Author: Dieter Buerssner
Date: 11:39:32 01/07/06
Go up one level in this thread
On January 06, 2006 at 20:30:22, Vincent Lejeune wrote:
Thanks for the interesting positions, Vincent.
Below, some analysis of Yace 0.9987 on Centrino 1.73 GHz (notebook with rather
slow HD, which slows down some of the analysis in the endgame positions).
Analysis time was varying, depending on my interest and some random factors. Not
too surprising, the pawn endgame postions were solved rather convincingly in
decent time. I skipped some of the positions for now.
It is a long post, I highlighted some positions (of which the correctness might
be doubtful, or of which I thought, I made some hopefully constructive comment).
It is a long post, perhaps one or the other reader wants to comment only one or
a few positions. I'd like to suggest, to delete the non relevant quoting and
only to keep the relevant context, to make it easier for all readers. I myself
kept all the context, because I commented almost everything :-)
>Let your favorite engine for loooong time on each position
>If you find refutation(s) or have some comments, please post here !
>
>8/5p1p/8/6k1/8/6P1/5PP1/7K w - -
>The line is 1. Kh2 f6 2. Kh1! f5 3. Kg1 Kf6 4. Kf1 Ke5 5. Ke1 Kd5 6. Kd1 Ke5
>7.Kc2! Kd4 8. Kd2 Ke4 9. Ke2 h6 10. Kf1! Ke5 11. Kg1 Kf6 12. Kh2 Kg6 13. Kh3,
>and now the win is not hard to find.
>see : http://members.aol.com/MuellerLamprecht/LoesungenBrainsoftheEarthengl.html
11 s for Yace to find the move, score of +8 after 1:30
15/36 0:09 +0.95 1.Kg1 h5 2.Kf1 Kf5 3.Ke2 Ke6 4.Kd3 Kd5
5.Ke3 Ke5 6.Kd3 Kd5 7.Ke3 Ke5 8.Kd3 Kd5 (12.966.731) 1402
TB:9
15/36 0:11 +0.96++ 1.Kh2 f6 2.Kh1 f5 3.Kg1 Kf6 4.Kf1 Ke5
5.Ke1 h5 6.Ke2 Ke6 7.Kd3 Kd5 8.f3 Ke5
9.Ke3 Kd6 10.g4 Ke5 11.gxf5 h4 12.f6 Ke6
13.Kd3 Kf5 14.f4 (15.374.688) 1320 TB:48
15/36 0:13 +1.35++ 1.Kh2 f6 2.Kh1 f5 3.Kg1 Kf6 4.Kf1 Ke5
5.Ke1 h5 6.Ke2 Ke6 7.Kd3 Kd5 8.f3 Ke5
9.Ke3 Kd6 10.g4 Kd7 11.gxf5 Kd8
12.f6 Ke8 13.f7+ Kf8 14.f4 (17.421.206) 1337 TB:48
15/36 0:13 +1.35 1.Kh2 f6 2.Kh1 f5 3.Kg1 Kf6 4.Kf1 Ke5
5.Ke1 Kd5 6.Kd1 Ke5 7.Kc2 Kd5 8.Kd3 Ke6
9.Kd4 Kf6 10.Ke3 Kg5 11.Kf3 h6
12.Ke3 Kg4 13.Kd3 Kg5 14.Kd4 (17.480.320) 1338 TB:48
20/48 1:30 +8.00 1.Kh2 f6 2.Kh1 f5 3.Kg1 Kf6 4.Kf1 Ke5
5.Ke1 Kd5 6.Kd1 Ke5 7.Kc2 Ke4 8.Kc3 h6
9.Kc4 h5 10.Kc3 Ke5 11.Kd3 Kf6
12.Ke3 Kf7 13.f3 Ke8 14.g4 (130.455.277) 1441 TB:310
>8/5k2/p7/2pP1Pp1/2P3P1/2N3K1/P4N2/5q2 b - - In this difficult ending taken from
>an actual game, the only move that wins is 1. ... Kf8 !!. The tempting move 1.
>... Qxc4 only gives a draw after White answers 2. Nfe4.
>-> 1. ... Kf8 !! 2. d6 Qc4 3. Nfe4 Qf1 4. Nf2 Qc1 5. Nce4 Qe3 6. Kg2 c4 7. f6
>Ke8 8. Ng3 c3 9. d7 Kd7 10. f7 Qe7 11. Nd3 c2 12. Nf5 Qf7 13. Ne5 Ke6
found after 16 minutes
>[D] 8/1p6/3k4/4n3/p2KP2p/P6N/1P6/8 b - -
>This endgame may seem simple, but it has some surprises. Here, the only move
>that wins is 1. ... b6!. If at once 1. ... b5 ?, Black cannot improve the
>position of his king. By playing the text, Black loses a move and forces White
>to play in the critical position.
>The Principal Variation as given in the book goes like this:
>1. ... b6! 2. Nf4 b5 3. Nh3 (if 3. Kc3 then 3. ... Ng6! 4. Nh3 Kc5 wins) 3. ...
>Nc6+ 4. Ke3 Kc5 5. Kd3 b4 6. ab+ Kxb4 7. Kc2 Nd4+ 8. Kb1 (or 8. Kd3 Ne6 9. Kc2
>Kc4 wins) 8. ... Ne6 9. Ka2 (if 9. Kc2 then 9. ... Kc4 wins) 9. ... Kc4 10. Ka3
>Kd4 11. Kxa4 Kxe4 12. b4 Kf3 13. b5 Kg2
Not found in 12 minutes. Yace prefers Ng4:
21/43 9:12 +1.04 1...Sg4 2.Sg5 Sf2 3.e5+ Kc6 4.Ke3 Sd1+
5.Kd3 Sxb2+ 6.Kc3 Sd1+ 7.Kd2 Sf2
8.Ke3 (845.197.744) 1529 TB:2.713
Playing on in analyse mode at a pace of about 1 move/min, black won after
1...Sg4.
[Event "?"]
[Site "?"]
[Date "2006.01.07"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[SetUp "1"]
[FEN "8/1p6/3k4/4n3/p2KP2p/P6N/1P6/8 b - -"]
1... Ng4 2. Ng5 Nf2 3. e5+ Kc6 4. Ke3 Nd1+ 5. Kd2 Nxb2
6. Kc3 Nd1+ 7. Kd2 Nf2 8. Ke3 Ng4+ 9. Kf4 Nh6 10. e6 Ng8
11. Nf3 h3 12. Kg3 Kd5 13. Kxh3 Kxe6 14. Kg3 Kd5 15. Kf4
Kc4 16. Ke3 Nf6 17. Kd2 Kb3 18. Nd4+ Kxa3 {TB mate due to Shredder Online TBs} *
>8/3R1P2/1ppP1p2/3r4/8/K7/p4k2/8 w - - 0 1
>1. Kxa2 ? f5; 2. Ra7 Re5 !
>1. Kb2 !! a1=Q+; 2. Kxa1 Ra5+; 3. Kb2 Rb5+; 4. Kc3 Rc5+; 5. Kd4 f5; 6. Ra7
>Rd5+;7. Kc3 Rc5+; 8. Kb2 Rb5+; 9. Ka1 Re5; 10 Ra2+ and wins
found after 1:32:
15/41 1:11 +7.34 1.Kxa2 f5 2.Ta7 Te5 3.f8D Te2+ 4.Kb3 Te3+
5.Ka4 Te4+ 6.Ka3 Te3+ 7.Kb4 Te4+
8.Kc3 Te3+ 9.Kb2 Te2+ 10.Ka1 Te1+
11.Kb2 Te2+ 12.Ka1 Te1+ 13.Kb2 Te2+ (78.843.871) 1107
TB:308
15/41 1:32 +7.35++ 1.Kb2 a1D+ 2.Kxa1 Ta5+ 3.Kb2 Kg1
4.f8D f5 (98.901.376) 1069 TB:749
15/41 1:49 +7.77 1.Kb2 Ta5 2.Ka1 Ta8 3.Te7 Tf8 4.d7 b5
5.Te8 Txf7 6.d8D Kg2 7.Db6 f5
8.Dxc6+ Kh2 9.Th8+ Kg1 10.Kxa2 (117.390.254) 1076 TB:899
>[D] 8/2k5/Kp3pp1/7p/P4P2/8/6PP/8 b - - 0 1 Azmaiparashvili played now Kc6? and lost
>the game. I belive that only move to draw is 46...h4!
found after 19 s:
12/32 0:15 -6.88 1...g5 2.fxg5 fxg5 3.Ka7 g4 4.g3 Kc6
5.Kb8 Kd6 6.Kb7 Kc5 7.Kc7 Kd4 8.Kxb6 Ke3
9.a5 Kf2 10.a6 Kg2 11.a7 Kxh2 12.Kc6 Kg2
13.Kd5 Kh3 14.a8D Kxg3 (22.902.683) 1436
12/32 0:19 -6.87++ 1...h4 2.g4 f5 3.g5 h3 4.Ka7 Kc6
5.Ka8 b5 6.a5 b4 7.Kb8 Kd6 8.a6 b3
9.a7 b2 10.a8D b1D+ 11.Db7 (28.596.877) 1453
12/32 0:29 -0.03 1...h4 2.g4 hxg3 3.hxg3 Kc6 4.g4 f5
5.g5 Kc5 6.Kb7 b5 7.a5 b4 8.a6 b3
9.a7 b2 10.a8D b1D+ 11.Kc8 Db4
12.Da6 (43.435.701) 1479
12/32 0:32 -0.03 1...h4 2.g4 hxg3 3.hxg3 Kc6 4.g4 f5
5.g5 Kc5 6.Kb7 b5 7.a5 b4 8.a6 b3
9.a7 b2 10.a8D b1D+ 11.Kc8 Db4
12.Da6 (47.906.191) 1481
13/34 0:38 0.00 1...h4 2.g4 hxg3 3.hxg3 Kc6 4.g4 f5
5.g5 Kc5 6.Kb7 b5 7.axb5 Kxb5 8.Kc7 Kc5
9.Kc8 Kd5 10.Kd7 Ke4 11.Ke6 Kxf4 (57.607.565) 1499
A search for the second best move yielded:
18/45 5:38 -6.81 1...Kc6 2.h4 f5 3.g3 Kc7 4.Ka7 Kc6
5.Kb8 Kd5 6.Kb7 Ke4 7.Kxb6 Kf3 8.a5 Kxg3
9.Kc5 Kxf4 10.a6 Kg4 11.a7 Kxh4
12.Kd4 g5 13.Ke5 f4 14.a8D Kg3 (534.883.968) 1578 TB:231
This is a rather convincing score, that supports (maybe even proves) that your
assumption is correct.
>r4rk1/1b1n1pb1/3p2p1/1p1Pq1Bp/2p1P3/2P2RNP/1PBQ2P1/5R1K w - - Nf5 or Nxh5
not found in 5 minutes.
>r3n3/3b2kp/1qNp2p1/1p1Pp2n/1P2Pp2/3B1N1P/1Q3PP1/2R3K1 w - - Rc5 (but not sure)
found after 13 min, no convincing score:
13/43 4:54 +0.32 1.Sfxe5 dxe5 2.Sxe5 Df6 3.Tc2 Lc8
4.Lxb5 Sd6 5.Ld3 Ta7 6.Dd4 Te7 7.Sc6 Tc7 (277.019.458)
941
13/45 13:06 +0.33++ 1.Tc5 Da6 2.Txb5 Lxc6 3.dxc6 Sc7
4.Td5 Dxc6 5.Ta5 Txa5 6.bxa5 Db6
7.axb6 Kg8 (698.203.598) 887
13/49 14:12 +0.33 1.Tc5 Da6 2.Txb5 Lxc6 3.dxc6 Sc7
4.Td5 Dxc6 5.Ta5 Sf6 6.Lb5 (761.325.890) 893
>r1bq1rk1/1p3ppp/p1pp2n1/3N3Q/B1PPR2b/8/PP3PPP/R1B3K1 w - - Rxh4 (but not sure)
Not found in 5 minutes.
>[D] 3rrk2/2p1n1bQ/1p3pp1/1q2B1N1/p1b1PP2/6PB/P1P5/3R1RK1 w - - Be6! mate
found after 1 minute (no mate score yet). Bxf6 seems to be a forced win, too.
Maybe other moves, too.
>[D] 4k1rr/1R1R3p/8/8/8/2K5/7P/8 w k - 1.Rd2 (zugzwang) Rf8 2.Kc2 Rhg8 3.Kb2 h6 4.h3
>h5 [ 4...Rh8 5.h4 Rhg8 6.h5] 5.h4 Rh8 6.Ra7 Rhg8 7.Ka2 +-
Not found in 5 minutes. Seems very difficult. Even at the end of the 4...Rh8
line, I don't get a convincing score in 2 minutes. Is the position known to be
correct?
>[D] 1q6/k7/3p2nP/pp1Qp3/1P2B3/P7/2P5/1K6 w - - h7! mate in 9
Not a good test position for a playing program, maybe good for a mate solver.
Bxg6 wins easily.
>[D] 8/2B2pPb/7k/4K3/3p2P1/8/8/8 w - - g8N+ hard for some
Found after 5 s. After 50 s changes to Kf6, after 1:17 back to g8N+ with score
of 2.5. Win seems not easy and rather far away.
>2r1r3/p3bk1p/1pnqpppB/3n4/3P2Q1/PB3N2/1P3PPP/3RR1K1 w - - Rxe6(not sure, Qh3 may
>be as strong)
Rxe6 found after 6 s, switched to Qh3 after 8 s, stays with Qh3 for at least 2
minutes.
>[D] r2qr1k1/pbp1bpp1/1p2p2p/3n4/3P1PN1/2PB3R/PP1BQ1PP/R5K1 w - - Nxh6!
Not found in 5 minutes, rather good score for 1. Qe4:
11/41 3:24 +1.35 1.De4 Ld6 2.Dh7+ Kf8 3.Se5 Sf6
4.Dh8+ Sg8 5.Lh7 Lxe5 6.fxe5 Ke7
7.Dxg7 c5 (204.740.327) 999
>2r3k1/2r1ppb1/3p2p1/pq4Pn/1p1BP3/1B3P2/PPPQ4/1K1R3R w - - Rxh5 (not sure yes)
Found after 11 s, lost after 54 s (Bxg7), found again after 2:17.
>8/8/p7/n1p5/PpPp4/1P1P4/2K1k1B1/8 b - - Nxb3 with >+2
Found after 57 s, score of 1.62 after 1:23 (note, pawn is 0.8 in Yace, so this
can be compared to score of > 2 with many engines), score > 2 fter 2:00, > 3
after 2:16, +8.01 after 5:47.
>r1q1rbk1/2pp1ppp/p1b5/1p1NP3/3Q4/3R2B1/PPP2PPP/5RK1 w - - Nf6+!
Not found in 3 minutes.
>1r2kb1r/2qb1pp1/p7/3Q2Pn/8/4B2P/PPP3B1/R4RK1 w k - 0 1 Rad1!
>[D] 3b4/3B4/4n3/5kp1/8/6P1/6K1/8 b - - 0 1 AM g4??
Not avoided in 10 minutes. Can you please show a sample winning line (I did not
search for one myself yet).
>rn1qk2r/p2nbp2/1p2p1p1/2ppP1Np/3P1PP1/P1N1B2Q/1PP4P/R3KR2 w Qkq - Nxf7!
>3r2k1/pp3ppp/5q2/8/1n2Q3/7P/PPP2PK1/R1B5 w - - 0 1 AM Qxb4??
>8/3k3p/1p4p1/p2p1p2/PP1K1P1P/2P3P1/8/8 b - - b5!! with >+2 or +3 for black
b5 found imediately, but only with small positive score. I have reason to
assume, that Yace would win this pawn endgame in a practical game with any
reasonable time control. My guess (from going into the main line), it would take
around 1 hour, to get a score > 3.
>1r1rb1k1/5ppp/4p3/1p1p3P/1q2P2Q/pN3P2/PPP4P/1K1R2R1 w - - bm Rxg7+; id "Aemis
>Test Suite No.30"; c0 "Vescovi G.,2455 - Gschwendtner S.,2270, Groningen (open)
>1994";1. Rxg7+ Kxg7 2. h6+ Kf8 3. Qf6 Bc6 4. c3 *
>4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 1. Qxe5 fxe5 2. Rf1 *
>8/6p1/P1b1pp2/2p1p3/1k4P1/3PP3/1PK5/5B2 w - - 0 1 bm Bg2; id Bishop Rivalry;
Not found in 5 minutes. I believe to have tried this position before, and it was
solved in over night analysis. The position was discussed in CCC several times?
>4B3/5K2/5P1b/2p5/2p5/1k6/8/8 w - - 0 1 bm Kg6; id Merry-go-round;
Kg6 found after 1s, no convincing score.
>8/1p5k/4pP2/6B1/8/1P6/8/1n4K1 w - - 0 1 bm f7; id Good bishop vs Limited knight;
f7 found after 6 s, score +1.93 after 26 s, > 2 after 57 s
>8/p1p2pp1/1pP3p1/3Pb2k/4P3/5P1P/8/3K2B1 w - - 0 1 bm f4; id The Winning Arrow;
f4 found after 18 s with draw score after 21 s, fail high after 1:35.
>8/6r1/Pp5p/7k/p7/7K/6P1/R7 w - - 0 1 bm Ra3; id The Rebel Rook;
Found after 2 s, score +8 after 7 s.
>[D] 1k3b1q/pP2p1p1/P1K1P1Pp/7P/2B5/8/8/8 w - - 0 1 bm Kd5; id Totally Paralysed;
Found after 31 s, score +0.41 after 1:50, after 2:07 Bb5 with 0.42, not
switching back to Kd5 within 5 minutes. Can you give some analysis?
>1knB4/8/7p/5R2/7K/8/3r4/7B w - - 0 1 bm Rb5; id The Weekend Nicholas;
Found after 1 s. Score stays at 0.9 for at least 1 min.
>4kr2/5p1K/3p1Q2/1p4P1/4P3/1PP5/7b/8 w - - 0 1 bm Kh6; id The Giant King; 1. Kh6
>Be5 2. Kg7!! Bh2 3. c4! bxc4 4. e5!! Bxe5 5. bxc4 Bxf6+ 6. gxf6 Rh8 7.
>Kxh8 Kd7 8. Kg8 Ke6 9. Kg7
Not found in 5 min.
>3k4/2Rnn3/1p2pbp1/pP1p1pBp/P2P1P1P/4KPP1/4B3/8 w - - 0 2 Rxd7+ draw !?
Found imediately, with slightly rising score (starting at +3.87)
>8/1p6/p4p1p/2p1P2k/5P2/6pP/1P4P1/6K1 w - - 0 1 bm f5; id Rare Pawn ending; 1. f5
> fxe5 2. h4 c4 3. Kf1 a5 4. Ke2 a4 5. Kf3 c3 6. Kxg3 cxb2 7. Kh3 b1=Q 8.
>g4+
Found after 12 s, score +3 after 22 s, +6.4 after 41 s.
>r7/6k1/5R2/p3p3/Pp1pPp2/1PpP1Pp1/2P3P1/1K6 w - - 0 1 bm Rh6;stalemate threat
Not found in 5 minutes
>8/8/pp6/3b2K1/1P6/PN2k3/8/8 w - - id A-pawn and wrong bishop; bm Na5;
This is from Jouni's endgame testsuite? It is not simple, even with rook pawn
vs. wrong B code. Found after 27 s, but showing bad score (-3.5).
>k4B1b/p6P/Kp6/1P3P2/4P3/R2P4/2P2p2/6n1 w - - id The Arrow's Point; bm Ra1;1.
>Ra1! Bxa1 2. c3! Bxc3 3. d4! Bxd4 4. e5! Bxe5 5. f6! Bxf6 6. Bg7 Bxg7
> 7. h8=Q+ Bxh8
Found after 9 s, score stays around -2.3 for a while, after 56 s fail high to
-0.45, which seems reasonable.
>8/2b2k1K/1pPp1p2/1P1P1P2/5B2/8/8/8 w - - id The Pawnbishop; bm Be3;
Found imediately, score > 8 after 8 s.
>8/5p2/6PR/7P/1k6/p7/7r/1K6 w - - id The Original Mozes; bm g7;
Found after 3:11
Regards,
Dieter
This page took 0.01 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.