Its a very long time since I did Statistics, choosing decision maths and mechanics as my modules at college.

Now, I'm fine with simple things, working out hits, wounds, saves ect, using the simple process of dividing by 6 and multiplying by 7-the dice roll needed.

So if you have one shot, and you need a 4 to hit, 1/6*(7-4) = 0.5

And if you have rerolls, original shots, - predicted hits, gives you predicted rerolls.

For Melta, I simply created a matrix and used a variety of ifs and lookups.

And now I have got to fleet.

Now, my matrix works for normal charges, one out of 36 results gets you a 12" charge, 3 in 36 get you an 11" charge, 6 in 36 give you a ten" charge and so on.

Now, with a flat reroll, thats workable.

But its not a flat reroll.

If you need an 8" charge, and roll a 6 and a 1, you would just reroll the one

Thats fairly simple, but I've seen people argue if you rolled a 5 and 2, if you rerolled the 2 for another 2, you could reroll the 5 hoping for a 6.

And that raises a lot more issues

Now, without mapping every possible combination, of which there are at least 1296, possibly a LOT more, I cant think of a way to make the numbers dance.

So, does anyone have any thoughts?

Has anyone brute forced the results?