IA de lolo introduces his AI

ULA (Uncle Lolo’s Abalone) is among the most powerful programs playing to Abalone. It has been developed by Laurent Pagli, best known as IA de lolo. I let the programmer introduce himself his work:

° ° ° ° ° °

ULA (Uncle Lolo’s Abalone) is a PC program running on Windows.
This program has been developed in a Borland Delphi language, a kind of Pascal language.
The functioning of this AI is classic: minimax with alpha–beta pruning, conversion tables, and pruning a priori move options, to go faster.
To give an idea of its game level, with a same time of reflection, ULA is a little bit less strong than Aba-Pro (to my despair), but its strongest level has already defeated it!
Graphical user interface allows playing (fortunately!), but I would have liked to have improved more if I would have had the time and patience… There again, I think that Aba-Pro is much better made.

So why use ULA program?
– Its playing style is not the same as Aba-Pro
– We can give it a playing style more or less aggressive and introduce randomness to add variety in the game
– We can force it to avoid a repetition of the position
– We can fix a starting position an solve a problem like “black wins in x moves”, or “how many moves to defeat Black?”
– ULA allows seeing the whole sequence of moves planned for it and its opponent
– It has a great number of variants, and you can ever take the option pillar (the central space is locked)
– I think that its highest level – if you have a powerful PC… and a lot of patience! – is probably unbeatable.

ULA lacks an opening book, and maybe a lot of other things. I did let I go a little and it is not completed, and still in process.

If you want to try it, and give me your opinion or suggestions, or if you want further information about its programming, send me a mail to:

iadelolo@laposte.net

I will send to you its temporary .exe.

iadelolo, march 12, 2014, 16h43 min (France)

Advertisements