deep.09

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

Moderators: hgm, Rebel, chrisw

Dann Corbit
Posts: 12540
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

deep.09

Post by Dann Corbit »

Does anyone have a purely mechanical (computer generated) solution to this problem:
6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - -

Best move is listed as Qf5, so the key move involves coughing up your queen. Hence, null move will have to be turned way down if your engine allows it.

Code: Select all

1rb2rk1/2q1ppbp/pn1p1np1/2pP4/4PB2/2N1N3/PP1QBPPP/R3K2R b KQ - bm Na4; id deep.01;
1rbq1rk1/1p2ppbn/p2p2p1/P1pP4/4P3/2PB1N2/2P3PP/R1B1QRK1 w - - bm e5; id deep.02;
2b3r1/8/1kp1p1P1/2b1P2B/1pprPP2/7p/1P5Q/2KN4 b - - bm Ba6; id deep.03;
2r1r1k1/1p4bp/p3p1p1/qnnpP3/2p2PQP/2P1B1P1/1PB2N2/3R1RK1 w - - bm Bxg6; id deep.04;
2r2rk1/2pR1p1b/1pq1pQp1/p3P2p/P1PR3P/1P3P2/6PK/8 w - - bm g4; id deep.05;
3r1rk1/1p1bbppp/pq1ppn2/n7/2PNP3/PPN3P1/1B1R1PBP/3Q1R1K w - - bm g4; id deep.06;
4r1k1/pp3p1p/3p2pb/5q2/2Pp2n1/PP1P1NPb/2Q1PPBP/3RR1K1 b - - bm Nxf2; id deep.07;
4rrk1/p1q1bpp1/b1n1p2p/1pppP1N1/n4B1P/3P2P1/PPP1QPBN/R3R1K1 w - - bm Qh5; id deep.08;
6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - - bm Qf5; id deep.09;
6rk/2qr2p1/5pQn/p1pPpP2/PpP1B2P/1P6/K5R1/6R1 w - - bm Qxh6; id deep.10;
8/5k2/p3pp2/2P3p1/p2Pn3/3R4/K7/8 b - - bm Ke7; id deep.11;
8/8/4kpp1/3p1b2/p6P/2B5/6P1/6K1 b - - bm Bh3; id deep.12;
r1b2rk1/5ppp/pq1ppn2/1pb2P2/3NP2Q/2NBBP1P/PPP5/R6K w - - bm Rg1; id deep.13;
r1b2rk1/pp1pqp2/2n1p2p/6p1/2PNn3/R3P1B1/PP3PPP/3QKB1R w K - bm Bd3; id deep.14;
r1bq1rk1/pp2bppp/1n2p3/3pP3/8/2RBBN2/PP2QPPP/2R3K1 w - - bm Bxh7; id deep.15;
r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - bm Bh6; id deep.16;
r1bqk2r/ppp2pp1/1bnp1n1p/8/P3P2B/1NN5/1PP2PPP/R2QKB1R b KQkq - bm Nxe4; id deep.17;
r1r3k1/1b1nbp1p/p3p1p1/q2pP3/Np3PP1/1P1BB2R/P1PQ3P/2KR4 w - - bm Bb6; id deep.18;
r2qrb1k/1p1b2p1/p2ppn1p/8/3NP3/1BN5/PPP3QP/1K3RR1 w - - bm e5; id deep.19;
r2qrb1k/1p1b2p1/p3p2p/4p2n/3NN3/1B6/PPP3QP/1K3RR1 w - - bm Qg6; id deep.20;
r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w - - bm f5; id deep.21;
r3qb1k/1b4p1/p2pr2N/3n4/Pn2N3/2p3RP/1B3PP1/1B1QR1K1 w - - bm Nf5; id deep.22;
r4rk1/1bq1bppp/ppn1p3/2n1P3/P2N1BQ1/2NB4/1PP3PP/R4R1K w - - bm Nd5; id deep.23;
r4rk1/1p1n1pp1/2p1pnp1/7P/p2P2P1/qBP2P2/P1PBQ3/1K1R3R w - - bm Bc4; id deep.24;
rnbqr1k1/ppp2ppp/3b4/3p4/3P2n1/P1NBP2P/1P2NPP1/R1BQ1RK1 b - - bm Nh2; id deep.25;
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.
peter
Posts: 3186
Joined: Sat Feb 16, 2008 7:38 am
Full name: Peter Martan

Re: deep.09

Post by peter »

Hi Dann!
Dann Corbit wrote:Does anyone have a purely mechanical (computer generated) solution to this problem:
6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - -

Best move is listed as Qf5, so the key move involves coughing up your queen. Hence, null move will have to be turned way down if your engine allows it.
Here's H4 with option FiftyMovesDistance set to 5 instead of 50:

I find it remarkable, that Houdini obviously treats this boundary more cleverly than I thought, because move-counting has to be reset by 2.Nxf5, yet it works

[D]6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - - 0 1

Code: Select all

Analysis by Houdini 4 Pro x640:

1...gxh3 2.Nxh5 gxh5 3.Rxh3 Rg2 4.Qe1 R8g4 5.Qf1 Rg8 
  +-  (1.98)   Depth: 7/24   00:00:08  5kN
...
1...gxh3 2.Nxh5 
  +-  (2.67 --)   Depth: 23/81   00:00:35  527MN
1...Qf5 
  +-  (1.65 ++)   Depth: 23/81   00:00:38  594MN
1...Qf5 2.Nxf5 gxf5 3.Rg3 h5 4.Rhg2 Rb7 5.Qf2 Rbb8 6.Rg1 Rb7 7.Rc1 Rh8 8.c4 bxc3+ 9.Rxc3 dxc3+ 10.Kc1 Rhb8 11.Qxc5 Kf7 12.Kc2 Bxb3+ 13.Kxc3 Bd5 14.Qxa5 Rb3+ 15.Kd2 R8b6 16.Qc5 Rb1 17.Ke3 Re1+ 18.Kd4 Rb7 19.a5 Rd7 20.Qa3 Re2 21.a6 Ra2 22.Qc5 Rxa6 23.Ke3 Ra1 24.Qb4 Rh1 25.Qc3 Rxh4 
  +/-  (0.98)   Depth: 23/90   00:01:29  1850MN
...
1...Qf5 2.Nxf5 gxf5 3.Rg3 h5 4.Re2 Rh7 5.Re1 Rb8 6.Qf2 Rd8 7.Rge3 Rg7 8.Re4 Rc8 9.Rxd4 g3 10.Qg1 cxd4 11.Qxd4 c5 12.Qe3 g2 13.Qf2 Rc6 14.Rd1 Rg4 15.Rg1 Rc8 16.c4 bxc3+ 17.Kxc3 Rc6 18.Kd2 Rc7 19.Kc2 Rc6 20.Qd2 Kd7 21.Kb2 Rxh4 
  +/-  (0.79)   Depth: 32/90   00:05:47  8561MN
Peter.
Dann Corbit
Posts: 12540
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: deep.09

Post by Dann Corbit »

peter wrote:Hi Dann!
Dann Corbit wrote:Does anyone have a purely mechanical (computer generated) solution to this problem:
6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - -

Best move is listed as Qf5, so the key move involves coughing up your queen. Hence, null move will have to be turned way down if your engine allows it.
Here's H4 with option FiftyMovesDistance set to 5 instead of 50:

I find it remarkable, that Houdini obviously treats this boundary more cleverly than I thought, because move-counting has to be reset by 2.Nxf5, yet it works

[D]6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - - 0 1

Code: Select all

Analysis by Houdini 4 Pro x640:

1...gxh3 2.Nxh5 gxh5 3.Rxh3 Rg2 4.Qe1 R8g4 5.Qf1 Rg8 
  +-  (1.98)   Depth: 7/24   00:00:08  5kN
...
1...gxh3 2.Nxh5 
  +-  (2.67 --)   Depth: 23/81   00:00:35  527MN
1...Qf5 
  +-  (1.65 ++)   Depth: 23/81   00:00:38  594MN
1...Qf5 2.Nxf5 gxf5 3.Rg3 h5 4.Rhg2 Rb7 5.Qf2 Rbb8 6.Rg1 Rb7 7.Rc1 Rh8 8.c4 bxc3+ 9.Rxc3 dxc3+ 10.Kc1 Rhb8 11.Qxc5 Kf7 12.Kc2 Bxb3+ 13.Kxc3 Bd5 14.Qxa5 Rb3+ 15.Kd2 R8b6 16.Qc5 Rb1 17.Ke3 Re1+ 18.Kd4 Rb7 19.a5 Rd7 20.Qa3 Re2 21.a6 Ra2 22.Qc5 Rxa6 23.Ke3 Ra1 24.Qb4 Rh1 25.Qc3 Rxh4 
  +/-  (0.98)   Depth: 23/90   00:01:29  1850MN
...
1...Qf5 2.Nxf5 gxf5 3.Rg3 h5 4.Re2 Rh7 5.Re1 Rb8 6.Qf2 Rd8 7.Rge3 Rg7 8.Re4 Rc8 9.Rxd4 g3 10.Qg1 cxd4 11.Qxd4 c5 12.Qe3 g2 13.Qf2 Rc6 14.Rd1 Rg4 15.Rg1 Rc8 16.c4 bxc3+ 17.Kxc3 Rc6 18.Kd2 Rc7 19.Kc2 Rc6 20.Qd2 Kd7 21.Kb2 Rxh4 
  +/-  (0.79)   Depth: 32/90   00:05:47  8561MN
Thanks for your assistance with this. It is very helpful.
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.
jhellis3
Posts: 546
Joined: Sat Aug 17, 2013 12:36 am

Re: deep.09

Post by jhellis3 »

Unless I am missing something this a draw by 50 move rule.... so I mean think about what you are asking....

Looking some more, it appears to me that white wins after Qf5.... not sure why it would be considered a best move for black?
jhellis3
Posts: 546
Joined: Sat Aug 17, 2013 12:36 am

Re: deep.09

Post by jhellis3 »

Yeah, as far as I can see black is just busted here... there is no "best move".
Sven
Posts: 4052
Joined: Thu May 15, 2008 9:57 pm
Location: Berlin, Germany
Full name: Sven Schüle

Re: deep.09

Post by Sven »

jhellis3 wrote:Unless I am missing something this a draw by 50 move rule.... so I mean think about what you are asking....

Looking some more, it appears to me that white wins after Qf5.... not sure why it would be considered a best move for black?
I think White wins by moving a rook to c1 and then playing c2-c3 which opens the position for the queen. Therefore I agree that "best move Qf5" for black sounds a bit misleading since it seems to be based on the assumption of a forced draw.
jhellis3
Posts: 546
Joined: Sat Aug 17, 2013 12:36 am

Re: deep.09

Post by jhellis3 »

Yeah, you don't even need to move the rook to c1 though, just Qe1, c3, gg.

gxh3 is the best try, but still losing AFAICS.
peter
Posts: 3186
Joined: Sat Feb 16, 2008 7:38 am
Full name: Peter Martan

Re: deep.09

Post by peter »

Hi Joseph!
jhellis3 wrote:Yeah, you don't even need to move the rook to c1 though, just Qe1, c3, gg.
So you mean

6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - - 0 1

1... Qf5 2. Nxf5 gxf5 3. Rg3 h5 4. Qe1 Rg6 5. c3 I guess, how to continue after 5...bxc3
?
Peter.
peter
Posts: 3186
Joined: Sat Feb 16, 2008 7:38 am
Full name: Peter Martan

Re: deep.09

Post by peter »

Hi Sven!
Sven Schüle wrote: I think White wins by moving a rook to c1 and then playing c2-c3
You mean e.g. like this?

6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - - 0 1


1... Qf5 2. Nxf5 gxf5 3. Rg3 h5 4. Rg1 Rc7 5. Rc1 Rg6 6. c3 bxc3+

Don't see a sure White win.
Rather think the fortress is stable, do you see Black not losing soon after not giving the Queen- sac
:?:
Last edited by peter on Tue Aug 23, 2016 11:28 pm, edited 1 time in total.
Peter.
Dirt
Posts: 2851
Joined: Wed Mar 08, 2006 10:01 pm
Location: Irvine, CA, USA

Re: deep.09

Post by Dirt »

peter wrote:Rather think the fortress is stable, do you see Black not losing soon after not giving the Queen- sac
I sure feel stupid. How does black not give up the queen?
Deasil is the right way to go.