Risk Programming Challenge Thread

Hello fellows,

Posting on the behalf of my university institute (i5 for computer sciences at the RWTH Aachen / Germany), I'd like to inform you about the 2nd CUELC Programming Challenge, this time featuring the well known classic Risk.
The challenge is open for all students, with the tasks to create a GUI and/or a working AI for Risk. You're invited to work in teams and to join the discussion with the other participants to determine issues like the communication protocols etc. As a challenge it mostly serves to satisfy your own interest and training, but to compensate at least little for your efforts we want to donate some prizes to the winners of the challenge.

If you have any questions, feel free to ask.

Regards,
Stephan

this contest does kinda sound cool ...

i think the GUI would be much much easier to make than the AI though ...

I tried something similar but a much larger based multiplayer network.

Frankly, its out of my league to tackle something like that.
I'd be more inspired if the prize was something more than a DS.

for mission 1 each of the members in the winner team gets an IPod
Shuffle; for mission 2 each of the members in the winner team gets a
Nintendo DS Lite.

Well, seeing as I am American I am definitely out but the Shuffle isn't very specific. Details, size, etc would attract people to it more.

im more curious as to the approaches people are taking for making the AI rather than competing ...

are they using a system of game theory to calculate an evolving long term strategy that gives the best possible outcomes kinda like deep blue does with chess? or are they doing something realistic for humans to play against ... (basically, do whats best in the current situation turn or so ahead)

game theory can be tricked by illogical, dumb luck though :/ ... its just the odds of it happening are ... small

also keep in mind that there are many many more unique possiblities than chess ... considering pieces can be taken or created ... also there are more than 2 teams available to play with, the game is not symmetrical (not each team has equal opportunities from the start)

EDIT: The actual criteria they list are:

Evaluation Criteria:

• consistency of interface
• efficiency of algorithm measured by score achieved

So doesn't look like they want genetic alorithms or fancy AI experiment. Just design the all-around best algorithm.

Edited by: GoodieStake