New Engine: Athena

[Poobear]

Hey!

My name is Nicolas and I'm from Canada. Over the past 2 or so years, I have been on and off working on a chess engine. It's gotten to the point where I think I'm finished with it, and would like to get it an official rating. Would it be possible to run it in some CCRL tournaments/gauntlets?

The source code is available at: https://github.com/NicolasSegl/Athena

And the executable is available at: https://github.com/NicolasSegl/Athena/r ... tag/v1.0.0

I estimate it has an elo of around 2000

Thanks!

[Gabor Szots]

Hi Nicolas,

I'm overwhelmed with other things at the moment but I'm going to take a look at it as soon as I have a chance.

[Poobear]

No rush! Thanks for considering

[Gabor Szots]

In the meantime, would you include version number in UCI id and engine name? As it is it may cause future problems when more versions will be in play.

[Poobear]

Sure thing! You should be able to download the newer version at: https://github.com/NicolasSegl/Athena/releases

[Gabor Szots]

Gauntlet started with version 1.0.1. Watch out for this week's update!

PS. I still have to rename the file before saving it.

[Poobear]

Thanks so much! Next update I'll change the name of the .exe

[Gabor Szots]

I have just submitted 512 Athena games. Unfortunately 16 of them ended by illegal moves of which I cannot tell more. Played under Arena.

[Poobear]

Huh that's weird I've never had it make any illegal moves before. Do you have the PGNs of FEN strings or any other way I could replicate the position on my own machine?

Thanks.

[Poobear]

I'll test out the engine in Athena, usually I use cutechess

[Gabor Szots]

Here is the PGN. I cannot provide further data.

[Guenther]

It's a pity that Arena did not include the assumed illegal move in the pgn.

At least the biggest part of the illegal moves seem to come from not being able to promote to a second Queen!
(in a very quick position fen test on one of those positions from the pgn it just sent a1a1 as a move here on cmd)
Probably you have tested with early adjudication, which hides a lot of bugs

A part of the other ones seem to have to do with giving up a castling Rook and may be still trying to castle?
(not verified nor test here)

The rest seems inexplicable w/o further info by 'Arena'.

Guenther

[Gabor Szots]

For this tournament I used the following adjudications:

Adjudicate game below -1100 centipawns.
Draw after 230 moves.
Draw by tablebases (both engines have to agree).

Yes, it's a shame Arena does not tell us what the illegal move was.