Rugger said:
This is really bugging me...
Now, I'm no statistician or even a "math person"....but the justification that light weapons get a greater percentage increase in damage vs bigger weapons strikes me as poop. The game is based on abstract concepts (HP and Damage) and its the FINAL damage total that matters....not the fact that a +2 to damage is a 50% increase for a dagger and only a 25% increase for a longsword.
If i'm bungling the mentality here, please let me know so I can remove my foot from my mouth...
You're bungling the mentality here.
Look at my analysis again -- it's not that Bob (the greatsword fighter) gets 3 extra damage while Joe (the two-shortsword fighter) gets 3 extra damage as well, but it's a bigger percentage for Joe.
In all scenarios, Joe actually gets
more total points of damage in his expected per-round total than does Bob. This is because Joe's percentage rate of damage increases
more than his percentage attack rate decreases, while Bob's percentage rate of damage increases
less than his percentage attack rate decreases.
If you're using a greatsword, in almost all cases in D&D3, it
lowers your per-round expected damage to use Power Attack. Not "increases by a flat rate, but not as much, percentage-wise."
Lowers.
Someone below you said that at 8th level, you use PA to keep up your rate of damage output on rounds that you have only one attack (for whatever reason). This is almost certainly incorrect, and I'll suggest why below:
You're an 8th level Fighter. You have a modified 20 Strength, (perhaps naturally, perhaps using Strength buffers), a +2 Greatsword, and Weapon Specialization in Greatsword. Your total damage is 2d6 + 11, for an average of 18 points of damage per round, before criticals. (Note that this character is not particularly optimized for damage).
When does it make sense for you to Power Attack, assuming you only get one attack a round?
Let's suppose you're just PAing for 1 point of damage. Let's suppose that x is the number on a D20 you have to roll, after all modifiers, to hit your opponent.
Your odds of hitting normally (without PA):
1 - .05x
Your odds of hitting with 1 point of PA:
1 - .05x -.05 = .95 - .05x
Your expected damage without PA: 18.
Your expected damage with PA: 19
We want:
(1 - .05x) * 18 < (.95 - .05x) * 19
For what values of x is this true?
18 - x * (18/20) < (361/20) - x * (19/20)
(1/20) * x < (361/20) - 18
x < 361 - 20 * 18
x < 361 - 360
x < 1
That's right! Unless your attack bonus is already so high that you already (would) hit on a 1 (if it weren't for the auto-fail rule), your 8th level Fighter expects a
decrease in per-round damage when using any PA at all,
even when he's denied his iterative attacks.