An interesting tidbit is how many rules sets make a unit degrade. Typically you roll to hit…usually a 4+. Then you roll a save…typically a 5+. That yields a 1/3 chance of hitting. Or, the expected value of 3 hits is 1. Now, many games don’t want to remove a stand straight away. So they make each stand take 2 hits. So now we remove a stand every 6 turns instead. Of course this is pretty broad brush but you get the idea. So, just roll a single roll and be done with it! Ultimately, for me anyway, it’s all statistics. No matter how involved you make the process, it can all be boiled down to a percent chance of succeeding at a task.
A friend of mine has a pet peeve. Some games we play seem to go too far with the rolling of the dice only to arrive at the conclusion that there is no effect. I’ve seen role play game masters do this over and over again. A complete waste of time. Wargame designers are often guilty of this as well. This brings me to the working title and subtitle of the game.
“Throw me a ‘6’!
(What part of ‘I hit’ didn’t you understand?)”
John's Wargame Page 
Exactly, John. Why randomise umpteen times? Even DBx can be boiled down to a simple percentage, and a small table – chances of lock, kill, or recoil. The whole “ratio of two scores” thing is unnecessary.
This dice-fest tendency came out of a completely nonseniscal notion that combat results tables were somehow tedious or confusing. Actually, being able to cross-reference something on a table is a pretty fundamental skill for just about any functioning adult in today’s world, and I always appreciated how explicit and clear they were for budding junior generals. Also you got, as a designer, to allow randomness, but keeping things within a plausible range. So they tended to be easy to control for effect.
But the simple rolling a single die and applying a result (as in 1HWG) is actually as simple a process as you can get without a combat results table. Modify the result up or down as needed, based on troop type, quality, terrain or situation. You don’t get simpler or clearer.
I suggest using at most 5%-20% as the significance level for randomness, and that would mean at most, a single d5-d20 for even the most complex interactions. Probabilistic effects less than 5% are not statistically significant, and can be pedagogically safely ignored.
In the few games I played, the usual amount of a successful hit has been 1. In fact, I only saw a double hit once and that was a lucky hit from a phalanx rolling 3 dice. There will be variations in this game…probably far greater than you subscribe since it simply is a dice rolling mechanic. Better units in better situations roll more dice thus getting a better chance for a hit.