cutechess-cli 0.8.0 beta (with pondering)

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

Moderator: Ras

User avatar
Ozymandias
Posts: 1537
Joined: Sun Oct 25, 2009 2:30 am

Re: cutechess-cli 0.8.0 beta (with pondering)

Post by Ozymandias »

SzG wrote:
Ozymandias wrote:Could be have 6-men syzygy adjudication, instead of 5-men gaviota?
Not sure that is a good idea. Even 5-man adjudication is questionable because the engines may not find the win/draw without tablebases.
I don't get your point, it's the GUI that adjudicates the game.
User avatar
Ozymandias
Posts: 1537
Joined: Sun Oct 25, 2009 2:30 am

Re: cutechess-cli 0.8.0 beta (with pondering)

Post by Ozymandias »

SzG wrote:
Ozymandias wrote:
SzG wrote:
Ozymandias wrote:Could be have 6-men syzygy adjudication, instead of 5-men gaviota?
Not sure that is a good idea. Even 5-man adjudication is questionable because the engines may not find the win/draw without tablebases.
I don't get your point, it's the GUI that adjudicates the game.
My point is that if an engine left to its own devices cannot win a 5-man ending then it is not justifiable to award it a win.
If the engine is capable of using EGTB's on its own, that's another thing. The GUI should not be allowed to help.

In my view only 4-man endings (but not KBNK) should be adjudicated.
That would be correct if engine testing was the only use for cutechess, but you also have book testing. Adjudication saves times, and TB adjudication is the only one that ensures 100% reliability. If be could use 7-men Lomonosov in cutechess, I would be asking for it.
User avatar
ilari
Posts: 750
Joined: Mon Mar 27, 2006 7:45 pm
Location: Finland

Re: cutechess-cli 0.8.0 beta (with pondering)

Post by ilari »

Ozymandias wrote:
SzG wrote:
Ozymandias wrote:
SzG wrote:
Ozymandias wrote:Could be have 6-men syzygy adjudication, instead of 5-men gaviota?
Not sure that is a good idea. Even 5-man adjudication is questionable because the engines may not find the win/draw without tablebases.
I don't get your point, it's the GUI that adjudicates the game.
My point is that if an engine left to its own devices cannot win a 5-man ending then it is not justifiable to award it a win.
If the engine is capable of using EGTB's on its own, that's another thing. The GUI should not be allowed to help.

In my view only 4-man endings (but not KBNK) should be adjudicated.
That would be correct if engine testing was the only use for cutechess, but you also have book testing. Adjudication saves times, and TB adjudication is the only one that ensures 100% reliability. If be could use 7-men Lomonosov in cutechess, I would be asking for it.
I agree. It should be up to the cutechess user to decide the adjudication rules.