Hi Everyone,

I'd would like to know your opinion regarding whether the community would consider a Monte Carlo simulation for Risk cheating or not.

The idea would be to define a number of attacking units, which attack a number of territories with a few defending units each. One would then simulate this attack a couple of thousand times to determine:

• The probability of the attack being successful
• The expected value of the number of remaining units of the attacker
• Any other fun statistics

Essentially, this would be like running the applet on the following site over and over again:

http://recreationalmath.com/Risk/

In my view this would just be a decision help, as it does not give you a black or white answer. You don't let a program do the work / decisions for you (=cheating), but rather still have to come up with the move yourself and decide whether you want to do the attack for that 73.12% probability of winning, or not.

Looking forward to your thoughts on this!

Cheers

Not quite sure, but I think one of our own has already created an applet that does this: The Wargear simulator on the other side returns the type of information I think you're looking for.

http://prestopnik.com/wargear/

There's also this calculator, which does almost the same thing:

http://gamesbyemail.com/games/gambit/battleodds

The biggest problem with this calculator is that it only does d6 v d6, which doesn't apply to boards that use border modifiers.

I don't use such tools myself because I prefer the challenge of working it out in my head, but I don't consider it cheating if my opponent use it. It's impossible to monitor anyway. 

The point of a probability is that it's not certainty. That's why the game has dice in the first place.  

If you know the probabilities, your probably going to be a better player than someone who doesn't.  Being better is not cheating.

M57 wrote:

Not quite sure, but I think one of our own has already created an applet that does this: The Wargear simulator on the other side returns the type of information I think you're looking for.

http://prestopnik.com/wargear/

I've been meaning to update that with an option to take multiple territories of defenders, and you just put the # in each territory separated by commas.  So something like:

10,6,1,1,1

Would mean you need to take a territory with 10, then 6, then 1, etc. 

I guess I'd need to add a checkbox for whether it's abandon or not, but other than that, it seems fairly straightforward.   Anything I'm missing? (other options, confounding variables, preferred output, possible pitfalls, etc.)

Litotes wrote:

I don't use such tools myself because I prefer the challenge of working it out in my head, but I don't consider it cheating if my opponent use it. It's impossible to monitor anyway. 

Of course it's not cheating.  I would say that the information that can be gleaned from these tools can't really be worked out in your head, but most good players can come up with a decent estimate.  For instance, say I'm playing a game and I'm wondering what the probability is that I can attack 11 v 11 and have at least 5 units left over.  Off the top of my head, I'm going to guess that it's 20%.  Quick check on the calc and it looks like it's about 30%.  Not bad.

Hi everyone,

since the general consensus was: "of course it's not cheating", I wrote a small script to do just this.

Since I felt like keeping it to myself would be an unfair advantage, I decided to share it with everyone. Feel free to mess around with it:

https://github.com/tehingo/risk-monte-carlo/blob/master/Risk.py

Warning: major code gore!!