Test position for weak (~2500) engines

Discussion of chess software programming and technical issues.

Moderators: hgm, Rebel, chrisw

Kieren Pearson
Posts: 70
Joined: Tue Dec 31, 2019 2:52 am
Full name: Kieren Pearson

Test position for weak (~2500) engines

Post by Kieren Pearson »

From a game between Halogen and CT800 this position arose:

[d]8/2R5/3r4/1b6/1k1pR3/4nP1P/3K4/8 w - - 2 83

As seen by the multi-pv output of Stockfish 12 only by sacrificing the rook can white draw here

Code: Select all

FEN: 8/2R5/3r4/1b6/1k1pR3/4nP1P/3K4/8 w - - 2 83

Stockfish 12:
 25/45	00:33	 58,803k	1,747k	-4.27	Rc7-b7 Ne3-c4+ Kd2-d3 Nc4-e5+ Kd3-c2 Rd6-d5 f3-f4 Ne5-c4 Kc2-d3 Nc4-d6+ Rb7xb5+ Nd6xb5 Re4-e2 Rd5-h5 Re2-h2 Rh5-h4 f4-f5 Nb5-d6 f5-f6 Kb4-c5 Kd3-e2 Kc5-c4 Ke2-f2 Nd6-e4+ Kf2-f3 Ne4xf6 Kf3-g3 Rh4-h6 h3-h4 d4-d3 Kg3-f3 Kc4-c3 Kf3-e3 Nf6-d5+ Ke3-f3 Rh6-f6+ Kf3-g3
 25/41	00:33	 58,803k	1,747k	-4.10	Rc7-c1 Ne3-c4+ Kd2-e2 d4-d3+ Ke2-f2 d3-d2 Rc1-d1 Bb5-c6 Re4-e1 Kb4-c3 Rd1xd2 Nc4xd2 Re1-a1 Rd6-f6 Kf2-g3 Bc6xf3 Kg3-f2 Bf3-e4+ Kf2-g3 Rf6-f3+ Kg3-g4 Be4-f5+ Kg4-g5 Nd2-e4+ Kg5-h6 Rf3xh3+ Kh6-g7 Kc3-d4 Ra1-a4+ Kd4-e3 Ra4-a5 Ke3-f4 Kg7-f7 Rh3-h2 Ra5-a8 Rh2-h6
 25/37	00:33	 58,803k	1,747k	-3.93	Re4-h4 Ne3-c4+ Kd2-d3 Nc4-e5+ Kd3-e4 Ne5-c6 Rh4-h7 Nc6-a5 Rc7-c1 d4-d3 Rc1-b1+ Na5-b3 Rh7-b7 d3-d2 Ke4-e5 Rd6-d8 Rb1-d1 Rd8-e8+ Ke5-f4 Kb4-c5 Rb7-c7+ Kc5-d4 Rc7-b7 Re8-f8+ Kf4-g5 Bb5-e2 Rb7-b4+ Kd4-c5 Rb4xb3 Be2xd1 Rb3-d3
 25/36	00:33	 58,803k	1,747k	-3.85	Kd2-e1 Ne3-c4 Re4xd4 Rd6xd4 Ke1-f2 Rd4-d3 Kf2-g3 Nc4-d2 Kg3-f4 Rd3xf3+ Kf4-g5 Nd2-e4+ Kg5-h6 Ne4-c5 Rc7-e7 Bb5-d3 h3-h4 Rf3-g3 Re7-c7 Nc5-a6 Rc7-f7 Rg3-g6+ Kh6-h5 Kb4-c3 Rf7-f8 Na6-b4 Rf8-c8+ Kc3-d4
 25/49	00:33	 58,803k	1,747k	-3.80	Re4-f4 Ne3-c4+ Kd2-d3 Nc4-a3+ Kd3-d2 Kb4-b3 Rc7-c1 Na3-c4+ Kd2-d1 Bb5-a6 h3-h4 Rd6-g6 Rf4xd4 Rg6-g1+ Kd1-e2 Rg1-g2+ Ke2-d3 Rg2-d2+ Kd3-e4 Ba6-b7+ Ke4-f4 Rd2xd4+ Kf4-g3 Kb3-b2 Rc1-e1 Rd4-d3 Kg3-g4 Bb7xf3+ Kg4-f4 Kb2-c3 Re1-e7 Bf3-a8 Re7-a7 Rd3-f3+ Kf4-g5 Ba8-d5 Ra7-a4 Rf3-h3 Ra4-a1 Rh3-h1
 26/53	00:33	 58,803k	1,747k	-3.75	h3-h4 Ne3-c4+ Kd2-d3 Nc4-e5+ Kd3-d2 Ne5xf3+ Kd2-d1 Bb5-a4+ Kd1-e2 Rd6-f6 Ke2-d3 Ba4-b5+ Kd3-c2 Bb5-c4 Rc7-b7+ Kb4-c5 Re4-e8 Bc4-e6 Rb7-a7 Be6-f5+ Kc2-d1 Kc5-c4 Kd1-e2 Bf5-g4 Ra7-c7+ Kc4-d5 Ke2-d3 Bg4-f5+ Kd3-e2 Nf3-e5 Ke2-d2 Ne5-c4+ Kd2-e1 d4-d3 Re8-d8+ Nc4-d6 Rc7-e7 Kd5-d4 Rd8xd6+ Rf6xd6 h4-h5 Rd6-a6
 26/15	00:33	 58,803k	1,747k	 0.00	Re4xd4+ Rd6xd4+ Kd2xe3 Rd4-h4 Rc7-g7 Kb4-c3 f3-f4 Rh4xh3+ Ke3-e4 Bb5-c4 f4-f5 Bc4-d3+ Ke4-f4 Bd3xf5
 26/17	00:33	 58,803k	1,747k	 0.00	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-f4 Rh6xh3 Kf4-g5 Rh3-h1 f3-f4 Bb5-d3 Rc7-d7 Kb4-c3 f4-f5 Rh1-f1 f5-f6 Rf1xf6
Halogen 6 played the disastrous Re4-h4 and went on to lose the game.

Halogen 7 on the other hand even without syzygy tables plays Re4xc3

Code: Select all

FEN: 8/2R5/3r4/1b6/1k1pR3/4nP1P/3K4/8 w - - 2 83

Halogen 7:
 1/3	00:00	 53	53k	-0.46	Re4xd4+ Rd6xd4+ Kd2xe3
 2/3	00:00	 348	348k	-0.47	Re4xe3 d4xe3+ Kd2xe3
 3/4	00:00	 1k	726k	-0.72	Re4xe3 d4xe3+ Kd2xe3 Bb5-f1
 4/6	00:00	 3k	957k	-0.73	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 h3-h4 Rh6xh4
 5/6	00:00	 8k	1,116k	-0.75	Re4xe3 d4xe3+ Kd2xe3 Bb5-c4 h3-h4 Kb4-c3
 6/8	00:00	 21k	1,091k	-0.81	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 h3-h4 Bb5-a4 Ke3-e4 Rh6xh4+
 7/9	00:00	 60k	1,102k	-1.00	Re4xd4+ Rd6xd4+ Kd2xe3 Rd4-h4 Rc7-c8 Rh4xh3 Rc8-b8 Rh3-h1 Ke3-e4
 8/9	00:00	 118k	1,254k	-1.02	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-f4 Rh6xh3 Kf4-g4 Bb5-f1 Rc7-e7
 9/7	00:00	 186k	1,184k	-1.15	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-e4 Rh6xh3 f3-f4
 10/11	00:00	 355k	1,226k	-1.14	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-e4 Rh6xh3 f3-f4 Kb4-b3 Ke4-e5 Rh3-h1 Ke5-f6
 11/12	00:00	 511k	1,226k	-1.21	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-e4 Rh6xh3 f3-f4 Rh3-h4 Ke4-e5 Rh4-h1 Rc7-b7 Rh1-h8
 12/12	00:00	 961k	1,346k	-1.21	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 h3-h4 Rh6xh4 f3-f4 Rh4-h5 Ke3-e4 Rh5-h4 Ke4-e5 Rh4-h1
 13/11	00:01	 1,643k	1,436k	-1.15	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 f3-f4 Rh6xh3+ Ke3-d4 Rh3-h4 Kd4-e5 Rh4-h5+ Ke5-e4
 14/16	00:02	 3,230k	1,456k	-1.22	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 f3-f4 Rh6xh3+ Ke3-e4 Rh3-c3 Rc7-a7 Kb4-c5 f4-f5 Kc5-d6 Ke4-f4 Rc3-c8 Kf4-g4 Rc8-h8
 15/13	00:03	 5,558k	1,454k	-1.25	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 f3-f4 Rh6xh3+ Ke3-e4 Bb5-d3+ Ke4-e5 Rh3-h5+ Ke5-d4 Bd3-b5 Rc7-b7
 16/19	00:05	 7,403k	1,474k	-1.22	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 f3-f4 Rh6xh3+ Ke3-e4 Bb5-d3+ Ke4-d4 Bd3-h7 Rc7-c6 Rh3-d3+ Kd4-e5 Rd3-e3+ Ke5-f6 Re3-h3 Kf6-e5 Rh3-h5+ Ke5-d4
Ras
Posts: 2487
Joined: Tue Aug 30, 2016 8:19 pm
Full name: Rasmus Althoff

Re: Test position for weak (~2500) engines

Post by Ras »

The CT800 V1.41 sees the sacrifice immediately, already at depth 2, and then just changes its mind on where to sacrifice (on e3 up to depth 10, on d4 from depth 11). Stockfish has both moves as draw.

The eval knowledge behind is that being a minor piece ahead without having pawns is pretty drawish.
Rasmus Althoff
https://www.ct800.net
JVMerlino
Posts: 1357
Joined: Wed Mar 08, 2006 10:15 pm
Location: San Francisco, California

Re: Test position for weak (~2500) engines

Post by JVMerlino »

Myrddin on 8 cores finds Rxd4 in 8 seconds, but then keeps waffling between Rxd4 and Rxe3. The score is never close to a draw, though, even after two minutes.
smcracraft
Posts: 737
Joined: Wed Mar 08, 2006 8:08 pm
Location: Orange County California
Full name: Stuart Cracraft

Re: Test position for weak (~2500) engines

Post by smcracraft »

Loaded this into Leela and it found it in 1/2 second.

I don't have a GPU and it was running on just one core.

My own program could not find it.
Dann Corbit
Posts: 12537
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Test position for weak (~2500) engines

Post by Dann Corbit »

Kieren Pearson wrote: Thu Sep 24, 2020 3:42 pm From a game between Halogen and CT800 this position arose:

[d]8/2R5/3r4/1b6/1k1pR3/4nP1P/3K4/8 w - - 2 83

As seen by the multi-pv output of Stockfish 12 only by sacrificing the rook can white draw here

Code: Select all

FEN: 8/2R5/3r4/1b6/1k1pR3/4nP1P/3K4/8 w - - 2 83

Stockfish 12:
 25/45	00:33	 58,803k	1,747k	-4.27	Rc7-b7 Ne3-c4+ Kd2-d3 Nc4-e5+ Kd3-c2 Rd6-d5 f3-f4 Ne5-c4 Kc2-d3 Nc4-d6+ Rb7xb5+ Nd6xb5 Re4-e2 Rd5-h5 Re2-h2 Rh5-h4 f4-f5 Nb5-d6 f5-f6 Kb4-c5 Kd3-e2 Kc5-c4 Ke2-f2 Nd6-e4+ Kf2-f3 Ne4xf6 Kf3-g3 Rh4-h6 h3-h4 d4-d3 Kg3-f3 Kc4-c3 Kf3-e3 Nf6-d5+ Ke3-f3 Rh6-f6+ Kf3-g3
 25/41	00:33	 58,803k	1,747k	-4.10	Rc7-c1 Ne3-c4+ Kd2-e2 d4-d3+ Ke2-f2 d3-d2 Rc1-d1 Bb5-c6 Re4-e1 Kb4-c3 Rd1xd2 Nc4xd2 Re1-a1 Rd6-f6 Kf2-g3 Bc6xf3 Kg3-f2 Bf3-e4+ Kf2-g3 Rf6-f3+ Kg3-g4 Be4-f5+ Kg4-g5 Nd2-e4+ Kg5-h6 Rf3xh3+ Kh6-g7 Kc3-d4 Ra1-a4+ Kd4-e3 Ra4-a5 Ke3-f4 Kg7-f7 Rh3-h2 Ra5-a8 Rh2-h6
 25/37	00:33	 58,803k	1,747k	-3.93	Re4-h4 Ne3-c4+ Kd2-d3 Nc4-e5+ Kd3-e4 Ne5-c6 Rh4-h7 Nc6-a5 Rc7-c1 d4-d3 Rc1-b1+ Na5-b3 Rh7-b7 d3-d2 Ke4-e5 Rd6-d8 Rb1-d1 Rd8-e8+ Ke5-f4 Kb4-c5 Rb7-c7+ Kc5-d4 Rc7-b7 Re8-f8+ Kf4-g5 Bb5-e2 Rb7-b4+ Kd4-c5 Rb4xb3 Be2xd1 Rb3-d3
 25/36	00:33	 58,803k	1,747k	-3.85	Kd2-e1 Ne3-c4 Re4xd4 Rd6xd4 Ke1-f2 Rd4-d3 Kf2-g3 Nc4-d2 Kg3-f4 Rd3xf3+ Kf4-g5 Nd2-e4+ Kg5-h6 Ne4-c5 Rc7-e7 Bb5-d3 h3-h4 Rf3-g3 Re7-c7 Nc5-a6 Rc7-f7 Rg3-g6+ Kh6-h5 Kb4-c3 Rf7-f8 Na6-b4 Rf8-c8+ Kc3-d4
 25/49	00:33	 58,803k	1,747k	-3.80	Re4-f4 Ne3-c4+ Kd2-d3 Nc4-a3+ Kd3-d2 Kb4-b3 Rc7-c1 Na3-c4+ Kd2-d1 Bb5-a6 h3-h4 Rd6-g6 Rf4xd4 Rg6-g1+ Kd1-e2 Rg1-g2+ Ke2-d3 Rg2-d2+ Kd3-e4 Ba6-b7+ Ke4-f4 Rd2xd4+ Kf4-g3 Kb3-b2 Rc1-e1 Rd4-d3 Kg3-g4 Bb7xf3+ Kg4-f4 Kb2-c3 Re1-e7 Bf3-a8 Re7-a7 Rd3-f3+ Kf4-g5 Ba8-d5 Ra7-a4 Rf3-h3 Ra4-a1 Rh3-h1
 26/53	00:33	 58,803k	1,747k	-3.75	h3-h4 Ne3-c4+ Kd2-d3 Nc4-e5+ Kd3-d2 Ne5xf3+ Kd2-d1 Bb5-a4+ Kd1-e2 Rd6-f6 Ke2-d3 Ba4-b5+ Kd3-c2 Bb5-c4 Rc7-b7+ Kb4-c5 Re4-e8 Bc4-e6 Rb7-a7 Be6-f5+ Kc2-d1 Kc5-c4 Kd1-e2 Bf5-g4 Ra7-c7+ Kc4-d5 Ke2-d3 Bg4-f5+ Kd3-e2 Nf3-e5 Ke2-d2 Ne5-c4+ Kd2-e1 d4-d3 Re8-d8+ Nc4-d6 Rc7-e7 Kd5-d4 Rd8xd6+ Rf6xd6 h4-h5 Rd6-a6
 26/15	00:33	 58,803k	1,747k	 0.00	Re4xd4+ Rd6xd4+ Kd2xe3 Rd4-h4 Rc7-g7 Kb4-c3 f3-f4 Rh4xh3+ Ke3-e4 Bb5-c4 f4-f5 Bc4-d3+ Ke4-f4 Bd3xf5
 26/17	00:33	 58,803k	1,747k	 0.00	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-f4 Rh6xh3 Kf4-g5 Rh3-h1 f3-f4 Bb5-d3 Rc7-d7 Kb4-c3 f4-f5 Rh1-f1 f5-f6 Rf1xf6
Halogen 6 played the disastrous Re4-h4 and went on to lose the game.

Halogen 7 on the other hand even without syzygy tables plays Re4xc3

Code: Select all

FEN: 8/2R5/3r4/1b6/1k1pR3/4nP1P/3K4/8 w - - 2 83

Halogen 7:
 1/3	00:00	 53	53k	-0.46	Re4xd4+ Rd6xd4+ Kd2xe3
 2/3	00:00	 348	348k	-0.47	Re4xe3 d4xe3+ Kd2xe3
 3/4	00:00	 1k	726k	-0.72	Re4xe3 d4xe3+ Kd2xe3 Bb5-f1
 4/6	00:00	 3k	957k	-0.73	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 h3-h4 Rh6xh4
 5/6	00:00	 8k	1,116k	-0.75	Re4xe3 d4xe3+ Kd2xe3 Bb5-c4 h3-h4 Kb4-c3
 6/8	00:00	 21k	1,091k	-0.81	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 h3-h4 Bb5-a4 Ke3-e4 Rh6xh4+
 7/9	00:00	 60k	1,102k	-1.00	Re4xd4+ Rd6xd4+ Kd2xe3 Rd4-h4 Rc7-c8 Rh4xh3 Rc8-b8 Rh3-h1 Ke3-e4
 8/9	00:00	 118k	1,254k	-1.02	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-f4 Rh6xh3 Kf4-g4 Bb5-f1 Rc7-e7
 9/7	00:00	 186k	1,184k	-1.15	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-e4 Rh6xh3 f3-f4
 10/11	00:00	 355k	1,226k	-1.14	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-e4 Rh6xh3 f3-f4 Kb4-b3 Ke4-e5 Rh3-h1 Ke5-f6
 11/12	00:00	 511k	1,226k	-1.21	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 Ke3-e4 Rh6xh3 f3-f4 Rh3-h4 Ke4-e5 Rh4-h1 Rc7-b7 Rh1-h8
 12/12	00:00	 961k	1,346k	-1.21	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 h3-h4 Rh6xh4 f3-f4 Rh4-h5 Ke3-e4 Rh5-h4 Ke4-e5 Rh4-h1
 13/11	00:01	 1,643k	1,436k	-1.15	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 f3-f4 Rh6xh3+ Ke3-d4 Rh3-h4 Kd4-e5 Rh4-h5+ Ke5-e4
 14/16	00:02	 3,230k	1,456k	-1.22	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 f3-f4 Rh6xh3+ Ke3-e4 Rh3-c3 Rc7-a7 Kb4-c5 f4-f5 Kc5-d6 Ke4-f4 Rc3-c8 Kf4-g4 Rc8-h8
 15/13	00:03	 5,558k	1,454k	-1.25	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 f3-f4 Rh6xh3+ Ke3-e4 Bb5-d3+ Ke4-e5 Rh3-h5+ Ke5-d4 Bd3-b5 Rc7-b7
 16/19	00:05	 7,403k	1,474k	-1.22	Re4xe3 d4xe3+ Kd2xe3 Rd6-h6 f3-f4 Rh6xh3+ Ke3-e4 Bb5-d3+ Ke4-d4 Bd3-h7 Rc7-c6 Rh3-d3+ Kd4-e5 Rd3-e3+ Ke5-f6 Re3-h3 Kf6-e5 Rh3-h5+ Ke5-d4
Not sure how you did it, but the picture does not match the EPD record shown. In the image, white has no rook to sac
Edit: nevermind. I see the white rook now. *Shrug*
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.
User avatar
hgm
Posts: 27788
Joined: Fri Mar 10, 2006 10:06 am
Location: Amsterdam
Full name: H G Muller

Re: Test position for weak (~2500) engines

Post by hgm »

KingSlayer also sees this instantly (at 1 ply, 12 nodes search). It just changes its mind after 0.01 sec where to sacrifice. Also here the reason is that drawishness of KRBKR is recognized in the evaluation.

Code: Select all

dep	score	nodes	time	(not shown:  tbhits	knps	seldep)
 18	 -0.14 	29.9M  	0:14.50	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 h3h4 e4e5 b5d3 e5d4 d3e2 d4e5 e2g4 f4f5 h4h5 c7b7 b4c5 b7c7 c5b5 c7f7 b5c6 e5e6
 17	 -0.14 	16.6M  	0:08.03	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 h3h4 e4e5 b5d3 e5d4 d3e2 d4e5 b4b5 f4f5 h4h5 e5f4 b5b6 c7e7
 16	 -0.16 	7.89M  	0:03.82	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 h3h4 e4e5 b5d3 e5d4 d3e2 c7b7 e2b5 d4e5 b4c5 f4f5 b5c6 b7c7
 15	 -0.15 	4.48M  	0:02.16	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 h3h4 e4e5 b5d3 e5d4 d3e2 c7b7 e2b5 d4e5 b4c5 f4f5 h4h5 e5e6
 14	 -0.16 	2.53M  	0:01.23	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 h3h4 e4e5 b5d3 e5d4 d3e2 c7b7 e2b5 d4e5 h4h2 e5d5 h2g2
 13	 -0.16 	1.29M  	0:00.64	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 b4a5 f4f5 a5b6 c7e7 b5c6 e4d4 h3h6 d4c4
 12	 -0.15 	804563	0:00.40	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 b4a5 f4f5 a5b6 c7e7 b5c6 e4d4
 11	 -0.17 	330918	0:00.17	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 b4a5 f4f5 a5b6 c7e7 b5d3 e4e5 h3h5 e7f7
 10	 -0.18 	171689	0:00.10	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 b4a5 f4f5 a5b6 c7f7 h3h4 e4e5 b6c6
  9	 -0.18 	88976  	0:00.06	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 h3a3 f4f5 b5d3 e4f4 a3a6 f4e5
  8	 -0.19 	34527  	0:00.04	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 h3g3 f4f5 b5d3 e4e5 g3e3 e5d4
  7	 -0.20 	18993  	0:00.03	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4 b5d3
  6	 -0.20 	12506  	0:00.01	e4d4 d6d4
  6	 -0.21 	9519    	0:00.01	e4e3 d4e3 d2e3 d6h6 h3h4 h6h4 f3f4
  5	 -0.20 	6716    	0:00.01	e4e3 d4e3 d2e3 d6h6 f3f4 h6h3
  4	 -0.20 	3247    	0:00.00	e4e3 d4e3 d2e3 d6h6 f3f4 h6h3
  3	 -0.12 	1208    	0:00.00	e4e3 d4e3 d2e3 b5d3
  3	 -0.20 	934      	0:00.00	e4d4 d6d4 d2e3 d4h4 f3f4 h4h3 e3e4
  2	 -0.11 	205      	0:00.00	s=1, phase=8, w=1 b=1 mm=1989250261 hm=2686616 ms=0 hs=2686480 P=fff8fffd e4d4 d6d4 d2e3 d4d3 e3e4
  2	 -0.12 	86        	0:00.00	s=1, phase=8, w=1 b=1 mm=1989250261 hm=2686616 ms=0 hs=2686480 P=fff8fffd e4e3 d4e3
  1	 -0.12 	12        	0:00.00	e4e3 d4e3 d2e3
  0	#