Ah, but keeping the opponent's three's and four's between me and his sixteen's is all that keeps me safe sometimes.

Sides, there'd have to be at least a chance of penalty or something.

Like stated previously, I think that this system is by far better if you make this ONE improvement. Treat the dice rolls like risk:

Each dice goes up against one other dice from the other side.

Therefore if you go 5 vs 2,
1 Dice beats the other dice, it becomes 5 vs 1
Rinse and repeat. It might keep things a bit less 'random' for the higher amounts of dice. Plus the computer will churn through it no problem through simulation.

I'm...not quite sure I understand, PointSeven. Run through an attack for me?

Player A: 5 dice, Player B: 2 dice
The simulation would run:
1 dice from each team face off
A rolls a 2, B rolls a 6

Player A: 4 dice, Player B: 2 dice
A rolls a 5, B rolls a 1

Player A: 4 dice, Player B: 1 dice

A rolls a 6, B rolls a 4

Player A: 4 dice, Player B: 0 dice
Player A wins spot.

Or alternatively, break down the attacks so that the ENTIRE force doesn't face off vs the opponents ENTIRE force. This really cripples the game at the moment due to the fact that the actual 'odds' of winning a roll are not spread out over the course of multiple rolls, thus having the effect of equally distributing the odds. Almost akin to the law of large numbers.

That post made no sense, update 2.0
----------------
Or alternatively, break down the attacks so that the ENTIRE force doesn't face off vs the opponents ENTIRE force. The lack of this really cripples the game at the moment due to the fact that the actual 'odds' of winning a roll are not spread out over the course of multiple rolls. By splitting the attacking / defending force (for example, half of attacking vs half of defending) the 'odds' have a greater chance of being true. Almost akin to the law of large numbers.

Or do it like risk:

Taken from wikipedia


In a typical attack in which both the attacker and defender have several armies, the attacking player may roll one, two, or three dice. The defending player roll one or two dice. The highest attacker's die is compared against the highest defender's die, and the second-highest attacker's die is compared against the second-highest defender's die. The lowest attacker's die is disregarded. The defender loses an army if the attacker's die is greater than the defender's own, but the attacker loses an army if the defender's die is greater than or equal to the attacker's die.

Intresting. SO, what does this do to the probabilities?

It is a lot easier to lose 16vs14 than 3vs1, 4vs2, etc...

With both stacks greater than 8, you start getting closer to 50/50 battles, even with a difference in stack height.

I think this is a good thing. There is supposed to be a random element in the game. the 8stack games happened from people trying to defeat the random element. It's going to be a lot tougher for them to defeat it now. They'll have to figure out strategies that incorporate the random element, rather than trying to get rid of it. One is to be more aggressive about taking states, rather sitting around stacking up dice, since that's now not as helpful.

Good point. Just look at the chart we already have, you can see the trends.
http://en.wikipedia.org/wiki/Kdice#Dice_probabilities

A dice advantage becomes less and less effective, while upsets (7 beating 8) become more and more likely.