Project Sting (4)

[lech]

The project is designed to prove that artificial intelligence is not necessary to solve complex chess positions and maintain a high level of play.
To achieve this it uses open code of Stockfish.
The part 4 corrects and improves the last version (3).
It was important to solve the problem of VALUE_DRAW in zugzwang positions too.
It allows to work with slower (also 2 threads) computers.
Soon I will publish the beta version.

Some interesting examples (no lucky results):
[d]n2Bqk2/5p1p/Q4KP1/p7/8/8/8/8 w - - bm Qc8; (Mario Matous - 1975 - study)
Sting-sf-4, i5, Hash 1024, Threads 4 (only 14 seconds)

Code: Select all

info depth 20 seldepth 67 multipv 1 score cp 105 nodes 47088541 nps 6073589 time
 7753 pv a6d6 f8g8 g6h7 g8h8 d8a5 e8c8 a5d2 a8b6 d6b6 c8g4 d2g5 h8h7 b6b1 h7g8 b
1b8 g8h7 g5f4 g4e6 f6g5 e6g6 g5h4 f7f5 h4h3 g6g4 h3h2 g4e2 h2g3 e2e1 g3f3 e1f1 f
3e3 f1e1 e3d3 e1e4 d3c3 e4c6 c3b2 c6g2 b2a3 g2f3 a3a2 f3g2 a2b1 g2f1 b1c2 f1e2 c
2c1 e2e1 c1b2 e1e2 b2c1
info depth 21
info currmove a6d6 currmovenumber 1
info nodes 89043959 nps 6486302 time 13728
info currmove a6a8 currmovenumber 2
info currmove g6f7 currmovenumber 3
info currmove g6h7 currmovenumber 4
info currmove d8a5 currmovenumber 5
info currmove a6a5 currmovenumber 6
info currmove f6f5 currmovenumber 7
info currmove f6g5 currmovenumber 8
info currmove g6g7 currmovenumber 9
info currmove d8e7 currmovenumber 10
info currmove a6e2 currmovenumber 11
info currmove a6b7 currmovenumber 12
info currmove a6c8 currmovenumber 13
info depth 21 seldepth 0 multipv 1 score cp 113 lowerbound nodes 90392353 nps 64
66758 time 13978 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 121 lowerbound nodes 90393259 nps 64
59891 time 13993 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 133 lowerbound nodes 90395080 nps 64
60021 time 13993 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 151 lowerbound nodes 90397623 nps 64
60203 time 13993 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 179 lowerbound nodes 90401234 nps 64
53082 time 14009 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 219 lowerbound nodes 90403814 nps 64
53266 time 14009 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 281 lowerbound nodes 90406702 nps 64
46570 time 14024 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 373 lowerbound nodes 90409736 nps 64
46786 time 14024 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 509 lowerbound nodes 90413229 nps 64
39688 time 14040 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 714 lowerbound nodes 90433959 nps 64
41165 time 14040 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 1021 lowerbound nodes 90463770 nps 6
435954 time 14056 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 1482 lowerbound nodes 90519285 nps 6
433038 time 14071 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 2174 lowerbound nodes 90561350 nps 6
428717 time 14087 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 3211 lowerbound nodes 90572762 nps 6
422689 time 14102 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 4766 lowerbound nodes 90577579 nps 6
423030 time 14102 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 7098 lowerbound nodes 90578297 nps 6
415802 time 14118 pv a6c8
info currmove a6c8 currmovenumber 1
info depth 21 seldepth 0 multipv 1 score cp 10596 lowerbound nodes 91269633 nps
6415240 time 14227 pv a6c8
info depth 21 seldepth 72 multipv 1 score cp 10596 nodes 91269633 nps 6415240 ti
me 14227 pv a6c8 f8g8 d8c7 e8c8 g6f7 g8h8 c7e5 c8c5 e5b2 a8c7 b2a1 a5a4 a1b2 a4a
3 b2a1 a3a2 a1b2 a2a1n b2a1

[d]5K2/k4p1p/5p1p/1p3p2/bP6/1p6/1P5p/7B w - - bm Ke8(e7);
Sting-sf-4, i5, Hash 1024, Threads 4 (24 minutes)

Code: Select all

info depth 47 seldepth 61 multipv 1 score cp 0 nodes 6139677931 nps 9722510 time
 631491 pv f8f7 a7b6 f7f6 f5f4 f6e5 b6c7 e5f4 c7d6 f4f5 h6h5 h1f3 d6d7 f5f4 d7d8
 f4e5 d8c7 f3g2 c7d7 e5f5 d7d8 f5f4 d8e8 g2h1 e8e7 f4f5 e7d8 f5f4 d8e7
info depth 48
info currmove f8f7 currmovenumber 1
info nodes 6590248990 nps 9729171 time 677370
info currmove h1f3 currmovenumber 2
info nodes 8549735790 nps 9770545 time 875052
info currmove f8e7 currmovenumber 3
info depth 48 seldepth 0 multipv 1 score cp 8 lowerbound nodes 9902380913 nps 97
74076 time 1013127 pv f8e7
info nodes 9902380913 nps 9774076 time 1013127
info currmove f8e7 currmovenumber 1
info depth 48 seldepth 0 multipv 1 score cp 16 lowerbound nodes 11322254964 nps
9737630 time 1162732 pv f8e7
info nodes 11322254964 nps 9737630 time 1162732
info currmove f8e7 currmovenumber 1
info depth 48 seldepth 0 multipv 1 score cp 28 lowerbound nodes 11344756994 nps
9734262 time 1165446 pv f8e7
info nodes 11344756994 nps 9734262 time 1165446
info currmove f8e7 currmovenumber 1
info depth 48 seldepth 0 multipv 1 score cp 46 lowerbound nodes 11713894103 nps
9721509 time 1204946 pv f8e7
info nodes 11713894103 nps 9721380 time 1204962
info currmove f8e7 currmovenumber 1
info nodes 12265507853 nps 9716970 time 1262277
info currmove f8f7 currmovenumber 2
info currmove h1f3 currmovenumber 3
info currmove h1g2 currmovenumber 4
info nodes 12856419423 nps 9723064 time 1322260
info currmove h1d5 currmovenumber 5
info nodes 13074968547 nps 9724737 time 1344506
info currmove h1c6 currmovenumber 6
info nodes 13252791034 nps 9725527 time 1362681
info currmove f8g7 currmovenumber 7
info nodes 13503790046 nps 9727098 time 1388265
info currmove f8g8 currmovenumber 8
info nodes 13760821693 nps 9731585 time 1414037
info currmove f8e8 currmovenumber 9
info depth 48 seldepth 0 multipv 1 score cp 73 lowerbound nodes 13862243113 nps
9725951 time 1425284 pv f8e8
info nodes 13862243113 nps 9725951 time 1425284
info currmove f8e8 currmovenumber 1
info depth 48 seldepth 0 multipv 1 score cp 114 lowerbound nodes 13927014341 nps
 9722228 time 1432492 pv f8e8
info nodes 13927014341 nps 9722228 time 1432492
info currmove f8e8 currmovenumber 1
info nodes 14082361820 nps 9719326 time 1448903
info currmove f8f7 currmovenumber 2
info currmove f8e7 currmovenumber 3
info currmove h1f3 currmovenumber 4
info currmove h1d5 currmovenumber 5
info currmove h1g2 currmovenumber 6
info currmove f8g7 currmovenumber 7
info currmove h1c6 currmovenumber 8
info currmove f8g8 currmovenumber 9
info currmove h1b7 currmovenumber 10
info currmove h1e4 currmovenumber 11
info currmove h1a8 currmovenumber 12
info depth 48 seldepth 81 multipv 1 score cp 109 nodes 14082365087 nps 9719121 t
ime 1448934 pv f8e8 a7b8 e8d7 b8a7 d7c7 h6h5 h1d5 h7h6 d5h1 h5h4 c7c8 a7b6 c8b8
h4h3 b8c8 b6a7 c8c7 h6h5 h1c6 h5h4 c6h1 a7a6 h1f3 a6a7 f3h1

[lech]

One more example:
[d]4r1k1/pp5p/2p1bR2/2P5/2B1n3/2B2K2/PP5P/8 w - - bm Rxe6;
Sting-sf-4, i5, Hash 1024, Threads 4 (2 seconds)

Code: Select all

info depth 23 seldepth 42 multipv 1 score cp 117 nodes 9190552 nps 4789240 time
1919 pv c4e6 e8e6 f6e6 e4g5 f3f4 g5e6 f4e5 e6c5 e5d6 c5a6 c3d4 a6b4 d4a7 b4a2 d6
c7 a2b4 c7b7 b4d3 b7c6 d3b2 c6d5 g8f7 a7d4 b2d3 d5e4 d3c1 d4e5 c1b3 e4d5 f7e7
info depth 24
info depth 24 seldepth 0 multipv 1 score cp 129 lowerbound nodes 10443168 nps 48
50519 time 2153 pv f6e6
info nodes 10443168 nps 4814738 time 2169
info currmove f6e6 currmovenumber 1
info depth 24 seldepth 0 multipv 1 score cp 141 lowerbound nodes 10657035 nps 48
09131 time 2216 pv f6e6
info currmove f6e6 currmovenumber 1
info depth 24 seldepth 0 multipv 1 score cp 159 lowerbound nodes 10864297 nps 48
02960 time 2262 pv f6e6
info currmove f6e6 currmovenumber 1
info depth 24 seldepth 0 multipv 1 score cp 186 lowerbound nodes 11173280 nps 48
05711 time 2325 pv f6e6
info currmove f6e6 currmovenumber 1
info depth 24 seldepth 0 multipv 1 score cp 227 lowerbound nodes 11541895 nps 48
03119 time 2403 pv f6e6
info currmove f6e6 currmovenumber 1
info depth 24 seldepth 0 multipv 1 score cp 288 lowerbound nodes 14138434 nps 48
98972 time 2886 pv f6e6
info currmove f6e6 currmovenumber 1
info nodes 48330599 nps 5378433 time 8986
info currmove c4e6 currmovenumber 2
info currmove f3e4 currmovenumber 3
info currmove f6f4 currmovenumber 4
info currmove h2h4 currmovenumber 5
info currmove c3e5 currmovenumber 6
info currmove f6f7 currmovenumber 7
info currmove f6g6 currmovenumber 8
info currmove c4d3 currmovenumber 9
info currmove a2a3 currmovenumber 10
info currmove f3e3 currmovenumber 11
info currmove f6f5 currmovenumber 12
info currmove f6h6 currmovenumber 13
info currmove f3f4 currmovenumber 14
info currmove f6f8 currmovenumber 15
info currmove f3g2 currmovenumber 16
info currmove f3e2 currmovenumber 17
info currmove c4b3 currmovenumber 18
info currmove c3b4 currmovenumber 19
info currmove b2b4 currmovenumber 20
info currmove h2h3 currmovenumber 21
info currmove c4d5 currmovenumber 22
info currmove c4b5 currmovenumber 23
info currmove c3d4 currmovenumber 24
info currmove b2b3 currmovenumber 25
info currmove a2a4 currmovenumber 26
info currmove c4e2 currmovenumber 27
info currmove c4a6 currmovenumber 28
info currmove c3a5 currmovenumber 29
info currmove c3d2 currmovenumber 30
info currmove c3e1 currmovenumber 31
info currmove c4f1 currmovenumber 32
info depth 24 seldepth 54 multipv 1 score cp 307 nodes 48346594 nps 5361716 time
 9017 pv f6e6 e4g5 f3g4 g5e6 g4f5 g8f7 c3f6 h7h6 h2h4 b7b5 c5b6 a7b6 a2a4 b6b5 a
4b5 c6b5 c4b3 e8a8 b3e6 f7e8 e6d5 a8a4 d5c6 e8f7 h4h5 a4a6 c6d5 f7e8 f6e5 e8d7 e
5f4 d7e7 f5e5

[jdart]

What is the source of the 2nd position?

Houdini seems to prefer Bxe6+ although it is similar in score to Rxe6. (I have not done a super-deep search though).

--Jon

[Ajedrecista]

Hello Jon:

What is the source of the 2nd position?

I just searched the FEN string in Google without the side to move (I removed w - - bm Rxe6; from the search). It is a position from the Scottish Correspondence Championship of 1998 (game Gillam - Mackintosh (Mackintosh or Macintosh, I do not know), played in 1999). Take a look at page 4 of this PDF.

Thanks to CSS I reach a link of the well-known Walter Eigenmann:

http://glareanverlag.wordpress.com/2009 ... achzug-37/

With the given solution:

(Gillam-Macintosh, CorrGame 1999)
35.Txe6!! [ 35.Lxe6+? Txe6 36.Txe6 Sg5+ 37.Kf4 ( 37.Ke3 Sxe6=) 37...Sxe6+ 38.Ke5 Sxc5=]
35…Sg5+ 36.Kg4 Sxe6 [ 36...Txe6 37.Kxg5+-] 37.Kf5 Kf7 38.Lf6 b5
[ 38...a5 39.a4+-; 38...h6 39.b4+-] 39.cxb6 axb6 40.a4 b5
[ 40...h6 41.La2+-] 41.axb5 cxb5 42.Ld5 b4 [ 42...h6 43.La2+-]
43.b3 Kf8 44.Lxe6 +-

It was the 35th move of the game according to the complete FEN string found with Google. Please do a search in Google with the FEN string for find further results.

Regards from Spain.

Ajedrecista.

[lech]

Sting-sf-4 beta : http://www.mediafire.com/?4hsf6b2qeoar203

[zullil]

Sting-sf-4 beta : http://www.mediafire.com/?4hsf6b2qeoar203

Since this is derived from Stockfish, shouldn't you include the source? All I see is a .exe file.

[lech]

For all programmers (solders too) the source of this beta is possible by PM.
The full version will be published with source code.
Don't worry Sting (based on 2.1.1 version) is not stronger (ELO!) than Stockfish 3.0.

[lech]

I would like to read more constructive comments.
Am I among thieves and crooks?
Some quick comments may change the version 4. Please go ahead.

[zullil]

I would like to read more constructive comments.
Am I among thieves and crooks?
Some quick comments may change the version 4. Please go ahead.

I am impressed by the examples you posted. The current version of Stockfish seems unable to solve these. Do the changes you've made weaken Stockfish in regular play? Is there a significant loss of strength?

[lech]

I would like to read more constructive comments.
Am I among thieves and crooks?
Some quick comments may change the version 4. Please go ahead.

I am impressed by the examples you posted. The current version of Stockfish seems unable to solve these.

Not Stockfish only!

[ Do the changes you've made weaken Stockfish in regular play? Is there a significant loss of strength?

Indeed, a hourse without tail may be faster.