My house rules mechanics for battle solves a lot of problems for me.
Don't get me wrong. I love d&d mechanics, but mainly because it just feels good rolling that d20 and then (if I am lucky) that d6. It is kind of like mechanical nostalgia. But it just doesn't make a whole heck of a lot of sense to me.
The folks that put that together were war gammers, scaling down down down to one to one ratio of man to figure. But they still had the mechanics of mass combat in their bones where things are chancier (fog of war!) and where you don't mind troops getting swept away turn after turn because, a.) you want the game to move and b.) you've got plenty more troops coming!
But random, linear progression for "to hit" and then random damage (allowing for low damage after a perfect hit) just doesn't make much sense in actual one to one combat. Actual human beings have a skill set and are going to succeed within that skill set following a bell curve, not a linear progression randomly meted.
Furthermore, the notion that one's defense is based entirely upon ones armor makes sense if you are a ship of the line (the real source of "armor class" was naval warfare!) but makes no sense considering that self-defense in battle is actually one of your skills!
So here is my solution:
Attacker and defender each get an attack and defense roll, respectively, on 2d. The "d" depends upon what kind of beast or PC you are. So, let's say a human fighter is 2d10 and a goblin is 2d6. Most of the time the fighter is going to hit that silly little goblin. Most of the time that goblin is not going to be able to get past a human fighter's defenses. Human fighter, on average, is going to hit at 11, and the goblin is going to defend at 7, and vice versa.
The mechanic for determining damage is then built into the throws: Successful attack minus unsuccessful defense equals damage. So, in the above scenario, the goblin takes 4 HP damage.
This gives everybody something to do at every phase of the battle. That in itself is psychologically satisfying, and my "play testers" attest to the fact!
I am leaning on a great online article "A Treatise on Dice Mechanics," for this house rule. According to the author's three goals, 1. speed and ease of use, 2. precision and 3. "correctness," that is to say, correspondence to what you are trying to "map" in the real world, I am purchasing 3 at the cost of both 1 and 2, well, at least 1. But it is worth it to me. Yes, the players and the GM have to do a little math, but not hard math. Addition and subtraction. We can handle it. The mechanic is more satisfying because both attacker and defender feel like they are playing a role, as in real combat, and because the target is no longer absolute but relative. In other words, the attacker has a moving target. That, also, is more like combat. We obtain reflexivity in combat. It is a bit more "complex," and that makes play move a bit "slower," but we have fun all the same.
The author of that really helpful online article also worries about how to morph a bell curve as we have PCs that raise in level. I do not see a problem here. With every level my "knight" class (basically "fighting man" but all my PCs are "lawful") adds another bonus to their attack and defense rolls. My "scouts" (alignment neutral "thieves") add a bonus every other level up. Wizards every four levels. So the bell curve of relative skill remains the same, but one's "sweet spot" continues to improve. Much like the development of skills in real life.
Yeay bell curves and a mechanic that maps more to "reality" (whatever that means in a fantasy game!)!