2 dice produces an easy results chart. Propose a formula, lay out the distribution, consider the probabilities, assign results accordingly.
For example suppose your formula is:
FP + Acc – (AR + Ev) = ODC
where
FP= Firepower, Acc= Accuracy D6,
AR= Armor, Ev = Evasion D6,
ODC = Opposed Dice Check
A 2D6 chart is easy to lay out, FP+Acc on the vertical axis, AR+Ev on the horizontal. Plot the results on the grid. I won't render that here.
For FirePower 5, Armor 6, the ODC range is (fractions rounded):
4 = 1/36 ~= 3%
3 = 2/36 ~= 6%
2 = 3/36 ~= 8%
1 = 4/36 ~= 11%
0 = 5/36 ~= 14%
rest = 21/36 ~= 58%
Your desired results range when hitting is:
. . . Normally . . . bounce off . .. But . . . might penetrate
So if "might" is 3% (once every 36 times), then require a 4 result to penetrate. If "might" is more often, a 3+ happens about 8% (once every 12 times), a 2+ about 17% (every 6th time).
If 58% of shots missing is a reasonable number, then that's your negative result.
The range in between (0 to 2, 0 to 3, 0 to 4 depending) represents "bounce off". A hit that didn't penetrate.
The flaw with this system is that higher firepower shots are inherently more accurate and harder to evade.
So what you may prefer to do is throw an accuracy die versus an evasion die (Acc – Ev = ODC) : Negative result = miss. Positive result= hit. High positive result, specially accurate hit (if you want to model that).
Then compare FP versus AR to determine effect. Typically you'll want a range of results here (i.e., a chart) which can be determined either by rolling a 3rd die, or just carry over the positive hit ODC result (so higher accuracy => more damage).