I saw that Nemorino 3.04 is available ...

[Frank Quisinsky]

Hi there,

made a little check for my engines overview.
http://www.amateurschach.de/main/_engines.htm

Nice ...
Nemorino is verry interesting!
If I understand ... some bugfixes.

Thanks to the programmer of Nemorino!

Best
Frank

[Norbert Raimund Leisner]

[Event "Computer Chess Game"]
[Site "QUADCORE-I5"]
[Date "2017.09.22"]
[Round "1"]
[White "StuckFish 0.1 (UCI2WB)"]
[Black "Nemorino (UCI2WB)"]
[Result "1/2-1/2"]
[TimeControl "40/240"]
[Annotator "1... -0.29"]

1. e4 e6 {-0.29/17 8} 2. d4 d5 {-0.38/16 4} 3. exd5 exd5 {-0.27/17 6} 4. h3
Nc6 {+0.01/16 3} 5. Nf3 Bd6 {+0.00/17 2.9} 6. Nc3 Nf6 {+0.00/17 2.2} 7. Bg5
Be6 {+0.00/17 5} 8. Qe2 O-O {+0.24/20 10} 9. Qd2 h6 {+0.19/18 6} 10. Bh4
Be7 {+0.24/18 4} 11. Bb5 Ne4 {+0.31/18 2.7} 12. Nxe4 dxe4 {+0.13/20 4} 13.
Bxc6 bxc6 {+0.13/21 2.7} 14. Bxe7 Qxe7 {+0.05/22 4} 15. Ne5 Qd6
{+0.08/21 7} 16. Qc3 Bd5 {+0.07/21 8} 17. Ng4 h5 {+0.18/19 10} 18. Ne3 a5
{+0.29/18 4} 19. Qc5 Qxc5 {+0.28/18 4} 20. dxc5 Be6 {+0.28/18 8} 21. b3 a4
{+0.39/18 10} 22. b4 a3 {+0.66/20 7} 23. c3 Rfd8 {+0.97/20 4} 24. Ke2 f5
{+0.74/21 4} 25. Rhd1 Rxd1 {+0.74/21 4} 26. Kxd1 Rd8+ {+0.99/24 5} 27. Ke1
Rd3 {+1.09/22 8} 28. Nc2 Rxc3 {+1.69/23 8} 29. Nd4 Kf7 {+1.95/22 4} 30.
Nxc6 Rc2 {+2.03/23 6} 31. Nd4 Rb2 {+2.09/23 7} 32. Nb5 Bc4 {+3.31/24 13}
33. Nxc7 Bd3 {+6.64/27 9} 34. Rc1 Rxa2 {+8.63/30 8} 35. Nd5 Re2+
{+8.63/29 3} 36. Kd1 Rxf2 {+8.83/29 6} 37. Rc3 a2 {+9.38/28 5} 38. Ra3 Ke6
{+9.38/27 15} 39. Nc3 a1=Q+ {+9.54/26 2.1} 40. Rxa1 Rf1+ {+9.97/27 3} 41.
Kd2 Rxa1 {+10.18/27 4} 42. c6 Rg1 {+10.70/28 4} 43. c7 Rxg2+
{+10.98/30 2.7} 44. Kd1 Kd7 {+13.51/29 4} 45. Nd5 g5 {+17.06/29 9} 46. Ne7
Kxc7 {+106.54/27 8} 47. Nxf5 Rh2 {+114.80/29 7} 48. Ne3 Rxh3 {+116.60/24 3}
49. Kd2 Rxe3 {+1000.11/25 43} 50. Kc3 Kc6 {+1000.08/18 4} 51. b5+ Kc5
{+1000.06/20 2.1} 52. b6 Re1 {+1000.05/25 2.0} 53. Kd2 Rb1
{+1000.04/56 1.8} 54. Kc3 Rxb6 {+1000.03/127 0.8} 55. Kd2 Kd4
{+1000.02/127 0.2} 56. Ke1 e3 {+1000.01/127 0.2}
{False win claim: 'mate' but bare king} 1/2-1/2

[Event "Computer Chess Game"]
[Site "QUADCORE-I5"]
[Date "2017.09.22"]
[Round "2"]
[White "Nemorino (UCI2WB)"]
[Black "StuckFish 0.1 (UCI2WB)"]
[Result "0-1"]
[TimeControl "40/240"]
[Annotator "1. +0.23"]

1. e4 {+0.23/17} d6 2. Nc3 {+0.37/18 8} Bd7 3. Nf3 {+0.55/19 11} e6 4. d4
{+0.59/19 6} Be7 5. Be2 {+0.59/18 2.7} a6 6. O-O {+0.79/18 6} h6 7. Bf4
{+0.78/18 5} Bg5 8. Nxg5 {+0.88/19 9} hxg5 9. Be3 {+0.91/19 3} Nc6 10. d5
{+0.91/17 2.2} Ne5 11. Qd4 {+0.90/16 4} Kf8 12. Rad1 {+0.91/15 4} exd5 13.
Nxd5 {+0.99/17 8} Rh4 14. Bxg5 {+1.66/18 7} Qxg5 15. f4 {+1.68/21 10} Qh6
16. fxe5 {+1.90/20 11} Rxh2 17. Qd3 {+0.16/16 6} Qh4 18. g3 {+0.69/19 19}
Rh1+ 19. Kf2 {+0.65/19 3} Rxf1+ 20. Bxf1 {+0.53/20 38} Qd8 21. Ne3
{+0.61/18 5} Be6 22. Bg2 {+0.59/15 1.9} Rc8 23. b3 {+0.87/15 4} Qg5 24.
exd6 {+0.76/21 5} Rd8 25. Qc3 {+0.83/20 4} cxd6 26. Qb4 {+0.85/19 2.8} Qb5
27. Qxb5 {+0.97/19 2.1} axb5 28. e5 {+0.96/19 3} Ke7 29. Nd5+
{+0.94/19 2.1} Bxd5 30. Rxd5 {+0.94/19 1.5} b4 31. Ke3 {+0.99/18 2.0} Nh6
32. exd6+ {+0.96/17 3} Rxd6 33. Rb5 {+0.86/18 2.2} Re6+ 34. Kd2
{+0.90/19 2.4} Rd6+ 35. Kc1 {+0.85/21 4} b6 36. Rxb4 {+0.83/20 4} Nf5 37.
g4 {+0.97/19 3} Nh4 38. Be4 {+1.22/20 3} Ng6 39. Bxg6 {+1.33/21 4} Rxg6 40.
Kd2 {+1.38/22 5} Rd6+ 41. Ke3 {+1.46/20 4} Re6+ 42. Re4 {+8.73/26 7} Rxe4+
43. Kxe4 {+10.04/30 13} Kd6 44. a4 {+11.04/30 4} Kc5 45. c3 {+11.82/30 12}
Kc6 46. b4 {+12.02/28 6} Kd6 47. c4 {+17.16/28 18} Kc6 48. Ke5
{+39.05/30 13} Kd7 49. Kd5 {+58.28/29 8} Kc7 50. a5 {+58.28/28 1:47} bxa5
51. bxa5 {+101.05/24 2.4} f6 52. c5 {+116.70/24 3} Kb7 53. c6+
{+116.80/25 1.0} Kc7 54. Kc5 {+1000.08/25 9} Kd8 55. Kd6 {+1000.07/18 0.7}
Ke8 56. c7 {+1000.06/24 0.6} Kf7 57. c8=Q {+1000.05/29 0.5} Kg6 58. Qf5+
{+1000.04/43 0.5} Kf7 59. Qh7 {+1000.03/127 0.3} f5 60. gxf5
{+1000.02/127 0.2} Kf8 61. Ke6 {+1000.01/127 0.2}
{False win claim: 'mate'} 0-1

[Event "Computer Chess Game"]
[Site "QUADCORE-I5"]
[Date "2017.09.22"]
[Round "3"]
[White "StuckFish 0.1 (UCI2WB)"]
[Black "Nemorino (UCI2WB)"]
[Result "1-0"]
[TimeControl "40/240"]
[Annotator "1... -0.29"]

1. e4 e6 {-0.29/17 8} 2. d4 d5 {-0.38/16 4} 3. exd5 exd5 {-0.27/17 6} 4.
Nf3 Bd6 {-0.23/16 4} 5. Bd3 Ne7 {-0.14/17 8} 6. c4 dxc4 {-0.09/18 7} 7.
Bxc4 O-O {-0.17/18 5} 8. h3 Nbc6 {+0.16/17 4} 9. Nc3 Nf5 {+0.17/17 5} 10.
Kf1 h6 {+0.13/16 4} 11. a3 a6 {+0.17/18 7} 12. Qd3 Nh4 {+0.32/19 5} 13.
Nxh4 Qxh4 {+0.19/20 8} 14. Be3 Re8 {+0.19/19 4} 15. Bd5 Ne7 {+0.37/17 6}
16. Be4 c6 {+0.46/18 3} 17. d5 Nxd5 {+0.82/20 9} 18. Nxd5 cxd5 {+0.90/21 7}
19. Bxd5 Be6 {+1.02/21 17} 20. Rd1 Be5 {+0.95/20 18} 21. Bxe6 Rxe6
{+0.91/21 3} 22. Qd7 Qc4+ {+1.37/19 3} 23. Qd3 Qa2 {+1.30/18 2.9} 24. Qb1
Qxb2 {+1.29/20 3} 25. a4 Qa3 {+1.40/20 12} 26. Qxb7 Rae8 {+1.48/19 2.9} 27.
Qe4 Bg3 {+1.91/20 5} 28. Qd4 Rd6 {+6.40/22 4} 29. Qg4 Rxd1+ {+7.14/23 2.9}
30. Qxd1 Rxe3 {+7.59/25 4} 31. fxg3 Re6 {+7.64/25 4} 32. Qd8+ Kh7
{+11.05/26 6} 33. Qd2 Qa1+ {+17.29/24 6} 34. Kf2 Rf6+ {+19.88/24 4} 35. Ke2
Qxh1 {+114.14/24 11} 36. Ke3 Qg1+ {+114.54/25 7} 37. Ke4 Re6+
{+114.54/23 1.4} 38. Kd5 Qb6 {+114.54/23 10} 39. Qf4 Qc6+ {+1000.17/18 2.1}
40. Kd4 Qxa4+ {+1000.16/17 1.0} 41. Kd5 Qd7+ {+1000.08/20 2.1} 42. Kc4 Rc6+
{+1000.06/21 2.2} 43. Kb4 a5+ {+1000.05/25 2.2} 44. Kxa5 Qd5+
{+1000.04/35 5} 45. Kb4 Rb6+ {+1000.03/45 2.2} 46. Kc3 Rb3+
{+1000.02/127 2.2} 47. Kc2 Qd3+ {+1000.01/127 0.2}
{False win claim: 'mate'} 1-0

[Event "Computer Chess Game"]
[Site "QUADCORE-I5"]
[Date "2017.09.22"]
[Round "4"]
[White "Nemorino (UCI2WB)"]
[Black "StuckFish 0.1 (UCI2WB)"]
[Result "0-1"]
[TimeControl "40/240"]
[Annotator "2. +0.33"]

1. e4 {+0.75/1} e6 2. d4 {+0.33/16 9} d5 3. Nc3 {+0.31/16 3} Nc6 4. Nf3
{+0.43/17 4} a6 5. e5 {+0.44/17 6} Be7 6. Bd3 {+0.47/19 4} Bd7 7. a3
{+0.58/20 13} Kf8 8. O-O {+0.77/18 2.3} h5 9. Be3 {+0.76/19 5} Nh6 10. Qd2
{+0.78/18 4} Ng4 11. Bf4 {+0.79/19 2.1} Kg8 12. Ne2 {+0.83/18 5} h4 13. h3
{+1.27/20 3} Nh6 14. c4 {+1.38/20 6} dxc4 15. Bxc4 {+1.38/20 2.5} Nf5 16.
Rac1 {+1.34/19 5} Rh5 17. d5 {+1.41/20 11} Na7 18. Rfd1 {+2.17/17 7} Ba4
19. dxe6 {+3.14/19 7} Bxd1 20. exf7+ {+3.99/22 9} Kf8 21. Ned4 {+8.39/20 7}
Bxf3 22. Ne6+ {+10.66/23 7} Kxf7 23. Nxd8+ {+11.06/23 6} Kf8 24. Ne6+
{+13.38/21 2.3} Kf7 25. Nxc7+ {+14.92/24 12} Kf8 26. gxf3 {+20.24/21 5} Rd8
27. Ne6+ {+28.49/22 8} Ke8 28. Nxd8 {+97.96/21 6} Bxd8 29. Be6
{+1000.05/26 2.2} Nc6 30. Qd7+ {+1000.01/127 0.3}
{False win claim: 'mate'} 0-1

Norbert

[Guenther]

{+1000.05/26 2.2} Nc6 30. Qd7+ {+1000.01/127 0.3}
{False win claim: 'mate'} 0-1

Norbert

To me it looks like all false claims are done by your strange 'Stuckfish'.
Please check again and repost in a different thread (if necessary). Thanks.

[hgm]

It seems like the mate score reported by Nemorino is off by one. E.g. after the last move (Ke6) in the second game the position is
[d]5k2/6pQ/4K3/P4P2/8/8/8/8 b - - 2 61
but the score reported by Nemorino with Ke6 seems to be mate in 1 (1000.01). While it obviously is mate in 2. I suppose that UCI2WB correctly translated this score from the UCI to WB standard for mate scores, or every UCI engine would have this problem, and I would have heard it long before. To make sure of that I would have to see the debug file (created whe WinBoard is ru with additional option -debug), where the debug option of UCI2WB is also switched on (I think this is on by default).

UCI2WB considers a mate-in-1 score sent with the PV info as a result claim, as UCI engines have no other way for claimig results.

[Guenther]

It seems like the mate score reported by Nemorino is off by one. E.g. after the last move (Ke6) in the second game the position is
[d]5k2/6pQ/4K3/P4P2/8/8/8/8 b - - 2 61
but the score reported by Nemorino with Ke6 seems to be mate in 1 (1000.01). While it obviously is mate in 2. I suppose that UCI2WB correctly translated this score from the UCI to WB standard for mate scores, or every UCI engine would have this problem, and I would have heard it long before. To make sure of that I would have to see the debug file (created whe WinBoard is ru with additional option -debug), where the debug option of UCI2WB is also switched on (I think this is on by default).

UCI2WB considers a mate-in-1 score sent with the PV info as a result claim, as UCI engines have no other way for claimig results.

In the first game there was this claim:

{False win claim: 'mate' but bare king} 1/2-1/2

The one who had the lonely K was 'Stuckfish' of course.

This makes me sure that it was always 'Stuckfish' who claimed the
false mate claims... (mostly with K+P - Stuckfish also was reported for other strange bugs before too - Nemorino never)

[Florentino]

According to UCI spec, the mate score should be given in moves not plies,
so after having played 61. Ke6 I think
[d]5k2/6pQ/4K3/P4P2/8/8/8/8 b - - 2 61 is mate in 1 (Black makes move and White gives mate)
Do I miss something?

[Florentino]

Please note Nemorino_3.04 is no new Release. It should have the same strength as Nemorino 3.00. It only contains some refactorings and bug fixes. Ne need for new tests!

[hgm]

In the first game there was this claim:

{False win claim: 'mate' but bare king} 1/2-1/2

The one who had the lonely K was 'Stuckfish' of course.

This makes me sure that it was always 'Stuckfish' who claimed the
false mate claims... (mostly with K+P - Stuckfish also was reported for other strange bugs before too - Nemorino never)

The "but bare king" is not part of the claim, but appended by XBoard to any result message that would make the bare King win. So Nemorino claims 'mate' here, but one move early, which would normally forfeit it, and grant Stuckfish the win. But because Stuckfish has a bare King, this win is then corrected to a draw.

[hgm]

According to UCI spec, the mate score should be given in moves not plies,
so after having played 61. Ke6 I think
[d]5k2/6pQ/4K3/P4P2/8/8/8/8 b - - 2 61 is mate in 1 (Black makes move and White gives mate)
Do I miss something?

Yes. Namely that the score is always for the position before the move, not the one after it.

After 61.Ke6 it would not be "mate in 2", but "mated in 1" (i.e. info score mate -1).

[Florentino]

I'm not convinced and I don't find any indication that UCI Specification says that the reported scores should be related to the position before the move!
Here is the excerpt of the spec

Code: Select all

	* score
		* cp 
			the score from the engine's point of view in centipawns.
		* mate 
			mate in y moves, not plies.
			If the engine is getting mated use negativ values for y.

This clearly states that a negative sign should only be sent by an engine, which is getting mated, so this is no statement whether the score is related to the position before or after the move.

But if I do a MultiPV analysis I think everyone expects to get the score after the move. Else I would always report the same score for all moves.

As it's irrelevant for UCI I will fix it and report mate scores (but only mate scores) for the position before the move.