NoCastlingChess openings-sets released

[pohl4711]

I created 2 openings sets with No Castling Chess openings for engine-testing:

Classical human openings, balanced endpositions. Queens still on the board.
All lines filtered out of MegaBase 2020. No double endpositions.
KomodoDragon 1.0 analyzed each endposition on a QuadcorePC with 15 seconds thinking-time.
Allowed Eval [-0.39;+0.59]
Files are sorted by Elo:
Both players 2600+ Elo followed by both players 2500+ Elo followed by both players 2400+ Elo

Then these 2 TAGs were added, which makes all 4 castlings illegal:
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - 0 1"]

All games containing (now) illegal castling-moves were deleted (around 25% of all games)

2 different openings sets (as PGN- and EPD-file):
NCC_6mvs_2400+_30k = 29990 different endpositions. Both players 2400 Elo or better
NCC_6mvs_2600+_6k = 6743 different endpositions. Both players 2600 Elo or better



Personally, I doubt, that this make much sense and give low draw-rates, but if anybody wants to try it out, here it is:

https://www.sp-cc.de/files/nocastlingchess_openings.7z

(or download from my website, "Downloads & Links"-section)

Enjoy!

[Chessqueen]

I created 2 openings sets with No Castling Chess openings for engine-testing:

Classical human openings, balanced endpositions. Queens still on the board.
All lines filtered out of MegaBase 2020. No double endpositions.
KomodoDragon 1.0 analyzed each endposition on a QuadcorePC with 15 seconds thinking-time.
Allowed Eval [-0.39;+0.59]
Files are sorted by Elo:
Both players 2600+ Elo followed by both players 2500+ Elo followed by both players 2400+ Elo

Then these 2 TAGs were added, which makes all 4 castlings illegal:
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - 0 1"]

All games containing (now) illegal castling-moves were deleted (around 25% of all games)

2 different openings sets (as PGN- and EPD-file):
NCC_6mvs_2400+_30k = 29990 different endpositions. Both players 2400 Elo or better
NCC_6mvs_2600+_6k = 6743 different endpositions. Both players 2600 Elo or better



Personally, I doubt, that this make much sense and give low draw-rates, but if anybody wants to try it out, here it is:

https://www.sp-cc.de/files/nocastlingchess_openings.7z

(or download from my website, "Downloads & Links"-section)

Enjoy!

Very interesting work, which make you the pioneer to start testing top engines with Non Castling Chess, I truly believe that Komodo Dragon 2.5 will excel in this type of chess, Just like in FRC or Chess960. I hope that CCRL, and CEGT start testing very soon.

[Chessqueen]

I created 2 openings sets with No Castling Chess openings for engine-testing:

Classical human openings, balanced endpositions. Queens still on the board.
All lines filtered out of MegaBase 2020. No double endpositions.
KomodoDragon 1.0 analyzed each endposition on a QuadcorePC with 15 seconds thinking-time.
Allowed Eval [-0.39;+0.59]
Files are sorted by Elo:
Both players 2600+ Elo followed by both players 2500+ Elo followed by both players 2400+ Elo

Then these 2 TAGs were added, which makes all 4 castlings illegal:
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - 0 1"]

All games containing (now) illegal castling-moves were deleted (around 25% of all games)

2 different openings sets (as PGN- and EPD-file):
NCC_6mvs_2400+_30k = 29990 different endpositions. Both players 2400 Elo or better
NCC_6mvs_2600+_6k = 6743 different endpositions. Both players 2600 Elo or better



Personally, I doubt, that this make much sense and give low draw-rates, but if anybody wants to try it out, here it is:

https://www.sp-cc.de/files/nocastlingchess_openings.7z

(or download from my website, "Downloads & Links"-section)

Enjoy!

Very interesting work, which make you the pioneer to start testing top engines with Non Castling Chess, I truly believe that Komodo Dragon 2.5 will excel in this type of chess, Just like in FRC or Chess960. I hope that CCRL, and CEGT start testing very soon.

What I would like to hear is the opinion of Mr. Kaufman on whether Top GM's can perform better against top engines with Non Castling chess then against FRC or Chess 960. I know that in either case either Magnus Carlsen would be crush, but I believe that in Non castling chess top GM's can perform a little bit better, since they can adjust faster by having only one single starting position instead of 960 unfamiliar positions.

Note: Reading and analyzing different games I have seen many many games where top players like Capablanca and Lasker in his chess manual book and so many others GM's they decided not to castle and beat their opponents, of course the position force them to do it, such as when they already made weakening pawn moves like moving their pawn in front of the side that they were planning to castle, and when the opponent already started launching a pawn storm in that side of the board, another reason not to castle is the attack that they initiated is so strong that delaying the attack by castling can give their opponent a chance to bring helping pieces to defend the attack in progress. also there are a lot of situation where your King is safer by staying in the center.

[pohl4711]

I created 2 openings sets with No Castling Chess openings for engine-testing:

Classical human openings, balanced endpositions. Queens still on the board.
All lines filtered out of MegaBase 2020. No double endpositions.
KomodoDragon 1.0 analyzed each endposition on a QuadcorePC with 15 seconds thinking-time.
Allowed Eval [-0.39;+0.59]
Files are sorted by Elo:
Both players 2600+ Elo followed by both players 2500+ Elo followed by both players 2400+ Elo

Then these 2 TAGs were added, which makes all 4 castlings illegal:
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - 0 1"]

All games containing (now) illegal castling-moves were deleted (around 25% of all games)

2 different openings sets (as PGN- and EPD-file):
NCC_6mvs_2400+_30k = 29990 different endpositions. Both players 2400 Elo or better
NCC_6mvs_2600+_6k = 6743 different endpositions. Both players 2600 Elo or better



Personally, I doubt, that this make much sense and give low draw-rates, but if anybody wants to try it out, here it is:

https://www.sp-cc.de/files/nocastlingchess_openings.7z

(or download from my website, "Downloads & Links"-section)

Enjoy!

Very interesting work, which make you the pioneer to start testing top engines with Non Castling Chess, I truly believe that Komodo Dragon 2.5 will excel in this type of chess, Just like in FRC or Chess960. I hope that CCRL, and CEGT start testing very soon.

I added a 1000 games testrun to the download of the NCC openings. The result was as bad as I predicted:

10'+5'' singlethread, (22 Threads of 24 used). No ponder, no bases (3-5 Syzygy for GUI)
1000 games each testrun. Cutechess-cli. AMD Ryzen 12core CPU.
Stockfish 12 avx2 vs. KomodoDragon 1.0 avx2. Contempt=0.
One run took about 20 hours. The Elo-number is calculated out of the score of Stockfish 12 in the head-to-head vs.
KomodoDragon 1.0 by ORDO.

Code: Select all

Classical openings-sets (for comparsion):

Balsa 2724:                             1000 (+ 67,=913,- 20), 52.4% +16 Elo, Draws 91.3%
GM_6mvs_balanced:                       1000 (+ 87,=896,- 17), 53.5% +24 Elo, Draws 89.6%
Chad 6mvs:                              1000 (+108,=859,- 33), 53.8% +26 Elo, Draws 85.9%
Hert500:                                1000 (+112,=857,- 31), 54.0% +28 Elo, Draws 85.7%
Feobos c3:                              1000 (+114,=852,- 34), 54.0% +28 Elo, Draws 85.2%
Stockfish 8mvs v3:                      1000 (+137,=817,- 46), 54.5% +32 Elo, Draws 81.7%

Average result of classical openings: 
Score (SF 12 point of view):                                   53.7% +26 Elo, Draws 86.6%

Here the result of the No Castling Chess openings testrun:
NoCastlingChess_openings_6mvs_2600+:    1000 (+122,=818,- 60), 53.1% +22 Elo, Draws 81.8%

As you can see, the NoCastlingChess (NCC) openings have a disappointing Elo-spreading and a draw-rate, which is only a little bit lower, than the classical openings. As I predicted, the NCC-concept of V.Kramnik makes no sense for using it in computerchess...

For comparsion here the average result of my AntiDraw openings-tests. Which make sense in computerchess...
Pohl AntiDraw Openings: Average result of AntiDraw openings: 58.5% +60 Elo, Draws 49.0%

I will keep the No Castling Chess openings download on my website for some time, but I can not recommend to use these openings! I strongly recommend to use my AntiDraw openings instead. Each of them is much, much better for computerchess:
https://www.sp-cc.de/anti-draw-openings.htm

[Chessqueen]

I created 2 openings sets with No Castling Chess openings for engine-testing:

Classical human openings, balanced endpositions. Queens still on the board.
All lines filtered out of MegaBase 2020. No double endpositions.
KomodoDragon 1.0 analyzed each endposition on a QuadcorePC with 15 seconds thinking-time.
Allowed Eval [-0.39;+0.59]
Files are sorted by Elo:
Both players 2600+ Elo followed by both players 2500+ Elo followed by both players 2400+ Elo

Then these 2 TAGs were added, which makes all 4 castlings illegal:
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - 0 1"]

All games containing (now) illegal castling-moves were deleted (around 25% of all games)

2 different openings sets (as PGN- and EPD-file):
NCC_6mvs_2400+_30k = 29990 different endpositions. Both players 2400 Elo or better
NCC_6mvs_2600+_6k = 6743 different endpositions. Both players 2600 Elo or better



Personally, I doubt, that this make much sense and give low draw-rates, but if anybody wants to try it out, here it is:

https://www.sp-cc.de/files/nocastlingchess_openings.7z

(or download from my website, "Downloads & Links"-section)

Enjoy!

Very interesting work, which make you the pioneer to start testing top engines with Non Castling Chess, I truly believe that Komodo Dragon 2.5 will excel in this type of chess, Just like in FRC or Chess960. I hope that CCRL, and CEGT start testing very soon.

I added a 1000 games testrun to the download of the NCC openings. The result was as bad as I predicted:

10'+5'' singlethread, (22 Threads of 24 used). No ponder, no bases (3-5 Syzygy for GUI)
1000 games each testrun. Cutechess-cli. AMD Ryzen 12core CPU.
Stockfish 12 avx2 vs. KomodoDragon 1.0 avx2. Contempt=0.
One run took about 20 hours. The Elo-number is calculated out of the score of Stockfish 12 in the head-to-head vs.
KomodoDragon 1.0 by ORDO.

Code: Select all

Classical openings-sets (for comparsion):

Balsa 2724:                             1000 (+ 67,=913,- 20), 52.4% +16 Elo, Draws 91.3%
GM_6mvs_balanced:                       1000 (+ 87,=896,- 17), 53.5% +24 Elo, Draws 89.6%
Chad 6mvs:                              1000 (+108,=859,- 33), 53.8% +26 Elo, Draws 85.9%
Hert500:                                1000 (+112,=857,- 31), 54.0% +28 Elo, Draws 85.7%
Feobos c3:                              1000 (+114,=852,- 34), 54.0% +28 Elo, Draws 85.2%
Stockfish 8mvs v3:                      1000 (+137,=817,- 46), 54.5% +32 Elo, Draws 81.7%

Average result of classical openings: 
Score (SF 12 point of view):                                   53.7% +26 Elo, Draws 86.6%

Here the result of the No Castling Chess openings testrun:
NoCastlingChess_openings_6mvs_2600+:    1000 (+122,=818,- 60), 53.1% +22 Elo, Draws 81.8%

As you can see, the NoCastlingChess (NCC) openings have a disappointing Elo-spreading and a draw-rate, which is only a little bit lower, than the classical openings. As I predicted, the NCC-concept of V.Kramnik makes no sense for using it in computerchess...

For comparsion here the average result of my AntiDraw openings-tests. Which make sense in computerchess...
Pohl AntiDraw Openings: Average result of AntiDraw openings: 58.5% +60 Elo, Draws 49.0%

I will keep the No Castling Chess openings download on my website for some time, but I can not recommend to use these openings! I strongly recommend to use my AntiDraw openings instead. Each of them is much, much better for computerchess:
https://www.sp-cc.de/anti-draw-openings.htm

We can not mix Non castling with standard chess where castling is allowed, since your AntiDraw Opening for standard chess will not be of any good if we use it in Non Castling chess, therefore, the draw rate of your Antidraw Opening become unsignificant when playing Noncastling Chess and vice versa.

[Chessqueen]

I created 2 openings sets with No Castling Chess openings for engine-testing:

Classical human openings, balanced endpositions. Queens still on the board.
All lines filtered out of MegaBase 2020. No double endpositions.
KomodoDragon 1.0 analyzed each endposition on a QuadcorePC with 15 seconds thinking-time.
Allowed Eval [-0.39;+0.59]
Files are sorted by Elo:
Both players 2600+ Elo followed by both players 2500+ Elo followed by both players 2400+ Elo

Then these 2 TAGs were added, which makes all 4 castlings illegal:
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - 0 1"]

All games containing (now) illegal castling-moves were deleted (around 25% of all games)

2 different openings sets (as PGN- and EPD-file):
NCC_6mvs_2400+_30k = 29990 different endpositions. Both players 2400 Elo or better
NCC_6mvs_2600+_6k = 6743 different endpositions. Both players 2600 Elo or better



Personally, I doubt, that this make much sense and give low draw-rates, but if anybody wants to try it out, here it is:

https://www.sp-cc.de/files/nocastlingchess_openings.7z

(or download from my website, "Downloads & Links"-section)

Enjoy!

Very interesting work, which make you the pioneer to start testing top engines with Non Castling Chess, I truly believe that Komodo Dragon 2.5 will excel in this type of chess, Just like in FRC or Chess960. I hope that CCRL, and CEGT start testing very soon.

I added a 1000 games testrun to the download of the NCC openings. The result was as bad as I predicted:

10'+5'' singlethread, (22 Threads of 24 used). No ponder, no bases (3-5 Syzygy for GUI)
1000 games each testrun. Cutechess-cli. AMD Ryzen 12core CPU.
Stockfish 12 avx2 vs. KomodoDragon 1.0 avx2. Contempt=0.
One run took about 20 hours. The Elo-number is calculated out of the score of Stockfish 12 in the head-to-head vs.
KomodoDragon 1.0 by ORDO.

Code: Select all

Classical openings-sets (for comparsion):

Balsa 2724:                             1000 (+ 67,=913,- 20), 52.4% +16 Elo, Draws 91.3%
GM_6mvs_balanced:                       1000 (+ 87,=896,- 17), 53.5% +24 Elo, Draws 89.6%
Chad 6mvs:                              1000 (+108,=859,- 33), 53.8% +26 Elo, Draws 85.9%
Hert500:                                1000 (+112,=857,- 31), 54.0% +28 Elo, Draws 85.7%
Feobos c3:                              1000 (+114,=852,- 34), 54.0% +28 Elo, Draws 85.2%
Stockfish 8mvs v3:                      1000 (+137,=817,- 46), 54.5% +32 Elo, Draws 81.7%

Average result of classical openings: 
Score (SF 12 point of view):                                   53.7% +26 Elo, Draws 86.6%

Here the result of the No Castling Chess openings testrun:
NoCastlingChess_openings_6mvs_2600+:    1000 (+122,=818,- 60), 53.1% +22 Elo, Draws 81.8%

As you can see, the NoCastlingChess (NCC) openings have a disappointing Elo-spreading and a draw-rate, which is only a little bit lower, than the classical openings. As I predicted, the NCC-concept of V.Kramnik makes no sense for using it in computerchess...

For comparsion here the average result of my AntiDraw openings-tests. Which make sense in computerchess...
Pohl AntiDraw Openings: Average result of AntiDraw openings: 58.5% +60 Elo, Draws 49.0%

I will keep the No Castling Chess openings download on my website for some time, but I can not recommend to use these openings! I strongly recommend to use my AntiDraw openings instead. Each of them is much, much better for computerchess:
https://www.sp-cc.de/anti-draw-openings.htm

The real question would be if the Top engines would perform worse with your NonCastling Openings then without it since top engines can create openings much better than any top Human GMs, specially humans rated players with 2600+ Elo