Breaking the Other Guy's Weapon

Weapons break.  Battlefields are littered with broken weapons as well as broken bodies.  In my view, most breaks should be essentially random—the result of a critical miss, or some such—or should (sometimes) be the result of a successful kill—ramming that sword into the dragon’s belly was forceful enough to kill it, but also forceful enough to break the sword.

But some breaks are the result of intentional actions in combat using weapons designed to break other weapons.  In my game this would cover larger members of the axe family—the military axe, battle axe and halberd as well as the two-handed sword.

The mechanic would be the same as the disarm mechanic of the previous post, with these differences/exceptions:

The base to break a weapon is 16 (as opposed to the 11 of disarm).

Only average to large axes and the two-handed sword can be used to break a weapon.

Only wooden hafted weapons can be targeted.  (Metal weapons would break on random rolls.  In other words, you can't use that battle axe to slice the other guy's dagger in two, but dagger blades are still pretty vulnerable to being chipped or even snapped off during a turn of furious combat.)

The two-handed sword gets a +4 modifier.  (As I understand it, one of its main uses in Renaissance combat was in breaking up pike formations by lopping the heads off of the pikes.)

Defending with a two-handed weapon gets no modifier.

I’m still working on a chart for random weapon breakage that’s both realistic seeming and simple.  You want a rapier to have a much greater chance of breaking than a mace, but you want to use the existing stats (weight? length? the damage roll?) to simulate that.  You don’t want some extra number—breakability or whatever—fussing things up.  The armor of the defender should also be a factor.  If you're attacking a guy in plate armor with your rapier, there's a much greater chance you'll break the point.  And again there’s that balance question.  You don’t want weapons breaking all the time, but you do want them to break often enough to be something that should be factored into the give and take of play (as opposed to a totally rare and obscure random event).  So does that mean that breaks should on average occur in, say, 1 out of 100 attack rolls? 1 out of 500?  1 out of 1,000?  I’m not sure I know the answer to that.