Probabilities of the attacks in Risk!!

[Marvaddin]

Probabilities of the attacks in Risk... I saw it in a faq about the board game. I think it can be helpful to you all

Attacker 3 X 2 Defender
Att lose 2: 29.26% (2275/7776)
Def lose 2: 37.17% (2890/7776)
Each lose 1: 33.58% (2611/7776)

Attacker 3 X 1 Defender
Att lose 1: 34.03% (441/1296)
Def lose 1: 65.97% (855/1296)

Attacker 2 X 2 Defender
Att lose 2: 44.83% (581/1296)
Def lose 2: 22.76% (295/1296)
Each lose 1: 32.41% (420/1296)

Attacker 2 X 1 Defender
Att lose 1: 42.13% (91/216)
Def lose 1: 57.87% (125/216)

Attacker 1 X 2 Defender
Att lose 1: 74.54% (161/216)
Def lose 1: 25.46% (55/216)

Attacker 1 X 1 Defender
Att lose 1: 58.33% (21/36)
Def lose 1: 41.67% (15/36)


If there is someone interested, how about note all rolls in a match and see if its really working here in Conquer Club?

Hmm, if prob of defense wins are 34% in a single roll 3 X 1, how much is the prob of the single defender wins 10 rolls? I think I never won a 1 X 1 roll here in Conquer Club

[Nobunaga]

... I've won several 1 x 1 rolls, though I've only made those roles for a lack of intelligent options or as a last, desperate fight for a card to continue survival.

... Defender wins all ties - so throwing anything less than 3 dice is unwise. in my humble opinion, that is.

[Marvaddin]

I agree, I never attack with less than 3 dice (unless its very necessary...)

Now, I remember, I won an 1 on 1 roll (that way I eliminated a player, then the game was a brise, it was a triples game; I got the cards, and held the Rockies in USA map )

[supermarcol]

Are you sure about those numbers? I remember doing the calculation once and getting a slight advantage on defence for 3vs2. oh well, I probably did some mistake somewhere. These calculations get so messy... I'll try to run through a program with all possibilities and figure out what it gives.

[EDIT] I have tried my program and it gives the same result as these. These should therefore be the actual odds.