Quote:It actually *is* in the manual
The description in the manual is clear, what I meant was that the way it works in practice isn't always clear. If the low and high casualty modifiers, and the associated randomness thrown in at the end, are far apart, it can become highly unpredictable. That's both a good and bad thing, as you will never know precisely how good your assault will be, firing can give you more stable results on the long term (depending on how good your soft/hard attack values are relative to the defence value of the target, obviously).
You could calculate where the low and high casualties are for the assault you're planning, but there's still substantial randomness involved. Twice even, first with determining what casualty figure is used between the high and low values and then for determining what the actual casualties are.
That's also why you need to get a feel for it, as if it would be as simple as a CRT, you would get immediate feedback as to why a certain result happened (for example: you rolled a four and the result is A1-D1R). Currently, you get feedback on casualties, but the assault having a bad or good result doesn't always directly relate to the forces involved.
It isn't immediately clear how bad, mediocre or good your roll was relative to the forces involved.
The randomness can make removing defenders from strong defensive positions a pain, also because both sides use the same stacking limitations, based on a single hex (this is the only serious wargaming series I can currently think of which imposes such a penalty on the attacker).
Basically, the calculations used are in the manual (for assaults, direct fire is trickier to calculate), but you don't see the system at work when you attack (as the calculations determining the result are not displayed), it remains abstract.