Page 4 of 4

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Wed Jun 13, 2018 8:47 pm
by Leo
Right. TCEC is very necessary. Maybe Spassky would have defeated Fischer after a scientific amount of games.

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Wed Jun 13, 2018 9:04 pm
by AndrewGrant
The best we can hope for from this thread is that corres will learn to use the quote function

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Wed Jun 13, 2018 9:11 pm
by Guenther
AndrewGrant wrote:
Wed Jun 13, 2018 9:04 pm
The best we can hope for from this thread is that corres will learn to use the quote function
He will not. He was already asked for using it correctly a few times in the past.
His answer was that he wants to keep his special style...and will never use the quote function.

(He was one of the first I added to my ignore list anyway for various other reasons)

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Wed Jun 13, 2018 9:28 pm
by AndrewGrant
special style
Special alright ....

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Fri Jun 15, 2018 8:27 am
by corres
Poor Stockfish is so unfortunate that after 48 turns its Elo loss enhanced to 121 Elo.
If the chance do not behave accidental-like it is not merely unfortunate - at least in my opinion.

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Fri Jun 15, 2018 11:05 am
by sovaz1997
corres wrote:
Fri Jun 15, 2018 8:27 am
Poor Stockfish is so unfortunate that after 48 turns its Elo loss enhanced to 121 Elo.
If the chance do not behave accidental-like it is not merely unfortunate - at least in my opinion.
Here begins the Superfinal, where Stockfish himself will show his strength

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Fri Jun 15, 2018 12:55 pm
by JJJ
Komodo has still the best contempt against weaker engines. If Komodo would be as strong as Stockfish dev, he might have lead this tournament with much more points. So Stockfish contempt is nice, but not as effective as his counterpart Komodo. Komodo might still be the best engine at imbalance as well.

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Fri Jun 15, 2018 2:00 pm
by sovaz1997
JJJ wrote:
Fri Jun 15, 2018 12:55 pm
Komodo might still be the best engine at imbalance as well.
Very much I doubt. We need tests to prove this fact.

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Sun Jul 01, 2018 1:54 pm
by Leo
corres wrote:
Tue Jun 12, 2018 11:36 am
Uri Blass wrote:
Mon Jun 11, 2018 8:53 pm

I think that the expected result may be 69% when stockfish was lucky in season 11 and unlucky in season 12
You would be right if the participants are the same.
But it is not the situation.
Now Stockfish is relative weaker than it was.
Its hard to imagine luck playing a role in the Super final. Maybe Stockfish is getting really lucky in season 12 and we need 20,000 games to prove it has won fairly. ( Assuming it goes on to win the Super final.)

Re: Developer tests of Stockfish need Stockfish 8 instead of Stockfish 7

Posted: Sun Jul 01, 2018 6:20 pm
by mjlef
JJJ wrote:
Fri Jun 15, 2018 12:55 pm
Komodo has still the best contempt against weaker engines. If Komodo would be as strong as Stockfish dev, he might have lead this tournament with much more points. So Stockfish contempt is nice, but not as effective as his counterpart Komodo. Komodo might still be the best engine at imbalance as well.
Since I wrote Komodo's COntempt features, I know quite a lot about it. But I would not claim its contempt is better without data. Perhaps you can give some data?

Mark