Page 1 of 2
For those that like to toy with Ethereal...
Posted: Wed Oct 30, 2019 10:22 pm
by Dann Corbit
You can get a very close approximation to the Stockfish null move pruning with this formula:
Code: Select all
R = ( ( 13 + depth ) >> 2 ) + MIN(3, (eval - beta) / 185);
Re: For those that like to toy with Ethereal...
Posted: Wed Oct 30, 2019 10:27 pm
by Dann Corbit
E.g. you can experiment with it like this easily:
Code: Select all
#ifdef OLD_PRUNE
R = 4 + depth / 6 + MIN(3, (eval - beta) / 200);
#else
R = ( ( 13 + depth ) >> 2 ) + MIN(3, (eval - beta) / 185);
#endif
Re: For those that like to toy with Ethereal...
Posted: Wed Oct 30, 2019 11:40 pm
by Dann Corbit
In a test (every early) the Ethereal pruning is doing better than the Stockfish style.
Re: For those that like to toy with Ethereal...
Posted: Thu Oct 31, 2019 12:05 am
by mar
Hmm, could this be a hint that Fritz 17 is Ethereal?
http://talkchess.com/forum3/viewtopic.php?f=2&t=72170
Re: For those that like to toy with Ethereal...
Posted: Thu Oct 31, 2019 2:13 am
by MikeB
Dann Corbit wrote: ↑Wed Oct 30, 2019 11:40 pm
In a test (every early) the Ethereal pruning is doing better than the Stockfish style.
Please define 'better' , can I assume you meant in position testing as opposed to head to head combat??
Re: For those that like to toy with Ethereal...
Posted: Thu Oct 31, 2019 4:53 am
by AndrewGrant
MikeB wrote: ↑Thu Oct 31, 2019 2:13 am
Dann Corbit wrote: ↑Wed Oct 30, 2019 11:40 pm
In a test (every early) the Ethereal pruning is doing better than the Stockfish style.
Please define 'better' , can I assume you meant in position testing as opposed to head to head combat??
http://chess.grantnet.us/viewTest/3607/
Going to figure things out
Re: For those that like to toy with Ethereal...
Posted: Thu Oct 31, 2019 5:43 am
by Dann Corbit
MikeB wrote: ↑Thu Oct 31, 2019 2:13 am
Dann Corbit wrote: ↑Wed Oct 30, 2019 11:40 pm
In a test (every early) the Ethereal pruning is doing better than the Stockfish style.
Please define 'better' , can I assume you meant in position testing as opposed to head to head combat??
I have a match running. It's set for 100K games, but I won't let it go that far.
When I left, SF pruning was ever so slightly ahead
Re: For those that like to toy with Ethereal...
Posted: Thu Oct 31, 2019 7:55 pm
by Dann Corbit
No difference in my test, but I discovered that Arena was logging like mad (I had a 550 MB arena.debug file) so I think the result cannot be trusted.
Code: Select all
Program Elo + - Games Score Av.Op. Draws
1 Ethereal-stopr : 3333 11 11 1475 50.1 % 3333 58.6 %
2 Ethereal-ethpr : 3333 11 11 1475 49.9 % 3333 58.6 %
Re: For those that like to toy with Ethereal...
Posted: Thu Oct 31, 2019 10:45 pm
by AndrewGrant
Dann Corbit wrote: ↑Thu Oct 31, 2019 7:55 pm
No difference in my test, but I discovered that Arena was logging like mad (I had a 550 MB arena.debug file) so I think the result cannot be trusted.
Code: Select all
Program Elo + - Games Score Av.Op. Draws
1 Ethereal-stopr : 3333 11 11 1475 50.1 % 3333 58.6 %
2 Ethereal-ethpr : 3333 11 11 1475 49.9 % 3333 58.6 %
NMPFormula vs master DIFF
ELO | 0.62 +- 1.98 (95%)
SPRT | 10.0+0.1s Threads=1 Hash=8MB
LLR | -2.96 (-2.94, 2.94) [0.00, 5.00]
Games | N: 46370 W: 9180 L: 9097 D: 28093
Not bad, but Ill stick with current master.
Re: For those that like to toy with Ethereal...
Posted: Thu Oct 31, 2019 10:46 pm
by AndrewGrant
AndrewGrant wrote: ↑Thu Oct 31, 2019 10:45 pm
Dann Corbit wrote: ↑Thu Oct 31, 2019 7:55 pm
No difference in my test, but I discovered that Arena was logging like mad (I had a 550 MB arena.debug file) so I think the result cannot be trusted.
Code: Select all
Program Elo + - Games Score Av.Op. Draws
1 Ethereal-stopr : 3333 11 11 1475 50.1 % 3333 58.6 %
2 Ethereal-ethpr : 3333 11 11 1475 49.9 % 3333 58.6 %
Code: Select all
NMPFormula vs master DIFF
ELO | 0.62 +- 1.98 (95%)
SPRT | 10.0+0.1s Threads=1 Hash=8MB
LLR | -2.96 (-2.94, 2.94) [0.00, 5.00]
Games | N: 46370 W: 9180 L: 9097 D: 28093
Not bad, but Ill stick with current master.