I know the dice slightly favor the attacker, but where is the tipping point? The spot where you and the other guy build and build but you should just attack him because the dice will win.

I'm in a crab game just now on hawaiian islands, and all three of us have around 150 units. Is this the time to attack?  I mean if I attack 150 of another players units with my 150 units, will I have 10 or 11 left with which to occupy the territories?

Which number of units would make it statistically reasonable to attack both players, and count on the attacker advantage for a win? 3 units? 10 units? 100? 1000?

You've reached the dreaded 3-player finale where each is too strong to take out without (a) pissing someone off royally or (b) taking so many units to kill one that you leave yourself too weak to compete with the remaining player.  Best to all agree on a vote to term.

I've always found this to be a great simulator when trying to answer such questions:

http://prestopnik.com/wargear/

Given equal numbers, I have it in the back of my mind that 9v9 is the tipping point.  Any less and defender has a better than even chance of surviving an attack.. Slander's post pointing (Ozyman's?) calculator is a great way to solve the problem.

.85 loss per army is to be expected..... 150 x 6/7 is....but Hawaiian islands has a few dice mods?

...is about 22 units left, which should be about right. 

It's got a pair of 6v7 borders but a max of 7 units on those. (which can totally be used to great effect)

I've definately crunched a thousand numbers with that calculator. :)

I guess where I fall down is when do the numbers become statistically significant? I've lost enough 7 units attack 3 unit battles to know that's not quite a guarantee.  

But thinking about it, 9 v 9 as M57 suggests - that would be 4 rolls that would have to be sour in order to make it an abysmal loss (which I think I might be quite surprised at the 4th one. - so that makes sense to me.)

Just pumping #'s through ozy's server, I've kinda landed on this:

(using stats from 1000 simulations) - I'm gonna need to start attacking when I've got about 300 units, and they have about 300 units so that I can get a guy on all 42 territories - just based on the slight attacker advantage. 

Yikes. That's gonna be forever...

I did some investigation on this topic. As mentioned above you need .85 units for each of you ropponents units on average. Another way to put it is that with .85 x defending units gives you a 50 % chance to win. A multiplier of 1.0 gives a 70% chance and a multiplier of 1.25 gives a 90% chance. So I will usually estimate the % chance of winning a battle and then decide to attack. For example, when considering a 'take out' I usually wait until I have a 70% chance of winning. Not sure if this helps..

Lol, SG (and all), the numbers are great, but what is the translation to what should actually be done?  I think it's an unwinnable situation. As ratsy says, it's gonna be...forever.

Yes, agreed

The only thing i could think of for a situation like that is to attack each opponent equally to keep numerical equality but have each players total be much lower. Then if there are cards to be turned in you could have enough to defeat both. So if the next set of cards is worth 50, then 300 v 300 v 300 is a stalemate, but 50 v 50 v 50 gives you a chance. Depends on how your opponents react to the tactic

I prefer to terminate in those situations, but occasionally I have had luck with a mutually assured destruction pact.   For example in a crab game where we all have 150 units, everyone attack with 50 units the next turn.  A attacks B, B attacks C, C attacks A.   Or each attacks each person with 25 units each.

Each person would lose ~50 attacking, and ~60 being attacked.  So at the end, you're all down to about 40 units, and the game is more destable.

Yes, that's an alternative if everyone agrees to it. Hadn't thought of that. 

SquintGnome wrote: The only thing i could think of for a situation like that is to attack each opponent equally to keep numerical equality but have each players total be much lower. Then if there are cards to be turned in you could have enough to defeat both. So if the next set of cards is worth 50, then 300 v 300 v 300 is a stalemate, but 50 v 50 v 50 gives you a chance. Depends on how your opponents react to the tactic

I've nuked this particular game two times now (attacked both players will all I had), as we've all hit the 100 unit mark before, and I can get em down to about 30 units apiece (which of course devastates me).  But the cards are not too helpful. The cycle from 4-9 units. 

Oh that's the worst! I'm not generally a fan of cycling card values for this reason. Surprised the other two didn't take advantage of your attacks!

Wish that board(more boards) had unlimited reserves...HI allows 15?....cute....