4 Nb3 g4 Rc7 f3 gxf3 Qxf3 a3 Rf8 Rc3 Qf2 Ng3 Bb7 Nc5 Rf3 Nge4 Bxe4 Nxe4 Qh4 Rxf3 gxf3 Bd3;
1rb2r1k/3p1pp1/p3p2p/1p2P3/4q3/1N2B1R1/PPP1N1P1/2KR1B2 w - - acd 38/57; acs 2308 ; acn 15011333554; ce 28 ; bm Ba7; pv Ba7 Rb7 Bc5 Rc7 Rd4 Qh7 Kb1 d6 exd6 Rxc5 Nxc5 Qf5 Rf3 Qxc5 d7 Bxd7 Rxd7 f5 Nf4 Re8 Nh5 Re7 Rd1 Qc6 Rfd3 Rc7 R1d2 Kh7 Ng3 Qb6 Ne2 e5 Rd5 Qe6 Rd6 Qc8 Nc3 Rc6 R2d5 Rxd6 Rxd6 Qb7 Ne2 e4;
1r2r1kb/2qn1p1p/3p2pB/pbpP1N2/1n2PQ2/R4N1P/1P3PP1/1B2R1K1 b - - acd 37/58; acs 2152 ; acn 6997355883; ce -106 ; bm Ne5; pv Ne5 Qh4 Bd7 Ne7+ Rxe7 Qxe7 Re8 Qg5 Nxd5 Nxe5 Rxe5 Qc1 Nb4 Bd2 Nc6 Bc3 Re8 Ba2 Bxc3 bxc3 a4 Qd2 Na7 Bd5 Nc8 c4 Nb6 Rb1 Bc6 Qb2 Rb8 Qc3 Re8 Ra2 Bxd5 exd5 Ra8 f4 Qd8 f5 Nd7 Rba1 Nb6 fxg6 hxg6 Rb1;
1r2r1kb/2qn1p1p/3p2pB/pbpP1N2/1n2P3/R4N1P/1P3PP1/1BQ1R1K1 w - - acd 39/71; acs 2995 ; acn 20072965130; ce 123 ; bm N3d4; pv N3d4 Bf6 Bg5 Bxg5 Qxg5 Qd8 Qxd8 Rexd8 Nxb5 gxf5 Nxd6 fxe4 Nxe4 Nxd5 Rd1 N7f6 Rxa5 c4 Bc2 Kg7 Nxf6 Nxf6 Rg5+ Kh8 Rxd8+ Rxd8 Rc5 Rb8 Rxc4 Rxb2 g3 h6 Kg2 Nd7 Rc3 Ne5 f4 Nd7 Kf3 Kg7 Bb3 h5 Ke3 Nf6 Rd3 Kg6 Kf3 Rb1 Bc4 Rb4 Ba2 Kg7;
1r2r1kb/2qn1p1p/3p2pB/pbpP1N2/1n2P3/R4N1P/1P3PP1/1BQ1R1K1 w - - acd 36/63; acs 2207 ; acn 7160539200; ce 114 ; bm N3d4; pv N3d4 Bf6 Bg5 Bxg5 Qxg5 Qd8 Qxd8 Rexd8 Nxb5 gxf5 Nxd6 fxe4 Nxe4 Nxd5 Rd1 N5f6 Rxa5 Kg7 Nd6 Rxb2 Ra7 Kg8 Nc4 Rbb8 Bf5 Nf8 Rd6 Ne8 Rc6 Ng7 Rxc5 Rd1+ Kh2 Rc1 Be4 Nge6 Rc6 Nd4 Rb6 Rxb6 Nxb6 Nfe6 Nd7 Kg7 Ne5 Rc7 Rxc7 Nxc7;
1nrqr3/1b4bk/pp3pp1/3pPp1p/P1pP1PPP/2P1BN2/1PB2Q2/R4RK1 b - - acd 36/58; acs 1640 ; acn 4904009432; ce -72 ; bm hxg4; pv hxg4 h5 gxf3 Qg3 Rh8 hxg6+ Kg8 Bxf5 Rc6 Rxf3 Nd7 Qg2 Nf8 Raf1 Rh5 Bg4 Rh4 Bh3 fxe5 fxe5 Qe8 Bg5 Rxh3 Rxh3 Rxg6 Rg3 Bc8 Bh6 Bxh6 Rxf8+ Qxf8 Rxg6+ Kh7 Rf6 Qg7 Kf1 Qxg2+ Kxg2 Kg7 Rc6 Bf5 Rxb6 Be4+ Kg3 Bc1 Kg4 Kf7 Rf6+ Ke7;
1nrqr3/1b4bk/pp3pp1/3pPp1p/P1pP1P1P/2P1BN2/1PB2QP1/R4RK1 w - - acd 38/47; acs 3485 ; acn 23139238967; ce 70 ; bm Bd2; pv Bd2 Rc7 Rae1 Bc8 b4 Nd7 Re3 Kg8 Ra1 Re6 a5 b5 Rae1 Re7 Bc1 Ra7 Kh1 Rb7 Kh2 Rc7 Qg3 Nf8 Kg1 Ra7 R1e2 Rac7 Qf2 fxe5 Nxe5 Qe8 Re1 Re6 Bd1 Bb7 Bf3 Rce7 g3 Nd7 Qe2 Nxe5;
1nrqr3/1b4bk/pp3pp1/3pPp1p/P1pP1P1P/2P1BN2/1PB2QP1/R4RK1 w - - acd 36/48; acs 2251 ; acn 7228070220; ce 70 ; bm Rae1; pv Rae1 Rc7 Bd2 Bc8 b4 Nd7 Qg3 Nf8 Bc1 Bd7 Re3 Bc6 Ba3 b5 a5 Kg8 Rfe1 Bb7 Bb2 Qc8 Bc1 Re6 Kh2 Ree7 Bd2 Kh7 Kg1 Kg8 R1e2 Re6 Bc1 Ree7 Qe1 Qd8 Bb2 fxe5 Nxe5;
1nrqr3/1b4bk/pp3pp1/3pPp1p/P1pP1P1P/2P1B3/1PB2QPN/R4RK1 b - - acd 34/44; acs 1945 ; acn 6174688021; ce -70 ; bm Rh8; pv Rh8 Nf3 Nc6 exf6 Bxf6 Ng5+ Kg8 Qg3 Qd7 Rfe1 Kg7 Nf3 Kh7 Bf2 Rhf8 Re2 b5 Rae1 Rce8 Ne5 Nxe5 fxe5 Bg7 Ra1 Ra8 Be3 Qe6 a5 Bc6 b4 cxb3 Bxb3 Rae8 Bf4 Kg8 Rae1;
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.

EvgeniyZh
Posts: 42
Joined: Fri Sep 19, 2014 2:54 pm
Location: Israel
Contact:

Re: Hard talkchess 2016 beta 1

Post by EvgeniyZh » Sat Jun 18, 2016 4:15 am

8/p2k1pp1/r3p1p1/8/1P6/P1R1PKPP/5P2/8 b - - bm Rd6

User avatar
Eelco de Groot
Posts: 4243
Joined: Sun Mar 12, 2006 1:40 am
Location: Groningen

Re: Hard talkchess 2016 beta 1

Post by Eelco de Groot » Sun Jun 19, 2016 7:06 am

EvgeniyZh wrote:8/p2k1pp1/r3p1p1/8/1P6/P1R1PKPP/5P2/8 b - - bm Rd6
Does White not still get a big advantage? I can't check with tablebases. This is what our tactical consultant, Tommy Cooper, says, without tablebases:

1...Rd6 2. Ke2 {best four Multi PV}

[D]8/p2k1pp1/3rp1p1/8/1P6/P1R1P1PP/4KP2/8 b - -

Engine: Tommy Cooper_006 MP (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

28 60:33 +2.16 2...Rb6 3.Kd3 g5 4.Kc4 Kc7 5.a4 a6
6.Kb3+ Kb7 7.Rc4 g6 8.Rg4 Rc6 9.Rxg5 Rc1
10.Rg4 Rb1+ 11.Kc4 Kc6 12.Rf4 f5
13.Rh4 Rc1+ 14.Kd4 Kd6 15.Rh8 Rb1 (19.311.793.873) 5314

28 60:33 +2.22 2...Ra6 3.h4 e5 4.Kd3 Rd6+ 5.Ke4 f6
6.a4 Rb6 7.Rc4 Rd6 8.b5 a6 9.Rb4 axb5
10.axb5 Rd2 11.b6 Kc8 12.Rc4+ Kb7
13.Rc7+ Kxb6 14.Rxg7 Rxf2 15.Rxg6 Kc6 (19.311.793.873) 5314

28 60:33 +2.31 2...Ke8 3.h4 Ra6 4.Kd3 Kd7 5.Kc2 Rb6
6.Kb3 a6 7.a4 e5 8.Rc5 Rf6 9.Rc2 Rb6
10.a5 Rd6 11.b5 axb5 12.Kb4 Rd1
13.Kxb5 Rb1+ 14.Ka6 Ra1 15.Rb2 Kc8 (19.311.793.873) 5314

28 60:33 +5.73 2...a6 3.Rd3 Kc6 4.Rxd6+ Kxd6 5.Kd3 g5
6.a4 Kc7 7.Kc4 Kd6 8.e4 f6 9.Kd4 Kc6
10.e5 f5 11.Kc4 Kb7 12.Kc5 Kc7 13.b5 axb5
14.axb5 Kb7 15.Kd6 Kb8 (19.311.793.873) 5314
_____________________________________________________

29 98:29 +2.11 2...Rb6 3.Kd3 e5 4.h4 a5 5.Kc4 Rc6+
6.Kb3 a4+ 7.Kc2 Rf6 8.f4 Re6 9.Kd2 Rf6
10.Rc5 Rd6+ 11.Ke2 e4 12.Ra5 Rc6
13.Rxa4 Rc2+ 14.Ke1 Rc3 15.Kd2 Rd3+ (31.348.055.141) 5304

29 98:29 +2.29 2...Ra6 3.h4 e5 4.Kd3 Rf6 5.Rc2 Ra6
6.Ra2 Rc6 7.b5 Rc1 8.a4 Rb1 9.Kc2 Rf1
10.Kc3 Kc7 11.Kb4 Kb6 12.a5+ Kb7
13.Kc5 a6 14.Rb2 axb5 15.Rxb5+ Ka6 (31.348.055.141) 5304

29 98:29 +2.37 2...Ke8 3.h4 Ra6 4.Kd3 Kd7 5.Kc2 Rb6
6.Kb3 a6 7.a4 e5 8.Rc5 Rf6 9.Rc2 Rb6
10.a5 Rb8 11.Rc5 Kd6 12.Kc4 f6 13.b5 axb5+
14.Rxb5 Rc8+ 15.Kb4 Rc1 (31.348.055.141) 5304

29 98:29 +6.08 2...Ke7 3.Rc7+ Ke8 4.Rxa7 e5 5.a4 Rf6
6.b5 Kd8 7.a5 Kc8 8.Kd3 Kb8 9.Re7 Rxf2
10.Kc4 Rc2+ 11.Kd5 f5 12.a6 Rc8
13.a7+ Ka8 14.b6 Rg8 15.Kxe5 Rf8 (31.348.055.141) 5304
_____________________________________________________

30 172:22 +2.32 2...Rb6 3.Kd3 g5 4.Kc4 e5 5.Rd3+ Kc8
6.Rd5 Re6 7.a4 Kc7 8.b5 Rf6 9.a5 Rxf2
10.Rxe5 Rc2+ 11.Kb3 Rg2 12.Re7+ Kb8
13.a6 Rf2 14.Kc4 g6 15.Rb7+ Ka8 (54.563.991.409) 5275

30 172:22 +2.38 2...Ra6 3.h4 e5 4.Kd3 Rf6 5.Rc2 Rd6+
6.Kc4 Kc6 7.a4 Kb6 8.a5+ Kb7 9.b5 Rd1
10.Ra2 Kc7 11.Rb2 Ra1 12.b6+ axb6
13.axb6+ Kb7 14.Kc5 Ra8 15.Rd2 Ra5+ (54.563.991.409) 5275

30 172:22 +2.52 2...Ke8 3.h4 Rb6 4.Kd3 Kd8 5.Kc4 a6
6.a4 e5 7.Rd3+ Kc7 8.Rd5 Re6 9.Rc5+ Kb7
10.Kd5 e4 11.Rc2 g5 12.hxg5 Rg6
13.Rc4 Rxg5+ 14.Kxe4 Rh5 15.Kf3 Rh8 (54.563.991.409) 5275

30 172:22 +6.53 2...Ke7 3.Rc7+ Kf6 4.Rxa7 Ke5 5.b5 f6
6.a4 Rb6 7.Rxg7 Rb8 8.Rxg6 Kf5 9.Rg4 Ke5
10.h4 Kd6 11.Rg6 Ra8 12.Rxf6 Rxa4
13.h5 Rb4 14.h6 Rxb5 15.g4 Rb8 (54.563.991.409) 5275


1...Rd6 2. Ke2 Rb6 3. Kd3


[D]8/p2k1pp1/1r2p1p1/8/1P6/P1RKP1PP/5P2/8 b - -

Engine: Tommy Cooper MP 006 (Latest Stockfish Master from 2016-06-18 with some added Retro extensions, i7 6700, Family 6, Model 94, Stepping 3, 3408MHz, No Overclock, 512 MB TT)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

17/40 0:01 +1.49 3...g5 4.Kc4 Kc7 5.Kb3+ Kb7 6.Rd3 Kc7
7.a4 a6 8.Rc3+ Kb7 9.Rc5 f6 10.b5 Rd6
11.a5 axb5 12.Rxb5+ Ka6 13.Kb4 (12.362.171) 7451

18/52 0:02 +1.56 3...g5 4.Kc4 Rc6+ 5.Kb3 Rd6 6.a4 Rd5
7.b5 Rd2 8.Rc6 Rxf2 9.Ra6 Kc7
10.Rxa7+ Kb6 11.Rd7 Re2 12.Kb4 Rb2+
13.Kc4 Ra2 14.Rd6+ Kc7 15.Ra6 Kb7
16.Kc5 Rc2+ (17.516.125) 7507
.
.
.
30/66 7:31 +2.06 3...Kd6 4.Kc4 Kc6 5.a4 a6 6.h4 Rb8
7.Kb3+ Kb7 8.Rc4 Ra8 9.Kc3 Rh8
10.Rf4 Rc8+ 11.Kd3 Rf8 12.Rd4 Kc7
13.Kc4 Rc8 14.Rf4 Kb6+ 15.Kd3 Rf8
16.Ke4 f6 (3.494.415.008) 7740

31/66 9:35 +2.11 3...Rd6+ 4.Kc2 Rb6 5.Kb3 a5 6.Rc5 axb4
7.axb4 Kd6 8.Kc4 Ra6 9.Ra5 Rc6+
10.Kd4 e5+ 11.Rxe5 Rc2 12.f4 Rd2+
13.Ke4 Rb2 14.Rd5+ Kc7 15.b5 Rb1
16.h4 Rg1 (4.451.644.996) 7732

32/82 17:22 +2.17 3...g5 4.Kc4 Rc6+ 5.Kb3 Rd6 6.a4 Rd2
7.Rc2 Rd3+ 8.Kc4 Ra3 9.a5 Kc7 10.b5 Kb8
11.Kb4 Ra1 12.Rc5 Rb1+ 13.Kc4 Rc1+
14.Kd4 Rd1+ 15.Ke4 f6 16.Rc6 Rd5 (8.073.795.804) 7742

33/86 25:20 +2.20 3...g5 4.Kc4 Kc6 5.a4 a6 6.Kb3+ Kb7
7.Rc4 Kb8 8.Re4 Kc8 9.Rd4 e5 10.Rg4 f6
11.h4 gxh4 12.gxh4 Rb7 13.Kc4 Rd7
14.Kc5 Kb7 15.Rg6 Kc7 16.h5 Kb7 (11.812.704.732) 7766

34/86 32:55 +2.21 3...Rd6+ 4.Kc2 Rb6 5.Kb3 a5 6.Rc5 axb4
7.axb4 Kd6 8.Kc4 Ra6 9.Ra5 Rc6+
10.Kd4 e5+ 11.Rxe5 Rc2 12.f4 Rd2+
13.Ke4 Rb2 14.Rd5+ Kc7 15.b5 Rb1
16.Rc5+ Kd6 (15.403.852.120) 7799

35/86 45:30 +2.28-- 3...Rd6+ 4.Kc2 (19.732.664.052) 7226

35/86 83:18 +2.35-- 3...Rd6+ 4.Kc2 (31.849.771.467) 6371

35/86 137:16 +2.39 3...a5 4.Kc4 axb4 5.axb4 e5 6.Ra3 e4
7.Ra7+ Ke8 8.h4 Rf6 9.Ra2 Ke7 10.Kd4 Rb6
11.Kc5 Rf6 12.Kd5 Rb6 13.Ra7+ Ke8
14.Ra4 Rf6 15.Ra2 Rb6 16.Kc4 Rf6 (49.089.540.332) 5959

36/86 168:06 +2.46-- 3...a5 4.Kc4 (58.615.809.265) 5811

Image
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

User avatar
Eelco de Groot
Posts: 4243
Joined: Sun Mar 12, 2006 1:40 am
Location: Groningen

Re: Hard talkchess 2016 beta 1

Post by Eelco de Groot » Mon Jun 20, 2016 2:51 am

Eelco de Groot wrote:
EvgeniyZh wrote:8/p2k1pp1/r3p1p1/8/1P6/P1R1PKPP/5P2/8 b - - bm Rd6
Does White not still get a big advantage? I can't check with tablebases. This is what our tactical consultant, Tommy Cooper, says, without tablebases:

1...Rd6 2. Ke2 {best four Multi PV}

[D]8/p2k1pp1/3rp1p1/8/1P6/P1R1P1PP/4KP2/8 b - -

Engine: Tommy Cooper_006 MP (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

28 60:33 +2.16 2...Rb6 3.Kd3 g5 4.Kc4 Kc7 5.a4 a6
6.Kb3+ Kb7 7.Rc4 g6 8.Rg4 Rc6 9.Rxg5 Rc1
10.Rg4 Rb1+ 11.Kc4 Kc6 12.Rf4 f5
13.Rh4 Rc1+ 14.Kd4 Kd6 15.Rh8 Rb1 (19.311.793.873) 5314

28 60:33 +2.22 2...Ra6 3.h4 e5 4.Kd3 Rd6+ 5.Ke4 f6
6.a4 Rb6 7.Rc4 Rd6 8.b5 a6 9.Rb4 axb5
10.axb5 Rd2 11.b6 Kc8 12.Rc4+ Kb7
13.Rc7+ Kxb6 14.Rxg7 Rxf2 15.Rxg6 Kc6 (19.311.793.873) 5314

28 60:33 +2.31 2...Ke8 3.h4 Ra6 4.Kd3 Kd7 5.Kc2 Rb6
6.Kb3 a6 7.a4 e5 8.Rc5 Rf6 9.Rc2 Rb6
10.a5 Rd6 11.b5 axb5 12.Kb4 Rd1
13.Kxb5 Rb1+ 14.Ka6 Ra1 15.Rb2 Kc8 (19.311.793.873) 5314

28 60:33 +5.73 2...a6 3.Rd3 Kc6 4.Rxd6+ Kxd6 5.Kd3 g5
6.a4 Kc7 7.Kc4 Kd6 8.e4 f6 9.Kd4 Kc6
10.e5 f5 11.Kc4 Kb7 12.Kc5 Kc7 13.b5 axb5
14.axb5 Kb7 15.Kd6 Kb8 (19.311.793.873) 5314
_____________________________________________________

29 98:29 +2.11 2...Rb6 3.Kd3 e5 4.h4 a5 5.Kc4 Rc6+
6.Kb3 a4+ 7.Kc2 Rf6 8.f4 Re6 9.Kd2 Rf6
10.Rc5 Rd6+ 11.Ke2 e4 12.Ra5 Rc6
13.Rxa4 Rc2+ 14.Ke1 Rc3 15.Kd2 Rd3+ (31.348.055.141) 5304

29 98:29 +2.29 2...Ra6 3.h4 e5 4.Kd3 Rf6 5.Rc2 Ra6
6.Ra2 Rc6 7.b5 Rc1 8.a4 Rb1 9.Kc2 Rf1
10.Kc3 Kc7 11.Kb4 Kb6 12.a5+ Kb7
13.Kc5 a6 14.Rb2 axb5 15.Rxb5+ Ka6 (31.348.055.141) 5304

29 98:29 +2.37 2...Ke8 3.h4 Ra6 4.Kd3 Kd7 5.Kc2 Rb6
6.Kb3 a6 7.a4 e5 8.Rc5 Rf6 9.Rc2 Rb6
10.a5 Rb8 11.Rc5 Kd6 12.Kc4 f6 13.b5 axb5+
14.Rxb5 Rc8+ 15.Kb4 Rc1 (31.348.055.141) 5304

29 98:29 +6.08 2...Ke7 3.Rc7+ Ke8 4.Rxa7 e5 5.a4 Rf6
6.b5 Kd8 7.a5 Kc8 8.Kd3 Kb8 9.Re7 Rxf2
10.Kc4 Rc2+ 11.Kd5 f5 12.a6 Rc8
13.a7+ Ka8 14.b6 Rg8 15.Kxe5 Rf8 (31.348.055.141) 5304
_____________________________________________________

30 172:22 +2.32 2...Rb6 3.Kd3 g5 4.Kc4 e5 5.Rd3+ Kc8
6.Rd5 Re6 7.a4 Kc7 8.b5 Rf6 9.a5 Rxf2
10.Rxe5 Rc2+ 11.Kb3 Rg2 12.Re7+ Kb8
13.a6 Rf2 14.Kc4 g6 15.Rb7+ Ka8 (54.563.991.409) 5275

30 172:22 +2.38 2...Ra6 3.h4 e5 4.Kd3 Rf6 5.Rc2 Rd6+
6.Kc4 Kc6 7.a4 Kb6 8.a5+ Kb7 9.b5 Rd1
10.Ra2 Kc7 11.Rb2 Ra1 12.b6+ axb6
13.axb6+ Kb7 14.Kc5 Ra8 15.Rd2 Ra5+ (54.563.991.409) 5275

30 172:22 +2.52 2...Ke8 3.h4 Rb6 4.Kd3 Kd8 5.Kc4 a6
6.a4 e5 7.Rd3+ Kc7 8.Rd5 Re6 9.Rc5+ Kb7
10.Kd5 e4 11.Rc2 g5 12.hxg5 Rg6
13.Rc4 Rxg5+ 14.Kxe4 Rh5 15.Kf3 Rh8 (54.563.991.409) 5275

30 172:22 +6.53 2...Ke7 3.Rc7+ Kf6 4.Rxa7 Ke5 5.b5 f6
6.a4 Rb6 7.Rxg7 Rb8 8.Rxg6 Kf5 9.Rg4 Ke5
10.h4 Kd6 11.Rg6 Ra8 12.Rxf6 Rxa4
13.h5 Rb4 14.h6 Rxb5 15.g4 Rb8 (54.563.991.409) 5275
Update:

8/p2k1pp1/3rp1p1/8/1P6/P1R1P1PP/4KP2/8 b - -

32 1344:17+2.49 2...Rb6 3.Kd3 e5 4.h4 a5 5.Kc4 axb4
6.axb4 Kc7 7.Kd5+ Kd7 8.Rb3 Kc7
9.Kxe5 f6+ 10.Kd4 Rd6+ 11.Kc3 Rc6+
12.Kd3 Rd6+ 13.Kc4 Rc6+ 14.Kd5 Rd6+
15.Ke4 Kb6 (431.253.309.723) 5346

32 1344:17+2.53 2...Ra6 3.h4 e5 4.Kd3 Rf6 5.Rc2 Ra6
6.Ra2 Kc7 7.Kc4 Rd6 8.Kc5 Rd7 9.a4 Rd6
10.a5 Rc6+ 11.Kd5 e4 12.Rb2 Rc1
13.Kxe4 a6 14.Kd5 Rd1+ 15.Ke5 Kd7 (431.253.309.723) 5346

32 1344:17+3.07 2...Ke8 3.h4 a6 4.Rc8+ Ke7 5.Rc7+ Kf8
6.Kf3 Rd3 7.Ra7 Rxa3 8.b5 Ke8 9.bxa6 Ra4
10.Ke2 Kf8 11.Kd2 Ra1 12.Kc3 Ra2
13.Kb4 Rb2+ 14.Ka3 Rb1 15.Ka4 Ra1+ (431.253.309.723) 5346

32 1344:17+6.70 2...f6 3.Rd3 Kc6 4.Rxd6+ Kxd6 5.Kd3 Kd5
6.a4 f5 7.h4 e5 8.Kc3 e4 9.b5 Kc5
10.Kb3 a6 11.bxa6 Kb6 12.Kc3 Ka7
13.Kd4 Kb6 14.Ke5 Ka7 15.Kd5 g5 (431.253.309.723) 5346
1...Rd6 2. Ke2 Rb6 3. Kd3


[D]8/p2k1pp1/1r2p1p1/8/1P6/P1RKP1PP/5P2/8 b - -

Engine: Tommy Cooper MP 006 (Latest Stockfish Master from 2016-06-18 with some added Retro extensions, i7 6700, Family 6, Model 94, Stepping 3, 3408MHz, No Overclock, 512 MB TT)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

17/40 0:01 +1.49 3...g5 4.Kc4 Kc7 5.Kb3+ Kb7 6.Rd3 Kc7
7.a4 a6 8.Rc3+ Kb7 9.Rc5 f6 10.b5 Rd6
11.a5 axb5 12.Rxb5+ Ka6 13.Kb4 (12.362.171) 7451

18/52 0:02 +1.56 3...g5 4.Kc4 Rc6+ 5.Kb3 Rd6 6.a4 Rd5
7.b5 Rd2 8.Rc6 Rxf2 9.Ra6 Kc7
10.Rxa7+ Kb6 11.Rd7 Re2 12.Kb4 Rb2+
13.Kc4 Ra2 14.Rd6+ Kc7 15.Ra6 Kb7
16.Kc5 Rc2+ (17.516.125) 7507
.
.
.
30/66 7:31 +2.06 3...Kd6 4.Kc4 Kc6 5.a4 a6 6.h4 Rb8
7.Kb3+ Kb7 8.Rc4 Ra8 9.Kc3 Rh8
10.Rf4 Rc8+ 11.Kd3 Rf8 12.Rd4 Kc7
13.Kc4 Rc8 14.Rf4 Kb6+ 15.Kd3 Rf8
16.Ke4 f6 (3.494.415.008) 7740

31/66 9:35 +2.11 3...Rd6+ 4.Kc2 Rb6 5.Kb3 a5 6.Rc5 axb4
7.axb4 Kd6 8.Kc4 Ra6 9.Ra5 Rc6+
10.Kd4 e5+ 11.Rxe5 Rc2 12.f4 Rd2+
13.Ke4 Rb2 14.Rd5+ Kc7 15.b5 Rb1
16.h4 Rg1 (4.451.644.996) 7732

32/82 17:22 +2.17 3...g5 4.Kc4 Rc6+ 5.Kb3 Rd6 6.a4 Rd2
7.Rc2 Rd3+ 8.Kc4 Ra3 9.a5 Kc7 10.b5 Kb8
11.Kb4 Ra1 12.Rc5 Rb1+ 13.Kc4 Rc1+
14.Kd4 Rd1+ 15.Ke4 f6 16.Rc6 Rd5 (8.073.795.804) 7742

33/86 25:20 +2.20 3...g5 4.Kc4 Kc6 5.a4 a6 6.Kb3+ Kb7
7.Rc4 Kb8 8.Re4 Kc8 9.Rd4 e5 10.Rg4 f6
11.h4 gxh4 12.gxh4 Rb7 13.Kc4 Rd7
14.Kc5 Kb7 15.Rg6 Kc7 16.h5 Kb7 (11.812.704.732) 7766

34/86 32:55 +2.21 3...Rd6+ 4.Kc2 Rb6 5.Kb3 a5 6.Rc5 axb4
7.axb4 Kd6 8.Kc4 Ra6 9.Ra5 Rc6+
10.Kd4 e5+ 11.Rxe5 Rc2 12.f4 Rd2+
13.Ke4 Rb2 14.Rd5+ Kc7 15.b5 Rb1
16.Rc5+ Kd6 (15.403.852.120) 7799

35/86 45:30 +2.28-- 3...Rd6+ 4.Kc2 (19.732.664.052) 7226

35/86 83:18 +2.35-- 3...Rd6+ 4.Kc2 (31.849.771.467) 6371

35/86 137:16 +2.39 3...a5 4.Kc4 axb4 5.axb4 e5 6.Ra3 e4
7.Ra7+ Ke8 8.h4 Rf6 9.Ra2 Ke7 10.Kd4 Rb6
11.Kc5 Rf6 12.Kd5 Rb6 13.Ra7+ Ke8
14.Ra4 Rf6 15.Ra2 Rb6 16.Kc4 Rf6 (49.089.540.332) 5959

36/86 168:06 +2.46-- 3...a5 4.Kc4 (58.615.809.265) 5811
This took a bit long but that is because of the extensions, (only change in the code) they scale with depth and distance to beta in the opposite direction of nullmove reductions, and so can become larger than ONE_PLY. This is a bit tricky but because concentrated near the root, and divided by three, I have not seen them explode yet. Bench is large though:

Engine: Tommy Cooper (Stockfish clone) 20160618_006 MP
===========================
Total time (ms) : 20444
Nodes searched : 36632857
Nodes/second : 1791863

Depths below therefore are larger than corresponding depths of Stock. And take quite a bit of time. That why I called them Retro extensions :)

36/101 230:12 +2.39++ 3...e5 (78.409.201.437) 5676

36/101 230:34 +2.28++ 3...e5 (78.527.246.738) 5676

36/101 232:16 +2.37-- 3...e5 4.Kc4 (79.060.285.394) 5673

36/101 364:13 +2.56 3...g5 4.Kc4 Kc7 5.a4 e5 6.Kd5+ Kb7
7.Rc4 Rh6 8.Kxe5 Ra6 9.b5 Rf6 10.f4 Re6+
11.Kd4 Rh6 12.Kc5 gxf4 13.Rxf4 Rh5+
14.Kb4 f6 15.h4 Re5 16.Rd4 Kc8 (122.134.972.566) 5588

37/101 609:08 +2.63-- 3...g5 4.Kc4 (202.818.734.433) 5549

37/101 815:18 +2.56++ 3...g5 (270.746.616.841) 5534

37/101 840:17 +2.63-- 3...g5 4.Kc4 (278.984.348.156) 5533

37/101 1159:38+2.67 3...g5 4.Kc4 Kc7 5.a4 e5 6.Kd5+ Kb7
7.Rc4 Ra6 8.Rg4 Rxa4 9.Rxg5 g6
10.Kxe5 Ra6 11.Rg4 Kc7 12.Rf4 Re6+
13.Kd5 Rd6+ 14.Kc5 Rc6+ 15.Kd4 f6
16.h4 Rb6 (383.208.726.545) 5507
Image
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

Post Reply