This thread has been dead a month, I know. But timing is everything, and my timing's off, so here's my rather-more-than-2 cp opinion. First off, here's a tool anyone can use to calculate their optimum PA for full attacks and single attacks, if they don't mind NOT being a PrC or epic character:
http://www.geocities.com/frisbeet/DandDandFightin.html
Now why would you want to do this?
Nail hit it right on:
In play, it's true you have to guess the AC of the opponent.....but frankly, that's pretty straightforward. And then you have (printed out in bright, cheery colors!) a graph that shows your optimal PA for each AC. Quite helpful and speedy.
But even this is probably unnecessary; knowing the range of probable ACs your opponent has gives a corresponding error on the PA guess, so it's usually safe enough to estimate, which can be done in a second or two at most.
Example:
"My barbarian charges at the Dire Tiger, and swings his sword. Mumble: (I guess the Dire Tiger's AC is ~ 18, so I power attack for 5.)"
"Hey, hey! I hit! ....and I did 44 hp of damage! ...Err, it's not dead yet, is it?! Oh."
Guessing AC isn't about meta-gaming. Korak's almost right:
Your interpretation of the feat is as valid as mine. I prefer to think of power attack as an experienced fighter sizing up his opponent and deciding how much power he can add at the expense of accuracy based on his opponent's skill.
Except the "sizing up" doesn't have to be conscious--maybe it's just the fighter's "athletic intelligence" at work--talent, if you will.
Now for jgsugden insightful yet overreaching comments:
1.) ... The truth of the matter is, unless you're hitting very easily with your worst attack (or are having trouble hitting at all with your best attack), you're unlikely to want to power attack for much, even with a 2 handed weapon (if your main goal is to increase your average damage per hit versus a set AC.) This means that the only times power attack really helps you out is in situations where your foe is pathetically easy to hit or is incredible hard to hit - which should be rare circumstances.
This point is somewhat subjective, so it's hard to dispute: who decides what AC is "patherically easy" or "incredibly hard" to hit? But consider this, taken directly from the sheet on my site--do a couple of example calculations yourself if you're untrusting of these #s--
Fighter 12
Str 24
Feats of WF, gWF, WS, gWS, iC for Greatsword
Wields a +3 Greatsword.
Here's his average full attack damage/round, taking into account critical hits, when not power attacking, and then when power attacking with the optimum PA.
FYI, there's nothing tricking about how this spreadsheet got these #s--you can't do calculus with this highly parameterized discrete problem to get a sol'n. My sheet brutally calculates all possible PA results and spits out the best one in the display. That's why it's so $#*&) big (and because of DR), sorry dialups.
Code:
opp AC 0 PA opt PA w/ PA
10 82.08 12 152.64
11 82.08 12 149.76
12 82.08 12 144.00
13 82.08 12 138.24
14 82.08 11 132.48
15 82.08 11 126.96
16 82.08 10 121.44
17 82.08 10 116.16
18 82.08 9 110.88
19 80.64 9 105.84
20 79.20 8 100.80
21 77.76 7 95.76
22 76.32 6 90.72
23 74.88 5 85.68
24 72.00 4 80.64
25 69.12 3 75.60
26 66.24 2 70.56
27 63.36 1 65.52
28 60.48 1 60.84
29 56.16 1 56.16
30 51.84 0 51.84
31 47.52 0 47.52
32 43.20 0 43.20
33 38.88 0 38.88
34 34.38 0 34.38
35 30.00 0 30.00
36 25.74 0 25.74
37 22.86 0 22.86
38 19.98 0 19.98
39 16.92 0 16.92
40 13.98 0 13.98
41 11.16 0 11.16
42 9.72 0 9.72
43 8.28 0 8.28
44 6.66 12 7.56
Get this far? My point: ACs 22-27 don't seem that uncommon for a level 12 encounter. This takes into account those 2nd & 3rd attacks, too. Power attack is
useful.
2.) Power attack requires information to optimize. If you don't know a foe's AC, it is very hard to know how to optimize your PA. As a result, the number of times a PC actually gets the maximum benefit out of power attack is rare.
Guess the AC. It makes combat a lot more interesting--makes it easier to roleplay, ironically.
3.) Power attack is most useful in single attack scenarios (slowed, attacking after moving, when you can't attack except for AoOs, etc ...) instead of full attack scenarios. In those instances, it is fine for a fighter to do an obscene amount of damage with one hit as it will be his only attack for the round (with the exception of AoOs).
Here's how the same fighter does single attacking, again from the sheet on my site:
Code:
opp AC 0 PA opt PA w/ PA
10 27.36 12 54.72
11 27.36 12 54.72
12 27.36 12 54.72
13 27.36 12 54.72
14 27.36 12 54.72
15 27.36 12 54.72
16 27.36 12 54.72
17 27.36 11 52.44
18 27.36 10 50.16
19 27.36 9 47.88
20 27.36 8 45.60
21 27.36 7 43.32
22 27.36 6 41.04
23 27.36 6 38.88
24 27.36 5 36.72
25 27.36 5 34.68
26 27.36 4 32.64
27 27.36 4 30.72
28 27.36 3 28.80
29 25.92 3 27.00
30 24.48 2 25.20
31 23.04 2 23.52
32 21.60 1 21.84
33 20.16 1 20.28
34 18.72 0 18.72
35 17.28 0 17.28
36 15.84 0 15.84
37 14.40 0 14.40
38 12.96 0 12.96
39 11.52 0 11.52
40 10.08 0 10.08
41 8.64 0 8.64
42 7.20 0 7.20
43 5.76 0 5.76
44 4.14 0 4.14
45 2.64 0 2.64
46 1.26 12 2.52
So whereas with full attack, you are best PA-ing through AC 27, in this case you you're best through AC 31. I'd have thought the difference would be more dramatic, as jgsugden implies.
Jgsugden's last two points are dead on.
All IMO, of course.
Edit: calculations corrected as of 3/22. Had parameterized the GS as doing 3 extra damage points/hit, which is silly. Architecture of spreadsheet unchanged.