Arena 3.5.1 and Mate Test epd requirements

Discussion of chess software programming and technical issues.

Moderator: Ras

chessica
Posts: 922
Joined: Thu Aug 11, 2022 11:30 pm
Full name: Esmeralda Pinto

Arena 3.5.1 and Mate Test epd requirements

Post by chessica »

Hello engine tester, i have a small database with mate tasks.
I would like to test the database with the help of arena with many engines.

Question:

How must the epd line look correct, e.g. h2-h4 and mate in 4?
Dann Corbit
Posts: 12777
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Arena 3.5.1 and Mate Test epd requirements

Post by Dann Corbit »

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.
chessica
Posts: 922
Joined: Thu Aug 11, 2022 11:30 pm
Full name: Esmeralda Pinto

Re: Arena 3.5.1 and Mate Test epd requirements #1

Post by chessica »

Is this fen correct, or incorrect?

in the automatic analysis, the train is not recognized as a solution, even though the mate was found???

[fen]5K2/8/2qk4/2nPp3/3r4/6B1/B7/3R4 w - e6 0 1 [/fen]

FEN: 5K2/8/2qk4/2nPp3/3r4/6B1/B7/3R4 w - e6 0 1

SOS 5.1 for Arena:

1/3 00:00 4 65 -1,62 d5xc6 Td4xd1
1/4 00:00 10 163 +M1 d5xe6/ep+
2/4 00:00 39 639 +M1 d5xe6/ep+

best move (SOS 5.1 for Arena): d5-e6
not found in: 00:04
2/2 00:00 31 508 +M1 d5xe6/ep+

what is wrong here?
JVMerlino
Posts: 1396
Joined: Wed Mar 08, 2006 10:15 pm
Location: San Francisco, California

Re: Arena 3.5.1 and Mate Test epd requirements

Post by JVMerlino »

The FEN is correct, but your EPD might not be correct in identifying the best move. What is the full EPD that you are using for this position?
chessica
Posts: 922
Joined: Thu Aug 11, 2022 11:30 pm
Full name: Esmeralda Pinto

Re: Arena 3.5.1 and Mate Test epd requirements

Post by chessica »

JVMerlino wrote: Wed Mar 08, 2023 12:22 am The FEN is correct, but your EPD might not be correct in identifying the best move. What is the full EPD that you are using for this position?
This is the epd-line:

5K2/8/2qk4/2nPp3/3r4/6B1/B7/3R4 w - e6 bm #1; ep; 00:00;
Dann Corbit
Posts: 12777
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Arena 3.5.1 and Mate Test epd requirements

Post by Dann Corbit »

I think that is a ChestUCI EPD row.
The bm expression is not correct according to the standard.
This part is OK:
5K2/8/2qk4/2nPp3/3r4/6B1/B7/3R4 w - e6

But the bm should be a best move (perhaps the e.p. pawn capture)
bm #1;
This is meaningless:
ep;

This is meaningless:
00:00;

Those things look like ChestUCI database annotations to me (which do not follow the standard)
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.
chessica
Posts: 922
Joined: Thu Aug 11, 2022 11:30 pm
Full name: Esmeralda Pinto

Re: Arena 3.5.1 and Mate Test epd requirements

Post by chessica »

Is it so correct?

5K2/8/2qk4/2nPp3/3r4/6B1/B7/3R4 w - e6 bm d5xe6 #1; ep;
chessica
Posts: 922
Joined: Thu Aug 11, 2022 11:30 pm
Full name: Esmeralda Pinto

Re: Arena 3.5.1 and Mate Test epd requirements

Post by chessica »

chessica wrote: Wed Mar 08, 2023 5:02 pm Is it so correct?
5K2/8/2qk4/2nPp3/3r4/6B1/B7/3R4 w - e6 bm d5xe6 #1;


Is my assumption correct that in Arena in the automatic analysis of
the epd file only the key move, here d5xe6, is checked and the mate
is not? If this is true, then my test is obsolete. I want to exclude
that the key move is played for wrong reasons.
JVMerlino
Posts: 1396
Joined: Wed Mar 08, 2006 10:15 pm
Location: San Francisco, California

Re: Arena 3.5.1 and Mate Test epd requirements

Post by JVMerlino »

chessica wrote: Thu Mar 09, 2023 10:48 pm
chessica wrote: Wed Mar 08, 2023 5:02 pm Is it so correct?
5K2/8/2qk4/2nPp3/3r4/6B1/B7/3R4 w - e6 bm d5xe6 #1;


Is my assumption correct that in Arena in the automatic analysis of
the epd file only the key move, here d5xe6, is checked and the mate
is not? If this is true, then my test is obsolete. I want to exclude
that the key move is played for wrong reasons.
I don't know of any GUIs or tools that check for more than just playing the correct move when processing an EPD. In other words, if you want to make sure that your engine solves this position not only by playing dxe6ep but also declaring mate in 1 as specified in your EPD, I don't think any program can do that.