Quiet - Hyatt the few loud positions

[Dann Corbit]

Given this file from Dr. Hyatt's site:
ftp://ftp.cis.uab.edu/pub/hyatt/tests/openings.epd

Most of the positions are fairly neutral. Those that are polar (centipawn evaluation of 3/4 of a pawn or more) are collected and shown below:

Code: Select all

2kr1b1r/pp3ppp/n2Pqn2/1Np5/3Pp1b1/2N1Q3/PPP2PPP/R1B1KB1R w KQ - acd 18; acn 64984000; ce 306; pv dxc5 Nb4 Nd4 Qe5 a3 Nbd5 f4 Qh5 Nxd5 Qxd5 b4 Bxd6 cxd6 Qxd6 h3 Bh5 g4 Qxd4 Qxd4 Rxd4 gxh5 Nxh5 Be3 Rd7 Rg1 Kb8 Be2 g6 Bxh5 gxh5 Rd1; bm dxc5; id "Hyatt-Quiet.0089"; acs 416;
r2qkbnr/2pb2pp/p2p4/1p2p3/4P3/1PN2N2/1PP2PPP/R1BQ1RK1 w kq - acd 19; acn 131295936; ce 195; pv Nxe5 dxe5 Qh5+ Ke7 Nd5+ Kd6 Rd1 Kc6 Qxe5 Bd6 Qd4 Kb7 e5 Be7 Nxe7 Qxe7 Qxd7 Qxe5 g3 Nf6 Qxg7 Rhg8 Qf7 Rae8 b4 Rgf8 Qa2 Qe6 Qxe6 Rxe6 Be3; bm Nxe5; id "Hyatt-Quiet.2563"; acs 444;
r2qk2r/pp1n1pp1/2pb1nb1/3pp2p/4P1PN/3P3P/PPPN1PB1/R1BQ1RK1 w kq - acd 19; acn 41560056; ce 159; pv g5 Nh7 Nxg6 fxg6 exd5 cxd5 h4 Nb6 a4 a5 c4 O-O cxd5 Rf4 Nf3 Rg4 Re1 Qc7 Re4 Rd8 Qb3 Nd7 Rxg4 hxg4 Nd2; bm g5; id "Hyatt-Quiet.2361"; acs 301;
r1b1kb1r/pp1n1ppp/5q2/2p5/Q1B5/2p2N2/PP3PPP/R1B1K2R w KQkq - acd 18; acn 77816925; ce 158; pv Bg5 Qc6 O-O-O Be7 Qxc6 cxb2+; bm Bg5; id "Hyatt-Quiet.0453"; acs 319;
r1b1kb1r/pp3ppp/2qppn2/2P5/2n2B2/1NN3P1/PP2PP1P/R2QKBR1 w Qkq - acd 19; acn 71510079; ce 158; pv e4 b5 cxd6 Bd7 e5 Ne4 Bg2 Nxb2 Qf3 Qxc3+ Qxc3 Nxc3 Bxa8 Nca4 Bf3 g5 Bxg5 Nd3+ Ke2 Nxe5 Rac1 Rg8 Bf4 Nxf3 Kxf3 f6 Rc7 e5 Be3 Bxd6 Rxa7; bm e4; id "Hyatt-Quiet.0467"; acs 418;
r1bk1b1r/ppppq1pp/8/1B3p2/3Bn3/6P1/PPP2P1P/RN1QR1K1 w - - acd 18; acn 68655894; ce 158; pv Nc3 Qg5 Nxe4 fxe4 Bf1 Be7 Rxe4 d6 Qf3 Bd7 Rf4 Bc6 Qb3 Rf8 Rxf8+ Bxf8 Qg8 Qf5 Bxg7; bm Nc3; id "Hyatt-Quiet.0760"; acs 319;
rn2kbnr/pp3p1p/3p4/4p1p1/4P3/5Q2/PPP1NP1q/RNB1K1R1 w Qkq - acd 19; acn 69059197; ce 157; pv Nbc3 Nd7 Nd5 Rc8 Bxg5 h6 Be3 Qh4 Nec3 Qd8 O-O-O a6 Kb1 b5 a3 Rc6 Rd3 Rc8 Rg2 Rc4 Rg3; bm Nbc3; id "Hyatt-Quiet.3368"; acs 325;
rn1q1rk1/ppp3bp/3p2p1/3Pp2n/2P3b1/2N1BP2/PP1Q3P/R3KBNR w KQ - acd 19; acn 59733712; ce 148; pv fxg4 Qh4+ Bf2 Qxg4 Nh3 Qf5 O-O-O Nd7 Rg1 e4; bm fxg4; id "Hyatt-Quiet.3139"; acs 431;
rnb2rk1/p3nppp/1p2p3/2qp4/5P2/2PB1NP1/PP1N1P1P/R2QK2R w KQ - acd 18; acn 98522878; ce 135; pv Bxh7+ Kxh7 Ng5+ Kg6 Qg4 f5 Qh4 Qd6 g4 Qxf4 Qh5+ Kf6 Nh7+ Ke5 Nxf8 Kd6 gxf5 Qe5+ Kd1 Nbc6 Re1 Qxf5 Qxf5 Nxf5 Ng6 Bd7 f4 Be8 Ne5 Nxe5 fxe5+; bm Bxh7+; id "Hyatt-Quiet.3596"; acs 385;
rn2k2r/pp3ppp/3qpn2/4NP2/PbBp4/2N2P2/1P4PP/R1BQK2R w KQkq - acd 17; acn 70162700; ce 115; pv Bb5+ Nc6 Bf4 a6 Bxc6+ bxc6 Nd3 Qe7 Qb3 Nd5 Qc4 Rd8 Nxb4 Qxb4 Qxc6+ Kf8 Bc7 Nxc7 Qxc7 Re8 O-O dxc3 Rad1 Qb8 Qxc3 exf5 Rfe1 h5 Rxe8+ Qxe8 Rd6; bm Bb5+; id "Hyatt-Quiet.3312"; acs 356;
r1b1kb1r/pp3ppp/1nnqp3/8/8/1BN2N2/PP3PPP/R1BQ1RK1 w kq - acd 20; acn 82902308; ce 113; pv Nb5 Qb8 g3 Be7 Bf4 e5 Nxe5 Nxe5 Re1 Bg4 Bxe5 Bxd1 Raxd1 Qd8 Rxd8+ Rxd8 Nxa7 O-O Bc7 Bc5 Bxd8 Rxd8 Nb5 Kf8 Rc1 Be7 Kg2 Bf6 Rc2 g6 f4; bm Nb5; id "Hyatt-Quiet.0465"; acs 297;
rnbqk2r/pp1p1ppp/8/2bp4/4P1P1/1P1Q1N2/P1P1K1P1/R1B2B1R w kq - acd 19; acn 88889425; ce 112; pv exd5 Qf6 Qe4+ Qe7 Qxe7+ Kxe7 Bb2 Kf8 Ng5 d6 Kf3 h6 Re1 Nd7 Bd3 Bb4 c3 Bc5 b4 Bb6 c4 a5 b5 Nf6; bm exd5; id "Hyatt-Quiet.3844"; acs 455;
rnbqkb1r/5ppp/p3p3/1p1n4/P7/3BBN2/1PQ2PPP/RN3RK1 w kq - acd 19; acn 110941150; ce 111; pv axb5 Nb4 Qc3 Bb7 Rd1 a5 Bg5 Qd6 Nbd2 Nd7 Nc4 Qb8 Be2 Nd5 Qd4 a4; bm axb5; id "Hyatt-Quiet.3876"; acs 541;
r3kbnr/Nppq2pp/8/8/2P1p3/8/PP1P1PPP/R1BQK2R w KQkq - acd 20; acn 106639702; ce 107; pv Qh5+ Ke7 Nb5 c6 Nc3 Nf6 Qe2 Kf7 O-O Qd4 b3 Bb4 Re1 Rhd8 a4 Ra5 Qe3 Bc5 Qxd4 Bxd4 h3 Rf5; bm Qh5+; id "Hyatt-Quiet.2824"; acs 421;
rnb1k2r/pp2nqpp/2pb4/3p4/8/2NBN3/PPP2PPP/R1BQR1K1 w kq - acd 19; acn 81173792; ce 105; pv Nexd5 cxd5 Nb5 O-O Nxd6 Qxf2+ Kh1 Bg4 Qd2 Qf6 Qg5 Nbc6 Qxf6 Rxf6 Nxb7 Ng6 Bg5 Rff8; bm Nexd5; id "Hyatt-Quiet.3501"; acs 361;
rnbq1rk1/ppp3bp/3p4/3Pp2n/2P2p2/2N1BP2/PP1Q2PP/2KR1BNR w - - acd 21; acn 90964463; ce 104; pv Bf2 a5 g4 Nf6 Bd3 Na6 a3 b6 Nge2 Nc5 Bxc5 bxc5 g5 Nh5 Rhg1 Bf5 Bxf5 Rxf5 Ne4 Qe8 N2c3 Qg6 Nb5 Rf7 Kb1 a4; bm Bf2; id "Hyatt-Quiet.3803"; acs 568;
r1bq1rk1/1pp2ppp/p2b4/8/3QP1n1/2N2N2/PPP2PPP/R1B2RK1 w - - acd 20; acn 65658422; ce 103; pv Rd1 Qe8 Nd5 Ne5 Bf4 Nxf3+ gxf3 Bxf4 Nxf4 Be6 Qc5 Rc8 b3 b6 Qe5 Qc6 Rd3 Kh8 c4 Rce8 Qd4 f5 e5; bm Rd1; id "Hyatt-Quiet.0923"; acs 321;
r1bq1rk1/1pp2ppp/1npb4/p7/P2PP3/2N2N2/1PP2PPP/R1BQ1RK1 w - - acd 21; acn 47643297; ce 101; pv h3 f6 b3 Be6 Ne2 Nd7 Qd3 Qe7 Bf4 Bf7 Bxd6 cxd6 Rfe1 Rfe8 Ng3 g6 Qd2 Rec8 Rad1; bm h3; id "Hyatt-Quiet.0920"; acs 310;
r1bq1rk1/ppp3bp/3p1np1/3Ppp1n/2P1P3/2NBBP2/PP1Q2PP/2KR2NR w - - acd 20; acn 86330221; ce 90; pv exf5 gxf5 Nh3 Qe7 Rhe1 Nd7 g4 fxg4 fxg4 Nc5 Be2 Nf4 Nxf4 exf4 Bxc5 dxc5 Bf3 Qh4 Nb5 c6 Nc7 Rb8 Ne6 Bxe6 Rxe6 Bd4 Qe2 Be3+ Kb1 cxd5 Bxd5; bm exf5; id "Hyatt-Quiet.1341"; acs 497;
r2q1rk1/ppp1ppbp/2n2Pp1/8/2bP4/2N2N2/PP3PPP/R1B1KB1R w KQ - acd 20; acn 83987068; ce -88; pv fxg7 Kxg7 Bxc4 Nxd4 Nxd4 Qxd4 Bb3 a5 O-O a4 Rd1 Qb4 Bc2 Rfd8 Bxa4 e5 a3 Rxd1+ Bxd1 Qc4 Be2 Qb3 Bd1 Qe6 Bc2 f5 Be3; bm fxg7; id "Hyatt-Quiet.2159"; acs 499;
rnbqk1r1/4bp1p/p2ppn1Q/1p6/4P3/1NNB4/PPP2PPP/R1B1K2R w KQq - acd 21; acn 98319866; ce 88; pv O-O Nc6 f3 Ne5 Kh1 Qc7 Bd2 Bd7 a3 Nc4 Bxc4 bxc4 Nd4 Rg6 Qe3 h5 Nde2 h4 Nf4 Rg8 Rab1 Bc6 Qe2; bm O-O; id "Hyatt-Quiet.3811"; acs 486;
r1b1k2r/2pn1ppp/p4n2/1p6/1bBNPB2/2N2P2/PP4PP/R3K2R w KQkq - acd 17; acn 6074439; ce 87; pv Bb3 c5 Nc6 Bb7 Nxb4 cxb4 Ne2 Nc5 Bc2 O-O Kf2 Rac8 Rac1 Rfd8 Be3 h6 Bb1 Nd3+ Bxd3 Rxd3 Rxc8+ Bxc8 Nf4; bm Bb3; id "Hyatt-Quiet.0260"; acs 35;
rnb2rk1/ppq2ppp/4p1n1/2ppP1N1/3P4/P1PB4/2P2PPP/R1BQ1RK1 w - - acd 18; acn 6719401; ce 86; pv Qh5 h6 Nxf7 Nf4 Bxf4 Qxf7 Qxf7+ Rxf7 Bg3 c4 Bg6 Rf8 f4 Nc6 Bh4 Bd7 f5 Be8 Bxe8 Raxe8 fxe6 Rxe6 Rxf8+ Kxf8 Rb1 g5 Bg3 b6 Kf2 Kf7; bm Qh5; id "Hyatt-Quiet.3668"; acs 31;
rnb1k2r/ppp1b1pp/3p1n2/8/2NP2q1/2NBQ3/PPP3PP/R1B1K2R w KQkq - acd 16; acn 13641110; ce 86; pv O-O Nc6 d5 Nb4 Rf4 Qd7 Bf5 Qd8 Bxc8 Rxc8 Qe2 O-O Ne3 c5 a3 Na6 Nf5 Rf7 Qe6 Ra8 Ne4 Nc7 Nxe7+ Qxe7 Nxf6+ gxf6 Qxe7 Rxe7 Rxf6 Nxd5 Rxd6; bm O-O; id "Hyatt-Quiet.3521"; acs 48;
rnbqk2r/ppp2ppp/8/2bPp3/N1P1P3/P4N2/1P3nPP/R1BQKB1R w KQkq - acd 16; acn 5926707; ce 85; pv Qc2 Nxh1 Nxc5 b6 Nd3 O-O; bm Qc2; id "Hyatt-Quiet.3853"; acs 38;
r1bqkb1r/pp2pppp/1nn5/8/8/NBPp1N2/PP3PPP/R1BQ1RK1 w kq - acd 15; acn 3967058; ce 84; pv Nb5 a6 Nbd4 e6 Qxd3 Be7 Re1 Nxd4 Nxd4 O-O Bf4 Nd5 Be5 Bf6; bm Nb5; id "Hyatt-Quiet.1519"; acs 18;
r4rk1/pp1bnppp/2n1p3/q1ppP3/P2P4/2PB1N2/2PB1PPP/R2QK2R w KQ - acd 14; acn 8362908; ce 84; pv Bxh7+ Kxh7 Ng5+ Kg6 h4 f5 c4 Qa6 h5+ Kh6 Nxe6+ f4 Nxc5 Qxc4 Nxd7 Nxd4 c3 Nb3 Nxf8 Rxf8 Ra2 Rf5 Qe2 Nxd2 Rxd2 Qxc3 Rh3 Qc1+ Rd1 Qc7 Kf1; bm Bxh7+; id "Hyatt-Quiet.2878"; acs 47;
r1bqr1k1/pppnbppp/8/8/3pP3/2P2N2/PP3PPP/RNBQR1K1 w - - acd 16; acn 3007773; ce 83; pv cxd4 c6 h3 h6 Bf4 Nf6 Qd3 Be6 Nbd2 Qb6 b3 Rad8 Nc4; bm cxd4; id "Hyatt-Quiet.1619"; acs 19;
r1b1k2r/1pqpbppp/p3pn2/4Pn2/8/2N3P1/PPP2PBP/R1BQR1K1 w kq - acd 16; acn 5412155; ce 83; pv exf6 gxf6 Qd3 d6 Be4 Ng7 Bh6 f5 Bf3 O-O Qd4 e5 Nd5 Qd8 Nxe7+ Qxe7 Rad1 Rd8 Bxg7 Re8 Qxd6 Kxg7 Bg2 Qxd6 Rxd6 Rb8 Re3 Re7 Rd8; bm exf6; id "Hyatt-Quiet.0258"; acs 25;
rnb1k2r/pp2n1pp/2pb1q2/3p4/3P4/4NN2/PPP2PPP/R1BQKB1R w KQkq - acd 17; acn 5543392; ce 83; pv Be2 O-O O-O Be6 c4 Nd7 Qb3 b6 Re1 Rae8 Qa4 a5 cxd5 Bxd5 Nxd5 Nxd5 Bd2; bm Be2; id "Hyatt-Quiet.3498"; acs 36;
rnb2rk1/p4ppp/1p1qpn2/2p5/2PP4/P2BPN2/5PPP/R1BQK2R w KQ - acd 17; acn 8910737; ce 83; pv e4 Bb7 e5 Bxf3 Qc2 Qxd4 Bb2 Qh4 gxf3 Ne8 Be4 Nc7 Rd1 Qh3 f4 Nba6 Bxa8 Nxa8 Qe4 Qb3 Qe2 N8c7 O-O b5 cxb5 Nxb5 Rd2; bm e4; id "Hyatt-Quiet.3602"; acs 51;
2rq1rk1/p2p1ppp/bpn1pn2/2p5/2PP1N2/P1P1PQ2/5PPP/1RB1KB1R w K - acd 17; acn 7958640; ce -81; pv Bd3 Na5 e4 Bxc4 e5 Nd5 Nxd5 exd5 O-O Bxd3 Qxd3 Nc4 Qg3 Kh8 Qf3 d6 Qxd5 cxd4 exd6 dxc3 Bf4 Qd7; bm Bd3; id "Hyatt-Quiet.0108"; acs 41;
r1bqk2r/pp3pp1/4pn1p/2Pp4/1P2n3/P7/1Q2PPPP/R1B1KBNR w KQkq - acd 17; acn 5821955; ce 81; pv f3 Ng5 h4 Ngh7 Bf4 O-O e3 Re8 Bd6 b6 Kf2 Nf8 Bd3 a5 Ne2 bxc5 bxc5 Ba6 Bxa6 Rxa6 Rab1 Ng6 Qb5; bm f3; id "Hyatt-Quiet.1454"; acs 37;
r1bq1rk1/pp3ppp/4p3/b1pn4/1n1P4/1QN1PN2/PP1B1PPP/R3KB1R w KQ - acd 17; acn 6569498; ce 80; pv dxc5 Nxc3 Qxc3 Nc6 Qa3 Bxd2+ Nxd2 Qd5 Nf3 e5 Rd1 Qe4 Qc3 Bg4 Bd3 Qa4 O-O Bxf3 gxf3 Qxa2 Be4 Rad8; bm dxc5; id "Hyatt-Quiet.1263"; acs 41;
rn2k2r/pp2ppbp/1qp3p1/3p3n/3PPP2/2N1BP1P/PPPQ4/2KR1B1R w kq - acd 17; acn 5856732; ce 79; pv exd5 Ng3 Rg1 Nxf1 Rgxf1 Qa5 dxc6 Nxc6 d5 O-O-O Kb1 e6 d6 Rd7 Ne4 Qxd2 Rxd2 Rhd8 Rfd1 Bf8 Bc5 b6; bm exd5; id "Hyatt-Quiet.3306"; acs 32;
rnb1kb1r/1p3p1p/p2ppp2/8/N2NPP2/q7/P1P1B1PP/R2QK2R w KQkq - acd 18; acn 84293221; ce -79; pv c3 Nd7 O-O b5 Bh5 Bb7 Qg4 O-O-O Bxf7 Qxa4 Bxe6 Bxe4 Qh5 d5 Rfe1 Qa3 Bxd5 Bxd5 Qxd5 Qc5 Qa8+ Nb8 a4 b4 Rab1 Qd5 Qxd5 Rxd5 Re8+ Rd8 Rxd8+; bm c3; id "Hyatt-Quiet.3528"; acs 394;
rnb1kb1r/p4N2/2p1p2p/1p1nP3/2pP3q/2N5/PP3PPP/R2QKB1R w KQkq - acd 19; acn 93194458; ce 79; pv Nxh8 Bb4 Qd2 c5 g3 Qxd4 Bg2 Qxd2+ Kxd2 Nc6 Bf3 Ba5 Bh5+ Kf8 Rhe1 Nd4 Ng6+ Kg7 Nf4; bm Nxh8; id "Hyatt-Quiet.3541"; acs 470;
rnbq1rk1/pp4bp/3p4/2pPpp1n/2P5/2NBBP2/PP2N1PP/R2QK2R w KQ - acd 17; acn 5017284; ce 79; pv O-O Na6 a3 Bd7 Qd2 Qe7 Rae1 Rab8 Bg5 Bf6 Bh6 Bg7 Be3 Nc7; bm O-O; id "Hyatt-Quiet.3797"; acs 27;
rnb1k2r/pp3ppp/4p3/q7/3PP3/5N2/P2b1PPP/1R1QKB1R w Kkq - acd 19; acn 9327933; ce 78; pv Qxd2 Qxd2+ Kxd2 O-O Bc4 b6 Ke3 Bb7 Ne5 Rd8 Rhc1 Nd7 Nxd7 Rxd7 Bb5 Re7 f3 Rd8 Rc3 a6 Be2 b5 a4 bxa4; bm Qxd2; id "Hyatt-Quiet.3517"; acs 49;
rnb1k2r/2q1bppp/pp1ppn2/8/P3PP2/1NNB4/1PP3PP/R1BQ1RK1 w kq - acd 14; acn 2048557; ce 77; pv e5 Nfd7 exd6 Bxd6 Be3 O-O a5 Nc6 Qh5 g6 Qh6 b5 Ne4 Be7 Ng5 Nf6; bm e5; id "Hyatt-Quiet.3480"; acs 10;
rn2k2r/p3nppp/4p3/qpPpP3/b7/2P2N2/2PQ1PPP/R1B1KB1R w KQkq b6 acd 19; acn 10768282; ce 77; pv c5xb6/ep a7xb6 Nf3d4 OO Bf1b5 Ba4xb5 Ra1xa5 Ra8xa5 Nd4xb5 Ra5xb5 OO Rb5a5 Qd2d4 Nb8d7 f2f4 Rf8c8 Qd4b4 Rc8c5 Bc1e3 Ne7c6 Qb4b1 Rc5b5 Qb1d1 Ra5a7 c3c4 d5xc4 Qd1e2 Ra7a4 c2c3; bm c5xb6/ep; id "Hyatt-Quiet.3276"; acs 47;
r2qkb1r/1b1n1p1p/pp1ppnp1/8/2PNPP2/2NB4/PP2Q1PP/R1B2RK1 w kq - acd 17; acn 10176756; ce 77; pv e5 dxe5 fxe5 Bc5 Be3 Nh5 Be4 Qc7 Bxb7 Qxb7 Rad1 O-O Bh6 Rfd8 Ne4 Qc7 Nxc5 Qxc5 Be3 Qc7 Bg5 Re8 g4 Ng7 Nf3 h5; bm e5; id "Hyatt-Quiet.2414"; acs 49;
r2qkb1r/2p2pp1/p2p3n/1p2p1N1/3nP1p1/1B6/PPP2PP1/RNBQ1RK1 w kq - acd 14; acn 2628334; ce 77; pv f4 g4xf3/ep Rf1xf3 Nd4xf3+ Qd1xf3 Qd8f6 Bb3d5 Qf6xf3 g2xf3 Ra8b8 Bd5c6+ Ke8d8 Nb1c3 Rb8b6 Bc6d5 f7f6 Ng5e6+ Kd8d7 a2a4 c7c6 Bc1e3 Rb6b8 Bd5b3; bm f4; id "Hyatt-Quiet.2474"; acs 14;
r1bqk2r/pp3pp1/4pn1p/2Pp4/1P1Qn3/P7/4PPPP/R1B1KBNR w KQkq - acd 17; acn 3924337; ce 77; pv f3 Ng5 g4 Ngh7 Bf4 a5 h4 Nf8 Bd6 Ng6 b5 Ne7 Nh3 O-O Nf4 Ne8; bm f3; id "Hyatt-Quiet.1453"; acs 25;
rnb1k2r/ppp1qp1p/1n2p1p1/4b3/4N3/1B3Q2/PPP2PPP/RNB1K2R w KQkq - acd 16; acn 6067187; ce 77; pv Bg5 Qb4+ c3 Qa5 Nf6+ Kf8 Bh6+ Ke7 Ne4 f5 Ng5 Bf6 Nd2 Qe5+ Kd1 Bxg5 Bxg5+ Kf7 Kc2 Nc6 Rae1 Qa5 Qf4; bm Bg5; id "Hyatt-Quiet.3522"; acs 30;
rn2k2r/pQ2pp1p/4b1p1/q1P5/8/5N2/PP1bPPPP/R3KB1R w KQkq - acd 20; acn 61447751; ce 76; pv Nxd2 O-O b4 Qa4 e3 Nd7 a3 Rad8 Qe4 a5 b5 Nxc5 Qxa4 Nxa4 Nc4 Bxc4 Bxc4 Rc8 Bb3 Nc3 a4 Rfd8 h4 Rd3 Rh3 Rc5 Bd1 Rd8 h5; bm Nxd2; id "Hyatt-Quiet.3325"; acs 265;
r2qkb1r/p2b1ppp/4pn2/1pp5/2QP4/5NP1/PP2PPBP/R1B2RK1 w kq - acd 18; acn 8898091; ce 75; pv Qc3 Nd5 Qd2 Rc8 e4 Nf6 Rd1 cxd4 e5 Nd5 Nxd4 Qc7 Qe2 Qc4 Qe4 Bc5 Qg4; bm Qc3; id "Hyatt-Quiet.2501"; acs 49;
r1bqk2r/p4ppp/1p2pn2/2Pp4/1P1Qn3/P7/4PPPP/R1B1KBNR w KQkq - acd 16; acn 6785572; ce 75; pv f3 bxc5 bxc5 O-O fxe4 Nxe4 Nf3 f6 Nd2 e5 Qb4 Bf5 Nxe4 Bxe4 e3 Rb8 Qc3 Qe7 Bd2 f5 h4 f4 c6 fxe3 Qxe3 Rf6 Rc1; bm f3; id "Hyatt-Quiet.1414"; acs 37;
rnbq1rk1/p4pbp/3p2p1/1NpP4/4n3/3B1N1P/PP3PP1/R1BQK2R w KQ - acd 16; acn 6739426; ce 75; pv Bxe4 Qa5+ Nc3 Ba6 Qc2 Re8 Kd1 Nd7 Ng5 Rab8 Re1 Qb4 Bf3 Ne5 Qa4 Qb6 Be2; bm Bxe4; id "Hyatt-Quiet.3742"; acs 41;
rnbqk2r/5ppp/4pn2/pp6/2pP4/1PB1PN2/5PPP/R2QKB1R w KQkq - acd 16; acn 6007748; ce 75; pv bxc4 bxc4 Bxc4 O-O Bb4 Re8 O-O Nc6 Bc5 Ne4 Ba3 Nd6 Bd3 Ba6 Bxa6 Rxa6 Qc2; bm bxc4; id "Hyatt-Quiet.3839"; acs 28;

Of which the most interesting (to me) is this one:
[d]rn2k2r/pQ2pp1p/4b1p1/q1P5/8/5N2/PP1bPPPP/R3KB1R w KQkq - acd 20; acn 61447751; ce 76; pv Nxd2 O-O b4 Qa4 e3 Nd7 a3 Rad8 Qe4 a5 b5 Nxc5 Qxa4 Nxa4 Nc4 Bxc4 Bxc4 Rc8 Bb3 Nc3 a4 Rfd8 h4 Rd3 Rh3 Rc5 Bd1 Rd8 h5; bm Nxd2; id "Hyatt-Quiet.3325"; acs 265;

How long does it take your engine to find a low score for this position?
[/code]

[Will Singleton]

Given this file from Dr. Hyatt's site:
ftp://ftp.cis.uab.edu/pub/hyatt/tests/openings.epd

Most of the positions are fairly neutral. Those that are polar (centipawn evaluation of 3/4 of a pawn or more) are collected and shown below:

How long does it take your engine to find a low score for this position?
[/code]

That position is way too deep, but amateur figures the pv more or less. Thanks for doing the verification on Hyatt's file.

Not sure if the positions > 3/4 pawn would disqualify from the suite, since ostensibly side are flipped in any testing. Plus the fact that the number of non-quiet positions represent a small fraction of the total.

Will

[Dann Corbit]

Given this file from Dr. Hyatt's site:
ftp://ftp.cis.uab.edu/pub/hyatt/tests/openings.epd

Most of the positions are fairly neutral. Those that are polar (centipawn evaluation of 3/4 of a pawn or more) are collected and shown below:

How long does it take your engine to find a low score for this position?
[/code]

That position is way too deep, but amateur figures the pv more or less. Thanks for doing the verification on Hyatt's file.

Not sure if the positions > 3/4 pawn would disqualify from the suite, since ostensibly side are flipped in any testing. Plus the fact that the number of non-quiet positions represent a small fraction of the total.

Will

I agree. The thing that is interesting about the polar positions is that they appear frequently in high level games. Hence, they might make excellent traps for white to spring on black, if they can guide black into these positions (ostensibly through an opening book).

[bob]

Given this file from Dr. Hyatt's site:
ftp://ftp.cis.uab.edu/pub/hyatt/tests/openings.epd

Most of the positions are fairly neutral. Those that are polar (centipawn evaluation of 3/4 of a pawn or more) are collected and shown below:

How long does it take your engine to find a low score for this position?
[/code]

That position is way too deep, but amateur figures the pv more or less. Thanks for doing the verification on Hyatt's file.

Not sure if the positions > 3/4 pawn would disqualify from the suite, since ostensibly side are flipped in any testing. Plus the fact that the number of non-quiet positions represent a small fraction of the total.

Will

I agree. The thing that is interesting about the polar positions is that they appear frequently in high level games. Hence, they might make excellent traps for white to spring on black, if they can guide black into these positions (ostensibly through an opening book).

Btw, for God's sake rewrite that from "given this file from Dr. Hyatt's site" to "given this file from Bob's site..."

I get plenty of the "Dr. Hyatt" stuff here at the office.

Edit:

As far as the small number of unbalanced positions, I am surprised. No effort was made to deal with detecting and eliminating those. In fact, they are actually very useful since those positions were not reached by dummies, and they were reached multiple times. Just means the program has to figure things out. And as mentioned, since I always alternate colors, if a couple are really busted, it turns into a wash since it will be a win and a loss after color alternation.

[jhaglund]

I'm working on a position file for your cluster test. It should be done in a few days, since I'm hand entering them all...

I hope you'll be willing to give it a test run.

Joshua

[Dann Corbit]

Given this file from Dr. Hyatt's site:
ftp://ftp.cis.uab.edu/pub/hyatt/tests/openings.epd

Most of the positions are fairly neutral. Those that are polar (centipawn evaluation of 3/4 of a pawn or more) are collected and shown below:

How long does it take your engine to find a low score for this position?
[/code]

That position is way too deep, but amateur figures the pv more or less. Thanks for doing the verification on Hyatt's file.

Not sure if the positions > 3/4 pawn would disqualify from the suite, since ostensibly side are flipped in any testing. Plus the fact that the number of non-quiet positions represent a small fraction of the total.

Will

I agree. The thing that is interesting about the polar positions is that they appear frequently in high level games. Hence, they might make excellent traps for white to spring on black, if they can guide black into these positions (ostensibly through an opening book).

Btw, for God's sake rewrite that from "given this file from Dr. Hyatt's site" to "given this file from Bob's site..."

I get plenty of the "Dr. Hyatt" stuff here at the office.

Edit:

As far as the small number of unbalanced positions, I am surprised. No effort was made to deal with detecting and eliminating those. In fact, they are actually very useful since those positions were not reached by dummies, and they were reached multiple times. Just means the program has to figure things out. And as mentioned, since I always alternate colors, if a couple are really busted, it turns into a wash since it will be a win and a loss after color alternation.

About the Dr. Hyatt thing...
Sorry, but you're the godfather of computer chess, as far as I am concerned. (Oh, sure, I know there were chess pioneers before you but your crafty program is the first chess program *I* ever saw where I found the source code fascinating and amazing). And I was on pins and needles watching the games progress in Jakarta, a few months after I first saw the crafty program. I had already seen GnuChess and some other GUI chess program, but it was Crafty that captured my imagination.

As to the small number of unbalanced positions -- > I think we can explain this away because if a large percentage of them were highly unbalance, then the good players would avoid them. The good bookwriters would eliminate them from the downside perspective from their books and so the only entities who play them will be those who are yet unaware of them. Since many seem to be quite beneficial for white, I suspect that black books will put a big, fat "??" on them. Of course, I never believe anything until Rybka has hammered on it for 60 minutes. These are preliminary findings.

P.S.
I will publish the full analysis for this run, and when I rerun them at longer time control I will post the analysis for that also.

[Dann Corbit]

Here is my analyzed version of the test set, decorated with the analysis:
http://cap.connx.com/epd/hyatt-quiet.epd.bz2

You will find that some of the positions do not have an ID field. That is because they were already in my database from previous analysis as they belong to other test sets as well (I have stored procedures that I use to update everything and they do not update ID fields for existing records).

[bob]

I'm working on a position file for your cluster test. It should be done in a few days, since I'm hand entering them all...

I hope you'll be willing to give it a test run.

Joshua

Be happy to. How many positions are you shooting for? It does need to be a significant number although I can test a small number just to give you feedback on how they look.

[bob]

Here is my analyzed version of the test set, decorated with the analysis:
http://cap.connx.com/epd/hyatt-quiet.epd.bz2

You will find that some of the positions do not have an ID field. That is because they were already in my database from previous analysis as they belong to other test sets as well (I have stored procedures that I use to update everything and they do not update ID fields for existing records).

Looks like they are pretty balanced given enough time, which is exactly what I want. Those kinds of positions will highlight search changes.

[jhaglund]

It is not clear just yet how many...

I'm shooting for 2k, but 1500 is more realistic.

Joshua