ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

Ghppn
Posts: 31
Joined: Wed Aug 11, 2021 5:15 pm
Full name: Gheorghe Pepene

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by Ghppn »

Kayra 1.3 bmi2 - ProteusSF JBE 007_bmi2 / Blitz 5m+2s
Image
Games pgn: https://mega.nz/file/asZmDDJQ#yDLcI_OGR ... 3FZGYs2L_Q
Openings: TCEC 21_Noomen Select
Processor: Intel ® Core ™ i7– 4790 CPU 3.60 GHz (4 Cores, 8 Threads)
Memoriy (RM): 32 GB, System type: 64-bit, Windows 10 PRO.
Graphic board: NVIDIA Geforce GTX 760
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by AlexChess »

cc2150dx wrote: Tue Apr 19, 2022 7:45 pm You were close on the link. But it's actually here: http://blogchess2016.blogspot.com/2021/ ... lable.html for future reference.
Thank you!
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by AlexChess »

Ghppn wrote: Tue Apr 19, 2022 8:13 pm Kayra 1.3 bmi2 - ProteusSF JBE 007_bmi2 / Blitz 5m+2s
Image
Games pgn: https://mega.nz/file/asZmDDJQ#yDLcI_OGR ... 3FZGYs2L_Q
Openings: TCEC 21_Noomen Select
Processor: Intel ® Core ™ i7– 4790 CPU 3.60 GHz (4 Cores, 8 Threads)
Memoriy (RM): 32 GB, System type: 64-bit, Windows 10 PRO.
Graphic board: NVIDIA Geforce GTX 760
Hi George!

Looking for the truth (I'm searching to get real and fair results but it's really difficult :) ) I've started using first sequentials 150 Pohl's UHO_2022_8mvs_big_+100_+129 and increased the number of games to 300 for each match (TC 3 +2) It would be interesting if you too could repeat the match against Kayra 1.3 (that I like a lot!)

First good results:

Image

https://www.sp-cc.de/uho_2022.htm

Whats new in UHO 2022:
The UHO 2022 opening sets, are filtered out of the Megabase 2022 (by ChessBase). And the only filter, I am using, is the eval of KomodoDragon 2.6 in the endposition of each opening-line. (UHO V3 used Megabase 2020 and KomodoDragon 1.0 for evaluating). And the UHO 2022 files are around +10% bigger than the UHO V3 files, because the Megabase 2022 contains more games than the Megabase 2020, of course.



What did I do exactly?

- sorted the games of the Megabase 2022 by Elo: both players 2700+ Elo, both players 2600+ Elo, both players 2500+ Elo, both players 2400+ Elo, both players 2300+ Elo.
- cutted all games after 6/8 moves (12/16 plies).
- removed all Chess960 games
- removed all lines with double endpositons, so each endposition is unique. And because the games are sorted by Elo of players, each opening line in the UHO-sets is played by the strongest players (if weaker players played the same opening line, this game was deleted by pgn-extract, because pgn-extract always keeps the first appearance of a line in a pgn-file)
- removed all lines, were not both queens are still on board in the endposition.
- removed all lines, starting with silly pawn-moves in move 1 and 2 (a4, b4, g4, h4 (same for black (...a5, ...b5, etc.)

After doing these work, there were 111789 6-moves deep opening-lines remaining and 316318 8-moves deep opening lines remaining (all with unique endpositions).
Finally, these opening-lines were evaluated (endpositions) with KomodoDragon 2.6 (8.5 seconds per endposition on a 12 core AMD Ryzen CPU) and kept all endpositions in an eval interval of [-2.99;+2.99] (this is the UHO 2022 RawData (can be found in the download, too)). That took around 40 days of computing...
The UHO 2022 RawData can be very useful for other purposes, because it contains all opening-lines from very good for black to very good for white. Feel free to filter these openings as you like...The eval (and search depth) is stored in the Annotator-Tag of each line:
[Annotator "depth=29 eval=+114"]
You can use ChessBase (for example) for searching. Examples of searching:
"eval=-05" finds all lines in an eval-interval of [-0.50;-0.59]
"eval=+00" finds all lines in an eval-interval of [+0.00;+0.09]
"eval=-1" finds all lines in an eval-interval of [-1.00;-1.99]


So the Unbalanced Human Openings contain:

- 100% moves played by humans, only. Both players had at least 2300 Elo.
- opening lines played by the stronges players are at the beginning of the pgn/epd-files.
- no manually constructed openings (like my Drawkiller openings)
- no manually added moves to make castling impossible (like my NBSC openings)
- no selection of piece-patterns
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by AlexChess »

Ghppn wrote: Tue Apr 19, 2022 8:13 pm Kayra 1.3 bmi2 - ProteusSF JBE 007_bmi2 / Blitz 5m+2s
Image
Games pgn: https://mega.nz/file/asZmDDJQ#yDLcI_OGR ... 3FZGYs2L_Q
Openings: TCEC 21_Noomen Select
Processor: Intel ® Core ™ i7– 4790 CPU 3.60 GHz (4 Cores, 8 Threads)
Memoriy (RM): 32 GB, System type: 64-bit, Windows 10 PRO.
Graphic board: NVIDIA Geforce GTX 760
Hi Gheorghe!

Looking for the truth (I'm searching to get real and fair results but it's really difficult :) ) I've started using sequentially first 150 Pohl's UHO_2022_8mvs_big_+100_+129 and increased the number of games to 300 for each match (TC 3 mins + 2 secs Tablebases 6 man 8 CPUs 6.5 Mn/s on Mac M1) It would be interesting if you too could repeat the match against Kayra 1.3 (that I like a lot!) .. until now +0 merging our results. On the bottom you find the benchmarks of our 2 systems: mine is 2 times faster, but your TC is double, so the comparative is almost perfect :D

First good not complete new results:

Image

https://www.sp-cc.de/uho_2022.htm

Whats new in UHO 2022:
The UHO 2022 opening sets, are filtered out of the Megabase 2022 (by ChessBase). And the only filter, I am using, is the eval of KomodoDragon 2.6 in the endposition of each opening-line. (UHO V3 used Megabase 2020 and KomodoDragon 1.0 for evaluating). And the UHO 2022 files are around +10% bigger than the UHO V3 files, because the Megabase 2022 contains more games than the Megabase 2020, of course.



What did I do exactly?

- sorted the games of the Megabase 2022 by Elo: both players 2700+ Elo, both players 2600+ Elo, both players 2500+ Elo, both players 2400+ Elo, both players 2300+ Elo.
- cutted all games after 6/8 moves (12/16 plies).
- removed all Chess960 games
- removed all lines with double endpositons, so each endposition is unique. And because the games are sorted by Elo of players, each opening line in the UHO-sets is played by the strongest players (if weaker players played the same opening line, this game was deleted by pgn-extract, because pgn-extract always keeps the first appearance of a line in a pgn-file)
- removed all lines, were not both queens are still on board in the endposition.
- removed all lines, starting with silly pawn-moves in move 1 and 2 (a4, b4, g4, h4 (same for black (...a5, ...b5, etc.)

After doing these work, there were 111789 6-moves deep opening-lines remaining and 316318 8-moves deep opening lines remaining (all with unique endpositions).
Finally, these opening-lines were evaluated (endpositions) with KomodoDragon 2.6 (8.5 seconds per endposition on a 12 core AMD Ryzen CPU) and kept all endpositions in an eval interval of [-2.99;+2.99] (this is the UHO 2022 RawData (can be found in the download, too)). That took around 40 days of computing...
The UHO 2022 RawData can be very useful for other purposes, because it contains all opening-lines from very good for black to very good for white. Feel free to filter these openings as you like...The eval (and search depth) is stored in the Annotator-Tag of each line:
[Annotator "depth=29 eval=+114"]
You can use ChessBase (for example) for searching. Examples of searching:
"eval=-05" finds all lines in an eval-interval of [-0.50;-0.59]
"eval=+00" finds all lines in an eval-interval of [+0.00;+0.09]
"eval=-1" finds all lines in an eval-interval of [-1.00;-1.99]


So the Unbalanced Human Openings contain:

- 100% moves played by humans, only. Both players had at least 2300 Elo.
- opening lines played by the stronges players are at the beginning of the pgn/epd-files.
- no manually constructed openings (like my Drawkiller openings)
- no manually added moves to make castling impossible (like my NBSC openings)
- no selection of piece-patterns

Image

Image
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by AlexChess »

Image

ALL 300 PGN GAMES: https://pixeldrain.com/u/BhdvSoxG

OPENINGS BOOK DETAILS: https://www.sp-cc.de/uho_2022.htm
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by AlexChess »

AlexChess wrote: Fri Apr 22, 2022 4:41 pm
ALL 300 PGN GAMES: https://pixeldrain.com/u/BhdvSoxG

OPENINGS BOOK DETAILS: https://www.sp-cc.de/uho_2022.htm
Image
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by AlexChess »

Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by AlexChess »

Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by AlexChess »

ProteusSF RBE 008 is out! https://banksiagui.com/forums/viewtopic.php?p=132#p132

Kind regards, Alex
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
Ghppn
Posts: 31
Joined: Wed Aug 11, 2021 5:15 pm
Full name: Gheorghe Pepene

Re: ProteusChessSF - Developing my Stockfish 14.1 Polyglot MTCS derivative

Post by Ghppn »

Kayra 1.4 bmi2 - SwordF PB 15.2 bmi2 / Blitz 5m+2s
Image
https://mega.nz/file/r8Bw3ABQ#ExtXbPu9L ... 8BgF7xfwsE
Openings: TCEC 22_Noomen 50 games
Processor: Intel ® Core ™ i7– 4790 CPU 3.60 GHz (4 Cores, 8 Threads)
Memoriy (RM): 32 GB, System type: 64-bit, Windows 10 PRO.
Graphic board: NVIDIA Geforce GTX 760