Myracle GUI

[mar]

ok but that was not the indended use for minimum ply...

so you want to accept the move if time is out and move is correct regardless of minimum ply limit?
could do it, but then the test will run much slower, always consuming max time

[chessica]

ok but that was not the indended use for minimum ply...

so you want to accept the move if time is out and move is correct regardless of minimum ply limit? ---> yes

could do it, but then the test will run much slower, always consuming max time ---> correct

[Wilhelm]

so you want to accept the move if time is out and move is correct regardless of minimum ply limit? ---> yes
could do it, but then the test will run much slower, always consuming max time ---> correct

Thats the only way to get correct results. You can't be sure that an engine didn't change the move after ply 4 eg.
or on the other side the engine has the correct move for 2 ply and the get failed result when not reached eg. 4ply in time. The problem is that the engines have very different outputs and speed.

For this testing I use Deep Shredder 12 x64, Senpai 1.0, Crafty 25.03 and Chessmaster. Test-Suite is WM-Test 2024, info and download here https://descent.at/files/wmtest.html

William

[mar]

ok, got it.
it will be fixed in the next build

[Elorejano]

Hi, Martin!
Sorry for this. For engines that not have ownbook (Obsidian, Plenty, Berserk), how can i use a book with those engines in tournament mode? I cant see this option.
Thanks in advance

[vencenda]

- attempt to parse eval, depth and time graphs when loading a pgn

Thanks for this, although as you may have seen, in games where there is a win (with a mate), the high values of +M5, +M4 etc, effectively destroy any visual information the graph gives, because all early normal eval values look like 0. I think a maximum value of 7 or 10 or something should be used to avoid that, since above that, it's always a win in practice.

[mar]

Thanks for this, although as you may have seen, in games where there is a win (with a mate), the high values of +M5, +M4 etc, effectively destroy any visual information the graph gives, because all early normal eval values look like 0. I think a maximum value of 7 or 10 or something should be used to avoid that, since above that, it's always a win in practice.

yes, I'll clamp eval scores to +-20 in next build

[mar]

Hi, Martin!
Sorry for this. For engines that not have ownbook (Obsidian, Plenty, Berserk), how can i use a book with those engines in tournament mode? I cant see this option.
Thanks in advance

you need to setup a book first in settings/books, then select it from the book combo in new tournament dialog

[Elorejano]

Hi
Thanks for the replay, but did not work.

[mar]

build 69 is up:
- score in eval graph clamped to +-20
- run epd suite: position is declared as passed if time is up and move is found
- show draw rate in crosstable footer