That's a nice result. I counted 13 solutions found by 11.75 but not 12.25, so 64 of the found solutions were the same and 34 solutions were found by 12.25 but not 11.75.
Going through the results, I found one position that only Ethereal 11.75 had solved among the tested engines.
One qualm I have with this sort of test suite is that it focuses on "only move" positions with a hard only move that wins or saves a draw. Positions with a superior move that doesn't outright win but gives better chances are critical in games but are absent from "hard positions" tests. Nevermind positions with multiple playable moves. Engines like Houdini 1.5 and Critter are good at finding some hard tactics, but they have much more trouble at playing well in quieter positions.
Hard-Talkchess-2020 set, final release
Moderators: hgm, Rebel, chrisw
-
- Posts: 550
- Joined: Tue Nov 19, 2019 8:48 pm
- Full name: Alayan Feh
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: Hard-Talkchess-2020 set, final release
Yes, it was pure luck.
I plan to make a criticism about a lot of the positions.
Yes, the positional play is important but tactical shots are also important.Alayan wrote: ↑Wed Jun 17, 2020 9:24 pm One qualm I have with this sort of test suite is that it focuses on "only move" positions with a hard only move that wins or saves a draw. Positions with a superior move that doesn't outright win but gives better chances are critical in games but are absent from "hard positions" tests. Nevermind positions with multiple playable moves. Engines like Houdini 1.5 and Critter are good at finding some hard tactics, but they have much more trouble at playing well in quieter positions.
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: Hard-Talkchess-2020 set, final release
2 new results :
Stockfish_20-06-27-10 : 153 sol (1252 sec)
Black-Diamond-XI-r2 : 135 sol (1419 sec)
Very good for Stockfish_20-06-27-10, if finds 12 more solutions the SF 11 !!!
Not so good for Black-Diamond-XI-r2, if finds fewer solutions than Black-Diamond-XI and Black-Diamond-XR7.
Results sheet updated : http://home.scarlet.be/vincentlejeune/c ... 2020v2.ods
Stockfish_20-06-27-10 : 153 sol (1252 sec)
Black-Diamond-XI-r2 : 135 sol (1419 sec)
Very good for Stockfish_20-06-27-10, if finds 12 more solutions the SF 11 !!!
Not so good for Black-Diamond-XI-r2, if finds fewer solutions than Black-Diamond-XI and Black-Diamond-XR7.
Results sheet updated : http://home.scarlet.be/vincentlejeune/c ... 2020v2.ods
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: Hard-Talkchess-2020 set, final release
2 more results :
ShashChess 12.0 : 111 solutions (1842 seconds)
Not a lot of solutions but hold some records now. This engine has surely something special.
Another curious thing : the search changes when I restart with the same position. There should be random things in this engine.
I ask ed the question in the ShashChess thread : http://talkchess.com/forum3/viewtopic.p ... 37#p849637
Crystal-2020-06-29 : 159 solutions (1061 seconds)
Very good result for Crystal : +13 solutions over Crystal 2020-05-02 !
It's the best engine so far if we except "Bluefish-FD-XI-Tact=2-defens=no" (who use multi threads) and "Eman 5.40, FAD=12" (who use full depth for first plies)
I downloaded the engine from here : https://chess.ultimaiq.net/stockfish.html
Results sheet updated : http://home.scarlet.be/vincentlejeune/c ... 2020v2.ods
ShashChess 12.0 : 111 solutions (1842 seconds)
Not a lot of solutions but hold some records now. This engine has surely something special.
Another curious thing : the search changes when I restart with the same position. There should be random things in this engine.
I ask ed the question in the ShashChess thread : http://talkchess.com/forum3/viewtopic.p ... 37#p849637
Crystal-2020-06-29 : 159 solutions (1061 seconds)
Very good result for Crystal : +13 solutions over Crystal 2020-05-02 !
It's the best engine so far if we except "Bluefish-FD-XI-Tact=2-defens=no" (who use multi threads) and "Eman 5.40, FAD=12" (who use full depth for first plies)
I downloaded the engine from here : https://chess.ultimaiq.net/stockfish.html
Results sheet updated : http://home.scarlet.be/vincentlejeune/c ... 2020v2.ods
-
- Posts: 12541
- Joined: Wed Mar 08, 2006 8:57 pm
- Location: Redmond, WA USA
Re: Hard-Talkchess-2020 set, final release
I wonder if the FAD = N could be applied to Bluefish to make a super tactical solver (that will probably stink up the joint when playing games)
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: Hard-Talkchess-2020 set, final release
I updated a lot of variations to explain all the "Hard-Talkchess-2020" points in these Studies.Vinvin wrote: ↑Fri Feb 07, 2020 1:33 pm I imported all the positions in Lichess Studies (but not all the mainlines yet) :
#1 to #60 : https://lichess.org/study/y6sM5eAf
#61 to #120 : https://lichess.org/study/Vfe59GOH
#121 to #180 : https://lichess.org/study/0FQa6riN
#181 to #213 : https://lichess.org/study/s2mTErEZ
Please tell me if you wish to add some variations.
-
- Posts: 1167
- Joined: Thu Dec 25, 2008 9:07 pm
- Full name: Herbert L
Re: Hard-Talkchess-2020 set, final release
Great, thanks for your work.
-
- Posts: 450
- Joined: Wed Nov 24, 2010 10:57 am
- Location: INDIA
Re: Hard-Talkchess-2020 set, final release
Stockfish_20-06-27-10 <---- Where to get this ?Vinvin wrote: ↑Mon Jun 29, 2020 5:02 am 2 new results :
Stockfish_20-06-27-10 : 153 sol (1252 sec)
Black-Diamond-XI-r2 : 135 sol (1419 sec)
Very good for Stockfish_20-06-27-10, if finds 12 more solutions the SF 11 !!!
Not so good for Black-Diamond-XI-r2, if finds fewer solutions than Black-Diamond-XI and Black-Diamond-XR7.
Results sheet updated : http://home.scarlet.be/vincentlejeune/c ... 2020v2.ods
Always Expect the Unexpected
-
- Posts: 6442
- Joined: Tue Jan 09, 2007 12:31 am
- Location: PA USA
- Full name: Louis Zulli
Re: Hard-Talkchess-2020 set, final release
I assume it's simply the June 27 version of Stockfish-dev. You can get it at https://abrok.eu/stockfish/Master Om wrote: ↑Mon Jul 06, 2020 5:18 pmStockfish_20-06-27-10 <---- Where to get this ?Vinvin wrote: ↑Mon Jun 29, 2020 5:02 am 2 new results :
Stockfish_20-06-27-10 : 153 sol (1252 sec)
Black-Diamond-XI-r2 : 135 sol (1419 sec)
Very good for Stockfish_20-06-27-10, if finds 12 more solutions the SF 11 !!!
Not so good for Black-Diamond-XI-r2, if finds fewer solutions than Black-Diamond-XI and Black-Diamond-XR7.
Results sheet updated : http://home.scarlet.be/vincentlejeune/c ... 2020v2.ods
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: Hard-Talkchess-2020 set, final release
Yes, this one is here :zullil wrote: ↑Mon Jul 06, 2020 5:24 pmI assume it's simply the June 27 version of Stockfish-dev. You can get it at https://abrok.eu/stockfish/Master Om wrote: ↑Mon Jul 06, 2020 5:18 pmStockfish_20-06-27-10 <---- Where to get this ?Vinvin wrote: ↑Mon Jun 29, 2020 5:02 am 2 new results :
Stockfish_20-06-27-10 : 153 sol (1252 sec)
Black-Diamond-XI-r2 : 135 sol (1419 sec)
Very good for Stockfish_20-06-27-10, if finds 12 more solutions the SF 11 !!!
Not so good for Black-Diamond-XI-r2, if finds fewer solutions than Black-Diamond-XI and Black-Diamond-XR7.
Results sheet updated : http://home.scarlet.be/vincentlejeune/c ... 2020v2.ods
https://abrok.eu/stockfish/builds/ca41e ... modern.exe