SPCC: Unbalanced Human Openings V3.0 and remastered Openings released

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

User avatar
pohl4711
Posts: 2435
Joined: Sat Sep 03, 2011 7:25 am
Location: Berlin, Germany
Full name: Stefan Pohl

SPCC: Unbalanced Human Openings V3.0 and remastered Openings released

Post by pohl4711 »

Because I can not say, when my website will go online again, I release my new Anti-Draw Openings right now here. You can download them directly from my Google Drive:



Information out of one ReadMe-file in the download:

This openings-sets collection contains 4 different openings concepts:
Drawkiller V4, No Black Short Castling V2, No Black Castling V2 and Unbalanced Human Openings V3. Additionally, the download contains the tools, which are needed for the Armageddon- and my Advanced Armageddon-rescoring of results.

Drawkiller, NBSC and NBC were remastered, which means, all endpositions were re-evaluated with KomodoDragon 1.0 (15 seconds/ position on a Quadcore PC) and the openings-sets were complete rebuilt. Additionally, Drawkiller openings now contain some unbalanced opening-sets, too (inspired by the Unbalanced Human Openings concept). Read more in the ReadMe-file in the Drawkiller-folder.

For UHO (Unbalanced Human Openings) V3 all endpositions of the raw-data of UHO V2 with an eval of at least +0.60 for white were re-evaluated on a 12core AMD Ryzen PC with 30 seconds per endposition, which means, that the eval of the UHO V3 endpositions is based on (around) 10x more calculated nodes than in UHO V2. That gives a search depth around 28-30 plies in UHO V3, which is around 4-5 plies deeper than in UHO V2 (!!!). Additionally, the UHO V3 opening-sets now contain not only one, but 7 UHO-openings sets with increasing advantage for white (from +0.90 up to +1.50). Read more in the ReadMe-file in the UHO-folder.

I believe, this opening collection with 4 different opening concepts is giving the perfect solution for all purposes in computerchess: Unbalanced Human Openings are 100% human chess moves - nothing else. Drawkiller is 100% constructed, non-human, for the best statistical results (take a look at the Drawkiller unbalanced test-results - just mind-blowing!). NBSC/NBC openings are in between: Half constructed and half human (a move-line, which forbids black castlings, is added to human played openings). All these openings are classical chess, no chess variant (like Chess960), because all lines start from the classical chesss starting position. So, all these openings can be used by all chessengines and chessGUIs on the planet!

Here the testing results with long thinking-time and comparsion with 6 classical openings-sets. I think, the results speak for themselves...But take a look at the results of my new Drawkiller unbalanced openings: Just mind-blowing!

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%


Pohl AntiDraw Openings:

UHO_V3_6mvs_+90_+99          : 1000 (+290,=546,-164), 56.3% +44 Elo,  Draws 54.6%
UHO_V3_8mvs_+90_+99          : 1000 (+286,=583,-131), 57.8% +54 Elo,  Draws 58.3%
UHO_V3_8mvs_big_+80_+109     : 1000 (+295,=585,-120), 58.8% +62 Elo,  Draws 58.5%

NBSC_Noomen_3mvs500          : 500  (+154,=242,-104), 55.0% +36 Elo,  Draws 48.4%
NBSC_V2_4mvs                 : 1000 (+302,=541,-157), 57.3% +52 Elo,  Draws 54.1%
NBSC_V2_6mvs                 : 1000 (+275,=564,-161), 55.7% +40 Elo,  Draws 56.4%

NBC_Noomen_3mvs500           : 500  (+170,=198,-132), 53.8% +26 Elo,  Draws 39.6%
NBC_V2_4mvs                  : 1000 (+313,=496,-191), 56.1% +42 Elo,  Draws 49.6%
NBC_V2_6mvs                  : 1000 (+323,=468,-209), 55.7% +40 Elo,  Draws 46.8%

Drawkiller_balanced_V4_2mvs  : 1000 (+376,=574,- 50), 66.3% +118 Elo, Draws 57.4%
Drawkiller_balanced_V4_3mvs  : 1000 (+324,=630,- 46), 63.9% +100 Elo, Draws 63.0%
Drawkiller_balanced_V4_4mvs  : 1000 (+329,=620,- 51), 63.9% +100 Elo, Draws 62.0%

Drawkiller_unbalanced_V4_2mvs: 1000 (+439,=407,-154), 64.3% +102 Elo, Draws 40.7%
Drawkiller_unbalanced_V4_3mvs: 1000 (+407,=461,-132), 63.8% +98 Elo,  Draws 46.1%
Drawkiller_unbalanced_V4_4mvs: 1000 (+383,=480,-137), 62.3% +88 Elo,  Draws 48.0%


Advanced Armageddon rescoring of opening-sets with clear white advantage:

UHO_V3_6mvs_+90_+99          : 1012 (+627,=  0,-385), 62.0% +86 Elo,  Draws 0%
UHO_V3_8mvs_+90_+99          : 1022 (+658,=  0,-364), 64.4% +104 Elo, Draws 0%
UHO_V3_8mvs_big_+80_+109     : 1015 (+675,=  0,-340), 66.5% +120 Elo, Draws 0%

NBSC_Noomen_3mvs500          : 506  (+302,=  0,-204), 59.7% +68 Elo,  Draws 0%
NBSC_V2_4mvs                 : 1020 (+646,=  0,-374), 63.3% +96 Elo,  Draws 0%
NBSC_V2_6mvs                 : 1025 (+616,=  0,-409), 60.1% +72 Elo,  Draws 0%

NBC_Noomen_3mvs500           : 503  (+288,=  0,-215), 57.3% +52 Elo,  Draws 0%
NBC_V2_4mvs                  : 1018 (+622,=  0,-396), 61.1% +80 Elo,  Draws 0%
NBC_V2_6mvs                  : 1023 (+616,=  0,-407), 60.2% +72 Elo,  Draws 0%

Drawkiller_unbalanced_V4_2mvs: 1050 (+787,=  0,-263), 75.0% +192 Elo, Draws 0%
Drawkiller_unbalanced_V4_3mvs: 1056 (+781,=  0,-275), 74.0% +182 Elo, Draws 0%
Drawkiller_unbalanced_V4_4mvs: 1035 (+749,=  0,-286), 72.4% +168 Elo, Draws 0%
jefk
Posts: 626
Joined: Sun Jul 25, 2010 10:07 pm
Location: the Netherlands
Full name: Jef Kaan

Re: SPCC: Unbalanced Human Openings V3.0 and remastered Openings released

Post by jefk »

nice, ik like the 1.Na3! Nh6!! opening, (anti-Durkin defense)
but...you haven't included the famous 'Bongcloud' ,
invented by Nakamura, and now already an awesome
weapon to avoid any draws (see here some highlevel
analysis (i like the version where the king goes to b5)
Alayan
Posts: 550
Joined: Tue Nov 19, 2019 8:48 pm
Full name: Alayan Feh

Re: SPCC: Unbalanced Human Openings V3.0 and remastered Openings released

Post by Alayan »

Games are being played with Drawkiller v4 at CCC currently.
User avatar
pohl4711
Posts: 2435
Joined: Sat Sep 03, 2011 7:25 am
Location: Berlin, Germany
Full name: Stefan Pohl

Re: SPCC: Unbalanced Human Openings V3.0 and remastered Openings released

Post by pohl4711 »

Alayan wrote: Fri Apr 02, 2021 9:20 pm Games are being played with Drawkiller v4 at CCC currently.
Cool.

Always good to see, that my openings are in use and my work is appreciated.
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: SPCC: Unbalanced Human Openings V3.0 and remastered Openings released

Post by MikeB »

pohl4711 wrote: Sat Apr 03, 2021 3:32 am
Alayan wrote: Fri Apr 02, 2021 9:20 pm Games are being played with Drawkiller v4 at CCC currently.
Cool.

Always good to see, that my openings are in use and my work is appreciated.
Well ... you do nice work! Thanks.
Image
Modern Times
Posts: 3548
Joined: Thu Jun 07, 2012 11:02 pm

Re: SPCC: Unbalanced Human Openings V3.0 and remastered Openings released

Post by Modern Times »

pohl4711 wrote: Sat Apr 03, 2021 3:32 am Cool.

Always good to see, that my openings are in use and my work is appreciated.
Your books are also used quite a lot in CCRL blitz testing - IM4moves and SuperGM4moves in particular, and I've been experimenting with the 5mvs_balanced book as well.
User avatar
Scally
Posts: 232
Joined: Thu Sep 28, 2017 9:34 pm
Location: Bermondsey, London
Full name: Alan Cooper

Re: SPCC: Unbalanced Human Openings V3.0 and remastered Openings released

Post by Scally »

Very nice ….

Thanks,

Al.
User avatar
pohl4711
Posts: 2435
Joined: Sat Sep 03, 2011 7:25 am
Location: Berlin, Germany
Full name: Stefan Pohl

Re: SPCC: Unbalanced Human Openings V3.0 and remastered Openings released

Post by pohl4711 »

Alayan wrote: Fri Apr 02, 2021 9:20 pm Games are being played with Drawkiller v4 at CCC currently.
Now, there is a match running at CCC between Lc0 and Komodo Dragon, using my NBSC openings (Black is not allowed to castle short) (NBSC_Noomen_3mvs - same set, which I use for my NN-testruns of Lc0 vs.SF).

After 25 of 100 games, the draw-rate is 56%. On the fast CCC-hardware, this is a very nice and low draw-rate. Thats, what I want to see...