Not logged inRybka Chess Community Forum
Up Topic The Rybka Lounge / Test Positions / 10 hard positions
- - By dorszcz (*) Date 2018-11-19 20:09 Edited 2018-11-19 20:21
Can your engine find best moves?

2r2r1k/qb3p1p/p2p1Np1/np1Pp1P1/2P5/3RP2P/5R2/4Q1K1 w - - 0 30

8/4k2p/8/1P3p2/p2q4/p5QK/P5PP/8 w - - 0 48

r3r1k1/1b1n1pbp/pq1p2p1/1pp1P3/3P1B2/nP1B1NNP/P4PP1/R2QR1K1 w - - 0 21

1rbq1r1k/7p/2np2p1/p2N1p2/R1B1PP1b/1PP1N3/8/3Q1K1R w - - 0 23

rqn2rk1/1p2ppb1/p1b3pp/2R1P1N1/8/4BB1Q/3R1PPP/6K1 w - - 0 31

6K1/p1p4p/2p2R2/p4p1k/8/8/8/8 w - - 0 1

r4r2/3pk1p1/3p3p/7q/P2bP1n1/5N2/1P1N1PP1/R1BQ1RK1 b - - 0 17

8/8/8/5P2/P2bP2P/P2r1kP1/8/1R2K3 b - - 0 89

1k3q2/2r2n2/5NR1/3p1p2/p2PbP1p/P1n3bK/8/5Q2 w - - 0 1

6r1/8/p7/5p2/1k3K2/7P/P1P3P1/4N3 b - - 0 2
Parent - By user923005 (****) [us] Date 2018-11-19 22:15
Checkmates:
8/4k2p/8/1P3p2/p2q4/p5QK/P5PP/8 w - - acd 36; acs 60; bm Qh4+; ce 32608; pm Qh4+;
pv Qh4+ Qxh4+ Kxh4 Kd6 Kg5 Kc5 Kxf5 Kxb5 h4 Kc5 g4 Kd6 g5 Ke7 g6 h5 Kg5 Kf8 Kh6 Kg8 g7 Kf7 Kh7 Kf6 g8=Q Ke5 Qg3+ Kf5 Qxa3 Kg4 Qxa4+ Kf5;
8/8/8/5P2/P2bP2P/P2r1kP1/8/1R2K3 b - - acd 36; acs 25; bm Bc3+; ce 32602; pm Bc3+;
pv Bc3+ Kf1 Rd2 g4 Rh2 Kg1 Kg3 g5 Bd4+ Kf1 Rh1+ Ke2 Rxb1 Kd3 Rd1+ Kc4 Kxh4 g6 Be5 a5 Rd4+ Kb5 Rxe4 a6 Bd4 Kc6 Kg5 g7 Re8 Kd7 Ra8 Kc6 Rxa6+ Kb5 Ra8 Kc4 Bxg7 Kd5 Kxf5 Kc4 Rxa3 Kb4;
6K1/p1p4p/2p2R2/p4p1k/8/8/8/8 w - - acd 40; acs 14; bm Kf7; ce 32598; pm Kf7;
pv Kf7 Kg4 Ke6 h5 Ke5 c5 Rg6+ Kf3 Kxf5 h4 Rh6 Kg3 Ke4 h3 Ke3 Kg2 Rg6+ Kf1 Rf6+ Kg2 Ke2 Kg3 Rf3+ Kg2 Rf7 Kg3 Rg7+ Kf4 Kf2 h2 Rh7 Ke4 Rxh2 Kd5 Rh6 c6 Rh4 a6 Ra4 c4 Rxa5+ Kd4 Rxa6 Kd5 Ra5+ Kd4 Ke1 Kd3 Rc5 Ke3 Rxc6 Kd3;

Easy wins:
6r1/8/p7/5p2/1k3K2/7P/P1P3P1/4N3 b - - acd 36; acs 882; bm Rf8; ce 1864; pm Rf8;
pv Rf8 Ke3 Ka3 Kd2 Kxa2 c4 Kb3 Nf3 a5 Kc1 a4 c5 a3 Nd4+ Kc3 Ne2+ Kd3 Nf4+ Ke4 Ne2 Rg8 g4 fxg4 Kc2 a2 Ng3+ Kf3 Nf1 a1=Q Nd2+ Kf4 h4 Re8 h5 Qa4+ Kc3;
rqn2rk1/1p2ppb1/p1b3pp/2R1P1N1/8/4BB1Q/3R1PPP/6K1 w - - acd 40; acs 177; bm Nxf7; c3 "Nxf7"; ce 309; pm Nxf7;
pv Nxf7 Rxf7 Rd8+ Kh7 Bh5 Be4 Rcxc8 Qxc8 Rxc8 Rxc8 Qxc8 gxh5 f3 Bg6 f4 Be4 Qc4 Bg6 Qd5 a5 Qxb7 e6 Qa6 Bf5 Qxa5 Bf8 Kf2 Be4 g3 Kg7 Qa4 Bf5 Qe8 Be7 Kf3 h4 Qc6 hxg3;
r4r2/3pk1p1/3p3p/7q/P2bP1n1/5N2/1P1N1PP1/R1BQ1RK1 b - - acd 40; acs 146; bm Rf4; ce 221; pm Rf4;
pv Rf4 e5 Nxf2 exd6+ Kd8 Rxf2 Rh4 Nxh4 Qxd1+ Nf1 Bxf2+ Kxf2 Qd4+ Kg3 g5 Nf3 Qxd6+ Kh3 Rb8 N1h2 Kc8 g4 Rb4 a5 Kb8 Ra3 Qd1 Rc3 Kb7 Kg3 Qd6+ Kh3 Ka6 Ra3 Qe6 Bd2 Rxb2 Rd3 Rb3 Rxb3 Qxb3 Kg3 Qc4 Kf2 Qe4 Kg3 Qe2 Bc3 Qd3 Bg7 Kxa5 Kf2 Qc2+ Kg3 Qg6 Be5 d6;

Advantage for side to move:
1rbq1r1k/7p/2np2p1/p2N1p2/R1B1PP1b/1PP1N3/8/3Q1K1R w - - acd 40; acs 294; bm Ra2; ce 143; pm Ra2;
pv Ra2 fxe4 Rah2 g5 Qh5 Be6 Ke2 Rb7 Qh6 Bg8 Rg2 Bxd5 Nxd5 Ne7 Rxg5 Nxd5 Rxd5 Bf6 Rdh5 Qe7 Bd5 Qg7 Qxg7+ Rxg7 Bxe4 Re8 Kf3 Ree7 c4 Bc3 Rh6 Rgf7 R1h2 Bb4 f5 Rd7 R2h5 Bd2 Re6 Rfe7 Bd5 Rxe6 fxe6 Re7 Rh2 Bc3 Rg2 Bf6 Ke4 Re8 Bc6 Rf8 Ra2 Bd8 Rd2 Be7 Rd5 Kg7 Bd7 Kf6 Rxa5 Kg6 Ra6;
r3r1k1/1b1n1pbp/pq1p2p1/1pp1P3/3P1B2/nP1B1NNP/P4PP1/R2QR1K1 w - - acd 40; acs 1082; bm Nh5; ce 142; pm Nh5;
pv Nh5 c4 Nxg7 Kxg7 bxc4 Nxc4 Bxc4 bxc4 Rb1 Qc6 exd6 Rxe1+ Qxe1 Qe4 Qxe4 Bxe4 Re1 Bxf3 gxf3 Rd8 Re7 h6 Kf1 g5 Bh2 Kf8 f4 c3 fxg5 hxg5 Ke2 Re8 Re3 Rc8 Kd1 Rc4 Kc2 Rxd4 Kxc3 Rd1 Kc4 Rc1+ Kd5 Rc2 Re7 Nb6+ Kd4 Rxf2 Kc5 Nc8;

I only found a draw for these two, so I will look a little harder:
2r2r1k/qb3p1p/p2p1Np1/np1Pp1P1/2P5/3RP2P/5R2/4Q1K1 w - - acd 36; acs 174; bm Rf4; ce 0; pm Rf4;
pv Rf4 h5 Qh4 Rxc4 Rxc4 Nxc4 Nxh5 gxh5 Qxh5+ Kg8 g6 Kg7 Qh7+ Kf6 Qh4+ Kf5 Qg4+ Kf6;
1k3q2/2r2n2/5NR1/3p1p2/p2PbP1p/P1n3bK/8/5Q2 w - - acd 40; acs 42; bm Nd7+; ce 0; pm Nd7+;
pv Nd7+ Rxd7 Qb5+ Rb7 Qxb7+ Kxb7 Rb6+ Kc7 Rc6+ Kd8 Rd6+ Kc8 Rc6+;
Parent - - By user923005 (****) [us] Date 2018-11-19 23:06
I would be interested in any background you can give me on these interesting positions.
I had only analyzed one of them previously.
Parent - - By dorszcz (*) Date 2018-11-20 15:05 Edited 2018-11-20 16:50
All your solutions are correct, at least first moves. These 2 are draws.
It seems to be easy for you. What program did you use and what cpu do you have?

What do you mean by background? Source?
Which one did you analyze previously?
Parent - By user923005 (****) [us] Date 2018-11-20 19:18
This is the one that I analyzed previously:
rqn2rk1/1p2ppb1/p1b3pp/2R1P1N1/8/4BB1Q/3R1PPP/6K1 w - - acd 47; acs 3900; bm Nxf7; c3 "Nxf7"; ce 572; pm Nxf7;
pv Nxf7 Rxf7 Rd8+ Kh7 Bh5 gxh5 Qxh5 Nd6 Rxb8 Rf5 Qxf5+ Nxf5 Rxa8 Nxe3 fxe3 Kg6 e6 Be4 Rg8 Kh7 Rd8 Bf6 Kf2 Kg7 g3 Kg6 Rg8+ Kh7 Rf8 Kg6 h4 h5 Rd8 Bb2 Rg5+ Kf6 Rxh5 Kxe6 g4 Be5 Rh6+ Kf7 g5 Bg7 Rb6 Be5 Rc8 Bc6 h5 a5 Ra8 a4 Rxc6 bxc6 Rxa4 Kg7 Ra7 Bd6 Kf3 Kf7 Ke4;

By background, I meant things like:
"It is a study by Loek van Wely"
or
"This is from Capablanca verses Marshall 1918"
or any other descriptive information on the positions.

I used a modified version of Stockfish that takes mobility into account in pruning decisions.
I gave it 11 threads at 3.2 GHz and 8 GB ram for analysis along with 6 man Syzygy tablebase files.

I reanalyzed some of the positions last night (along with some others that I was interested).
Attachment: dix-ana.epd - Several engines take a look at some interesting positions (11k)
Parent - By user923005 (****) [us] Date 2018-11-20 19:22
They were hard positions (not quite as toxic as Walter's but good clean fun).

I suspected the draws were correct, since the evaluations were large negative values for some time, but sometimes engines will cling to a draw in a precarious position when a mate is just around the corner.

I have some high powered machines, but I don't use them for puzzles very often.  They mostly analyze frequently played positions.
Parent - By user923005 (****) [us] Date 2018-11-20 20:07
Here is a link to the current version of the program I use for most of my analysis:
https://drive.google.com/open?id=1Rmmi0Ni2-zsmVuNYyfU_-qtQNfOjllJa
It includes two 64-bit binaries and also the source code with changes I make.

You will probably want to turn off logging.  I have some new UCI parameters and I also change some defaults.
Parent - - By user923005 (****) [us] Date 2018-11-20 20:03
My current analysis records (somewhat improved by last night's analysis):
8/4k2p/8/1P3p2/p2q4/p5QK/P5PP/8 w - - acd 36; acs 60; bm Qh4+; ce 32608; pm Qh4+;
pv Qh4+ Qxh4+ Kxh4 Kd6 Kg5 Kc5 Kxf5 Kxb5 h4 Kc5 g4 Kd6 g5 Ke7 g6 h5 Kg5 Kf8 Kh6 Kg8 g7 Kf7 Kh7 Kf6 g8=Q Ke5 Qg3+ Kf5 Qxa3 Kg4 Qxa4+ Kf5;

8/8/8/5P2/P2bP2P/P2r1kP1/8/1R2K3 b - - acd 36; acs 25; bm Bc3+; ce 32602; pm Bc3+;
pv Bc3+ Kf1 Rd2 g4 Rh2 Kg1 Kg3 g5 Bd4+ Kf1 Rh1+ Ke2 Rxb1 Kd3 Rd1+ Kc4 Kxh4 g6 Be5 a5 Rd4+ Kb5 Rxe4 a6 Bd4 Kc6 Kg5 g7 Re8 Kd7 Ra8 Kc6 Rxa6+ Kb5 Ra8 Kc4 Bxg7 Kd5 Kxf5 Kc4 Rxa3 Kb4;

6K1/p1p4p/2p2R2/p4p1k/8/8/8/8 w - - acd 40; acs 14; bm Kf7; ce 32598; pm Kf7;
pv Kf7 Kg4 Ke6 h5 Ke5 c5 Rg6+ Kf3 Kxf5 h4 Rh6 Kg3 Ke4 h3 Ke3 Kg2 Rg6+ Kf1 Rf6+ Kg2 Ke2 Kg3 Rf3+ Kg2 Rf7 Kg3 Rg7+ Kf4 Kf2 h2 Rh7 Ke4 Rxh2 Kd5 Rh6 c6 Rh4 a6 Ra4 c4 Rxa5+ Kd4 Rxa6 Kd5 Ra5+ Kd4 Ke1 Kd3 Rc5 Ke3 Rxc6 Kd3;

6r1/8/p7/5p2/1k3K2/7P/P1P3P1/4N3 b - - acd 36; acs 882; bm Rf8; ce 1864; pm Rf8;
pv Rf8 Ke3 Ka3 Kd2 Kxa2 c4 Kb3 Nf3 a5 Kc1 a4 c5 a3 Nd4+ Kc3 Ne2+ Kd3 Nf4+ Ke4 Ne2 Rg8 g4 fxg4 Kc2 a2 Ng3+ Kf3 Nf1 a1=Q Nd2+ Kf4 h4 Re8 h5 Qa4+ Kc3;

rqn2rk1/1p2ppb1/p1b3pp/2R1P1N1/8/4BB1Q/3R1PPP/6K1 w - - acd 47; acs 3900; bm Nxf7; c3 "Nxf7"; ce 572; pm Nxf7;
pv Nxf7 Rxf7 Rd8+ Kh7 Bh5 gxh5 Qxh5 Nd6 Rxb8 Rf5 Qxf5+ Nxf5 Rxa8 Nxe3 fxe3 Kg6 e6 Be4 Rg8 Kh7 Rd8 Bf6 Kf2 Kg7 g3 Kg6 Rg8+ Kh7 Rf8 Kg6 h4 h5 Rd8 Bb2 Rg5+ Kf6 Rxh5 Kxe6 g4 Be5 Rh6+ Kf7 g5 Bg7 Rb6 Be5 Rc8 Bc6 h5 a5 Ra8 a4 Rxc6 bxc6 Rxa4 Kg7 Ra7 Bd6 Kf3 Kf7 Ke4;

r4r2/3pk1p1/3p3p/7q/P2bP1n1/5N2/1P1N1PP1/R1BQ1RK1 b - - acd 54; acs 3900; bm Rf4; ce 238; pm Rf4;
pv Rf4 e5 Nxf2 exd6+ Kd8 Rxf2 Rh4 Nxh4 Qxd1+ Nf1 Bxf2+ Kxf2 Qd4+ Kg3 g5 Nf3 Qxd6+ Kh3 Rc8 g4 Qd5 Kg3 Rc2 Bd2 Qd6+ Kh3 Rxb2 Re1 Qd3 Kg2 Ra2 a5 Ra4 N1h2 Kc8 Re3 Qd5 Re5 Qc4 Re8+ Kb7 Re7 Qc6 Re5 Ra2 Kh3 Ka6 Re3 Qd5 Re5 Qd6 Re2 Ra4 Re3 Qd5 Re5 Qd3 Re3 Qb5 Re7 Ra3 Kg2 Qc6 Kf2 Ra2 Nf1 Qd6;

1rbq1r1k/7p/2np2p1/p2N1p2/R1B1PP1b/1PP1N3/8/3Q1K1R w - - acd 50; acs 3900; bm Ra2; ce 150; pm Ra2;
pv Ra2 fxe4 Rah2 g5 Qh5 Be6 Ke2 Rb7 Qh6 Bg8 Rg2 Bxd5 Nxd5 Ne7 Rxg5 Nxd5 Rxd5 Bf6 Rdh5 Qe7 Bd5 Qg7 Qxg7+ Rxg7 Bxe4 Re8 Kf3 Ree7 c4 Bc3 Rh6 Rgf7 R1h2 Rd7 Ke3 Rg7 R2h3 Bb2 Kf3 Rgf7 R3h5 Bc1 Rh4 Bd2 Bc6 Rd8 Bd5 Re7 Be4 Rdd7 Rh2 Bb4 R2h5 Bd2 Rg5 Rf7 Bf5 Rd8 Rgh5 Rb8 Rxd6;

r3r1k1/1b1n1pbp/pq1p2p1/1pp1P3/3P1B2/nP1B1NNP/P4PP1/R2QR1K1 w - - acd 43; acs 3900; bm Nh5; ce 144; pm Nh5;
pv Nh5 c4;

2r2r1k/qb3p1p/p2p1Np1/np1Pp1P1/2P5/3RP2P/5R2/4Q1K1 w - - acd 56; acs 3900; bm Rf4; ce 0; pm Rf4;
pv Rf4 h5 Qh4 Rxc4 Rxc4 bxc4 Nxh5 gxh5 Qxh5+ Kg8 g6 Kg7 Qh7+ Kf6 Qh4+ Kf5 Qg4+ Kf6;

1k3q2/2r2n2/5NR1/3p1p2/p2PbP1p/P1n3bK/8/5Q2 w - - acd 72; acs 3900; bm Nd7+; ce 0; pm Nd7+;
pv Nd7+ Rxd7 Qb5+ Rb7 Qxb7+ Kxb7 Rb6+ Kc7 Rc6+ Kb8 Rb6+;
Parent - - By dorszcz (*) Date 2018-11-20 20:59
I'm glad you liked it.

Looking on my first post:
-positions number 3,4,5 are taken from ICCF. I didn't save players' names and tournaments, but in the future i will try to do it.
-position 10 is taken from old Arasan test suite (probably 10 or 11), but I played some moves. As I remember in original version some engines were finding best move by luck, but the key move is here if it plays Rf8 not Ka3 or other.
-positions 1,2,6,7,8,9 are mine.

In an hour I'll post you something more.
Parent - By user923005 (****) [us] Date 2018-11-20 21:02
Thanks.  I am a big fan of chess puzzles.
Parent - - By Vinvin (***) [be] Date 2018-11-23 00:36
Parent - By dorszcz (*) Date 2018-11-23 15:35
Thanks. I saw it, I visit talkchess quite regulary.
Up Topic The Rybka Lounge / Test Positions / 10 hard positions

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill