A complete 2000 lines of code engine

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

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
xr_a_y
Posts: 543
Joined: Sat Nov 25, 2017 1:28 pm
Location: France

Re: A complete 2000 lines of code engine

Post by xr_a_y » Thu Jan 10, 2019 7:48 pm

Thanks a lot for this tourney. Very good opposition !

In the meantime, Minic 0.30 is released with a very simple mobility term in eval => +45elo !
Attachments
Capture du 2019-01-10 20-47-13.png
Capture du 2019-01-10 20-47-13.png (5.22 KiB) Viewed 2209 times

User avatar
CMCanavessi
Posts: 832
Joined: Thu Dec 28, 2017 3:06 pm
Location: Argentina

Re: A complete 2000 lines of code engine

Post by CMCanavessi » Thu Jan 10, 2019 8:38 pm

Code: Select all

    Engine                    Score   Dr Fr Co Be Mi Ga Fr Or Ru TJ Ab Fr Ar Ee TC Ze Gi Be Di Ab Ga Pl Si Ja Ev Ne Gr Su If Ki    S-B
01: Drosophila 1.5 x64        30.0/41 ·· 0  0  =  1  =1 1= 1  1  1  0  1  0  11 01 1= =1 0  0  1  1  11 1  =  1= 11 1  11 11 1=  566,00
02: Francesca M.A.D. 0.19 x32 29.0/41 1  ·· =  00 =  1  1  =1 11 1  1  01 0  11 11 1  10 11 01 =1 0  01 1  1  =  0  1  1  =1 1   579,75
03: Coiled 0.4 x64            28.0/41 1  =  ·· 1  =  11 01 =  1  0  01 0  =1 1  0= 01 1  1  0  1  0  00 1  1  11 11 =  11 11 11  540,75
04: Betsabe II 1.84 x32       27.0/41 =  11 0  ·· 1= 0  =  00 10 01 01 1= 1  0  1  =  1  11 01 11 =  1  11 0= =  1  1  1  1  1   535,00
05: Minic 0.29 x64            26.0/41 0  =  =  0= ·· 0  0  01 =  1  1  00 0  11 10 1  10 11 11 11 1  =1 1  0  1  =  =  1  11 1=  484,00
06: Galjoen 0.39.2 x64        24.0/41 =0 0  00 1  1  ·· 1= 1  =  0  1= 1  01 =  0  10 0  1  0  =  10 1  0  11 10 11 11 11 =  =   464,25
07: FrankWalter 2.2.8 x64     24.0/41 0= 0  10 =  1  0= ·· =  0  == =  0  =1 =  0  0  0  =  1  0  11 1  11 11 11 11 01 == 1  1   448,25
08: Orion 0.5 x64             23.5/41 0  =0 =  11 10 0  =  ·· =0 == =1 =1 0  1  1  1  1  0  =0 10 0  1  10 01 1  1  == 1  1  1   468,25
09: RubiChess 0.8.1 x64       23.5/41 0  00 0  01 =  =  1  =1 ·· 0  == 01 0  =1 0  0  11 1= 10 00 1  11 1  1  1  1  1  1  10 1   445,50
10: TJchess 1.3 x64           23.0/41 0  0  1  10 0  1  == == 1  ·· 0  =  1  1= =0 1  11 =1 00 =0 1  10 0  0  1  1  1  1  01 1=  449,75
11: Abrok 5.0 x32             23.0/41 1  0  10 10 0  0= =  =0 == 1  ·· 0= 1  0  1  == 0  0  == =1 =  1  1  =  11 10 1  11 1  =   447,25
12: Fruit 1.0 x32             22.5/41 0  10 1  0= 11 0  1  =0 10 =  1= ·· 0  1= =  0  1= 01 =1 11 1  =  0  1  1  =  0  =  01 0   466,25
13: Ares 1.005.2.1 x64        22.5/41 1  1  =0 0  1  10 =0 1  1  0  0  1  ·· 0  01 =0 =0 =  1  1  1  =1 1  1  00 == 1  01 10 =1  441,25
14: Eeyore 1.52 x64           22.0/41 00 00 0  1  00 =  =  0  =0 0= 1  0= 1  ·· 1  0  1  1  11 01 1= 1  10 11 =  1  11 0  1  =   407,00
15: TCB 0052 x32              20.0/41 10 00 1= 0  01 1  1  0  1  =1 0  =  10 0  ·· 1  1  1  1  0  1= 1  0= 10 =0 0  01 00 0  1   415,00
16: Zevra v2.1.1 r216 x64     19.5/41 0= 0  10 =  0  01 1  0  1  0  == 1  =1 1  0  ·· 0  0  1  =  01 =  00 11 =1 01 00 10 1  =   387,50
17: Giraffe 20161023 x64      19.5/41 =0 01 0  0  01 1  1  0  00 00 1  0= =1 0  0  1  ·· 0  1  10 10 1  10 11 1  1  00 1  1  0   383,50
18: Bearded Neural v44.5 x64  18.5/41 1  00 0  00 00 0  =  1  0= =0 1  10 =  0  0  1  1  ·· 1= 01 =0 =  == 01 1  0  01 1  1  1   349,75
19: Dimitri 3.93 x32          17.5/41 1  10 1  10 00 1  0  =1 01 11 == =0 0  00 0  0  0  0= ·· 01 0  1  11 0  0  0  1  0  0  1   381,25
20: Absolute Zero 2.4.7.2 x64 17.5/41 0  =0 0  00 00 =  1  01 11 =1 =0 00 0  10 1  =  01 10 10 ·· 0  0  0  1  =  0  1  =  1  1   346,75
21: Gaia 3.5 x64              17.0/41 0  1  1  =  0  01 00 1  0  0  =  0  0  0= 0= 10 01 =1 1  1  ·· 00 =  0  =  =1 1  01 00 01  335,00
22: Plisk 0.2.7_d x64         17.0/41 00 10 11 0  =0 0  0  0  00 01 0  =  =0 0  0  =  0  =  0  1  11 ·· 10 00 10 1  11 =  1  1   326,50
23: Simplex 0.9.8 x64         16.5/41 0  0  0  00 0  1  00 01 0  1  0  1  0  01 1= 11 01 == 00 1  =  01 ·· 1  =  0  0  1  00 10  316,00
24: Jazz Orchestra 840 x64    16.0/41 =  0  0  1= 1  00 00 10 0  1  =  0  0  00 01 00 00 10 1  0  1  11 0  ·· 0  1  1  1  0= 10  311,75
25: EveAnn 1.72 x32           16.0/41 0= =  00 =  0  01 00 0  0  0  00 0  11 =  =1 =0 0  0  1  =  =  01 =  1  ·· =1 =  =1 0  ==  305,00
26: Nemeton 1.7 x32           16.0/41 00 1  00 0  =  00 00 0  0  0  01 =  == 0  1  10 0  1  1  1  =0 0  1  0  =0 ·· 10 01 1  11  292,75
27: Gromit3 3.0.0 x32         15.5/41 0  0  =  0  =  00 10 == 0  0  0  1  0  00 10 11 11 10 0  0  0  00 1  0  =  01 ·· 0  11 10  293,00
28: Sungorus 1.4 x64          15.0/41 00 0  00 0  0  00 == 0  0  0  00 =  10 1  11 01 0  0  1  =  10 =  0  0  =0 10 1  ·· 1  11  268,00
29: Ifrit m1.8 x64            13.5/41 00 =0 00 0  00 =  0  0  01 10 0  10 01 0  1  0  0  0  1  0  11 0  11 1= 1  0  00 0  ·· 0   262,50
30: Kingfisher v1.1 x64       12.0/41 0= 0  00 0  0= =  0  0  0  0= =  1  =0 =  0  =  1  0  0  0  10 0  01 01 == 00 01 00 1  ··  231,50

615 of 1740 games played
Name of the tournament: 190 - Qualy League
This new tournament is quite competitive and Minic is doing great!
Follow my tournament and some Leela gauntlets live at http://twitch.tv/ccls

User avatar
xr_a_y
Posts: 543
Joined: Sat Nov 25, 2017 1:28 pm
Location: France

Re: A complete 2000 lines of code engine

Post by xr_a_y » Fri Jan 11, 2019 6:41 pm

Lastest results :)

Code: Select all

   1 fruit_21                      278      51     250   83.2%   13.6%
   2 drosophila-win64              117      41     250   66.2%   18.8%
>  3 Minic 0.30                     55      39     250   57.8%   21.2%
>  4 Minic 0.29                     -8      37     250   48.8%   27.2%
   5 myrddin                       -13      39     250   48.2%   18.0%
   6 MadChess.Engine               -55      38     250   42.2%   23.6%
   7 Bitfoot                       -69      39     250   40.2%   20.4%
   8 sungorus64                    -78      40     250   39.0%   16.4%
>  9 Minic 0.28                    -88      39     250   37.6%   21.6%
  10 Horizon_4_4                   -94      40     250   36.8%   18.4%

User avatar
xr_a_y
Posts: 543
Joined: Sat Nov 25, 2017 1:28 pm
Location: France

Re: A complete 2000 lines of code engine

Post by xr_a_y » Sat Jan 19, 2019 10:24 am

Minic lichess bot was lucky enough to be challenged by some lichess user this week and most notably by a multi titled candidate master (https://lichess.org/@/vovaches, https://ratings.fide.com/card.phtml?event=34105414) and won the 4 games :) .

https://lichess.org/wx0UFgCE/black#0 (slow positional win and a too optimistic rook exchange from white)
https://lichess.org/jWv84yBV#0 (Minic playing white is putting too much pressure near black king)
https://lichess.org/ZLfMSpHd#0 (Minic playing white is pushing a passed pawn to the win)
https://lichess.org/5LAymnvz/black#0 (almost a draw but Vladimir blunder at the end)

User avatar
xr_a_y
Posts: 543
Joined: Sat Nov 25, 2017 1:28 pm
Location: France

Re: A complete 2000 lines of code engine

Post by xr_a_y » Sun Jan 27, 2019 9:56 am

Minic 0.34 is out. It should be a +40 versus 0.32 mainly thanks to time management stuff ; so heading to 2500 soon...
Minic 0.34 also adds some (not activated) code to prepare endgame management and syzygy use.

User avatar
xr_a_y
Posts: 543
Joined: Sat Nov 25, 2017 1:28 pm
Location: France

Re: A complete 2000 lines of code engine

Post by xr_a_y » Tue Feb 19, 2019 6:33 am

0.37 was a +35 versus 0.34
but the 0.40 is +70 versus 0.37 thanks to pruning even in end-game :o
Should be near 2570 now
Attachments
Capture du 2019-02-19 07-32-26.png
Capture du 2019-02-19 07-32-26.png (4.45 KiB) Viewed 1726 times

User avatar
xr_a_y
Posts: 543
Joined: Sat Nov 25, 2017 1:28 pm
Location: France

Re: A complete 2000 lines of code engine

Post by xr_a_y » Thu Feb 28, 2019 8:50 am

Capture.PNG
Capture.PNG (1.96 KiB) Viewed 1628 times
Strange but good surprise at 40/3 TC

Release 0.41 is available.

User avatar
xr_a_y
Posts: 543
Joined: Sat Nov 25, 2017 1:28 pm
Location: France

Re: A complete 2000 lines of code engine

Post by xr_a_y » Mon Mar 18, 2019 7:11 am

Minic probably crossed the 2600 elo line :

Code: Select all

   4 minic_dev_linux_x64             -85      29     437   38.0%   22.9%
   5 minic_0.43_linux_x64_see4.2     -97      29     438   36.4%   22.6%
   6 asymptote-v0.4.2               -118      32     437   33.6%   13.7%
   7 minic_0.40_linux_x64_see4.2    -175      31     437   26.8%   22.0%
A new 0.44 release will come soon... :D

User avatar
M ANSARI
Posts: 3395
Joined: Thu Mar 16, 2006 6:10 pm

Re: A complete 2000 lines of code engine

Post by M ANSARI » Tue Mar 19, 2019 7:30 am

I guess if anything ... this shows the importance of bug free code in a chess program. I think the first to show that was Fruit. While it is nice to have many features and many minor little tweaks to get the program to play "proper" chess ... bugs are the Achilles Heel of any chess program. Each additional improvement needs to come without letting any bugs creep in.

Good job on your program and I must say it is very interesting to follow up on the progress!

User avatar
xr_a_y
Posts: 543
Joined: Sat Nov 25, 2017 1:28 pm
Location: France

Re: A complete 2000 lines of code engine

Post by xr_a_y » Sun Mar 24, 2019 9:26 am

yes 2600 elo ;-) : http://ccrl.chessdom.com/ccrl/404/cgi/e ... _43_64-bit

Thanks a lot too CCRL members for testing.

Post Reply