I know that at least a few of you have already crunched the numbers and determined that Power Attack does indeed increase your DPR.
I'm no math guru, but intuitively it seems like your attack bonus and the defender's AC doesn't actually impact the math. Rather, its the difference between the two that determines how often you land hits. In that case, there must be some threshold where the difference between your attack bonus and the defender's AC is great enough that PA starts to hurt your DPR.
I would also expect this threshold to vary depending on which tier your at, as well as one handed vs two handed. Perhaps the number of [W] that your attack deals probably plays a role as well?
Anyone care to take a stab at laying out a matrix that will indicate exactly which situations you can be certain using PA will be beneficial?
I know that at least a few of you have already crunched the numbers and determined that Power Attack does indeed increase your DPR.
I'm no math guru, but intuitively it seems like your attack bonus and the defender's AC doesn't actually impact the math. Rather, its the difference between the two that determines how often you land hits. In that case, there must be some threshold where the difference between your attack bonus and the defender's AC is great enough that PA starts to hurt your DPR.
I would also expect this threshold to vary depending on which tier your at, as well as one handed vs two handed. Perhaps the number of [W] that your attack deals probably plays a role as well?
Anyone care to take a stab at laying out a matrix that will indicate exactly which situations you can be certain using PA will be beneficial?
The change in your DPR depends on 1) The impact on your hit chance percentage-wise, 2) Your damage on a hit. (and sometimes 3) your damage on a miss, and 4) the impact on your miss chance) For example: If your hit chance is halved and your damage is doubled, you'll even out.
It depends on the exact damage you'll be dealing on a hit and miss and the exact impact that it'll have on your chance to hit and miss.
Favored Enemy wrote:
I'll start it off by suggesting the use of Power Attack anytime you hit the opponent's defense+2 on a 2+.
I'm no math guru, but intuitively it seems like your attack bonus and the defender's AC doesn't actually impact the math. Rather, its the difference between the two that determines how often you land hits. In that case, there must be some threshold where the difference between your attack bonus and the defender's AC is great enough that PA starts to hurt your DPR.
The to-hit roll is what actually matters, tis so. And there is a point at which you see declining returns.
However, the point depends on not just your to-hit roll, but how much damage you deal in the first place; the more damage you deal in the first place, the worse power attack becomes.
The to-hit roll is what actually matters, tis so. And there is a point at which you see declining returns.
However, the point depends on not just your to-hit roll, but how much damage you deal in the first place; the more damage you deal in the first place, the worse power attack becomes.
Exactly.
You'd need a separate table for each tier, and each would look like: --Average Damage on a hit w/out PA - Miss damage = Damage Hit chance = Hit | change in DPR (or simply Yes/No, for PA or not)
The formula for the table would be: (Damage+(PA damage))*(Hit - 0.1) - Damage*Hit = (PA dmg)*(Hit - 0.1) - Damage*0.1 (makes sense, right? You're sacrificing 10% of your regular damage over the damage you'd do on a miss (damage*0.1 = average damage on hit - average damage on miss, the part you'll miss because of the hit reduction) for PA damage on all of your new hit %s (which is now 10% lower)) The formula would be incorrect/ignore the end points.
Running away from any math way of view this(the math part i let for other ppl =P)
Some conclusion:
-The penalty is fixed. The damage bonus is progressive. So, it work better at higher tiers. Also, using a two-handed(or a versatile as two-handed). (But also, you must notice, as people said above, that the more damage you already do without PA... the more significant is the risk of miss. If your at-will damage is pitty... often the risk of a +9 worth. But, if your damage is already incredible high, then the 10% miss chance can become dangerous)
-Power attack got amazing sinergy with powers that give absurd attack bonus, like Lead the attack(Warlord). A swordmage, for example, using lead the attack(maybe + 5 to +8 power attack bonus) should definitelly grab power attack, as the penalty become less relevant. There are other powers, like armor splinter, imperiling strike, etc. Same happens with all the penalties. Stack penalty, CA, etc.
-Also, elements that do some kind of damage on a miss, like the feats, somehow punish you less when the penalty you draw from power attack make you miss your target.
-Sometimes, more damage is important to make your hit relevant against a resistant foe. If your foe got resist 15 against your attack, and your damage range is 10-18... then you should always use power attack.
I'm a little nuts for the math, so I ran some of the numbers, borrowing DemonLord57's basic math, and adding in controls for the top and bottom of the scale (i.e. 1s always miss, 20s always hit), and accounting for Miss Damage - though I still haven't added controls for crits.
I'm including the "break even" points where using Power Attack becomes beneficial. The numbers are based on the average expected damage per attack and the difference between the target's AC bonus and your Hit bonus (e.g. if the target has AC 17 and you have a +5 to hit, then the difference is -2). So to find out when you should power attack, figure out your expected damage on a hit, find that number along the top row, and if your hit bonus is greater than the listed value, you should power attack (statistically). (Taking the assumption that my math/formulae are correct)
Adding miss damage shifts the number down slightly, but never more than a few points.
I'm working on tweaking the spreadsheet to handle more information, if anyone wants a copy, let me know and I can post it someplace.
Another interesting note - if you can normally only hit with a 20, then it's in your best interest to Power Attack. Even with the penalty, you still get the same hit% and more damage (slightly).
I'm a little nuts for the math, so I ran some of the numbers, borrowing DemonLord57's basic math, and adding in controls for the top and bottom of the scale (i.e. 1s always miss, 20s always hit), and accounting for Miss Damage - though I still haven't added controls for crits.
I'm including the "break even" points where using Power Attack becomes beneficial. The numbers are based on the average expected damage per attack and the difference between the target's AC bonus and your Hit bonus (e.g. if the target has AC 17 and you have a +5 to hit, then the difference is -2). So to find out when you should power attack, figure out your expected damage on a hit, find that number along the top row, and if your hit bonus is greater than the listed value, you should power attack (statistically). (Taking the assumption that my math/formulae are correct)
Adding miss damage shifts the number down slightly, but never more than a few points.
I'm working on tweaking the spreadsheet to handle more information, if anyone wants a copy, let me know and I can post it someplace.
Another interesting note - if you can normally only hit with a 20, then it's in your best interest to Power Attack. Even with the penalty, you still get the same hit% and more damage (slightly).
Cool. By the way, the miss damage should be subtracted directly from the damage, to calculate 'Damage'. Think of it as finding the damage you lose by not hitting. (damage you would normally do over what you do on a miss)
As for the 'always PA when you only hit on a 20', if you're critting (barely hitting with a +20, so you'd hit normally with a 20 while PAing) that's almost assuredly unsound advice.
Cool. By the way, the miss damage should be subtracted directly from the damage, to calculate 'Damage'. Think of it as finding the damage you lose by not hitting. (damage you would normally do over what you do on a miss)
This simplification works for the majority of the range, but fails at the edge cases (i.e. where Base Hit % - 10% != PA Hit %). The final formula I'm using is:
If the result is greater than 0, you're better off power attacking.
DemonLord57]As for the 'always PA when you only hit on a 20', if you're critting (barely hitting with a +20, so you'd hit normally with a 20 while PAing) that's almost assuredly unsound advice.
This is correct - I'd forgotten the 4E rule that while 20s always hit, you only crit if you would still have hit the target's AC. As such, you definitely don't want to PA when you're "in the margin" where a 20 will hit the target's AC - but when you need a 22 or higher to hit, then the swinging for the fences rule still applies. (Of course, that situation should almost never come up in a real game, unless your DM is a total jerk. :evil wrote:
As for the 'always PA when you only hit on a 20', if you're critting (barely hitting with a +20, so you'd hit normally with a 20 while PAing) that's almost assuredly unsound advice.[/quote] This is correct - I'd forgotten the 4E rule that while 20s always hit, you only crit if you would still have hit the target's AC. As such, you definitely don't want to PA when you're "in the margin" where a 20 will hit the target's AC - but when you need a 22 or higher to hit, then the swinging for the fences rule still applies. (Of course, that situation should almost never come up in a real game, unless your DM is a total jerk. :evillaugh )
This simplification works for the majority of the range, but fails at the edge cases (i.e. where Base Hit % - 10% != PA Hit %). The final formula I'm using is:
If the result is greater than 0, you're better off power attacking.
Ah, right, you were including the edge cases.
drunkenewok wrote:
This is correct - I'd forgotten the 4E rule that while 20s always hit, you only crit if you would still have hit the target's AC. As such, you definitely don't want to PA when you're "in the margin" where a 20 will hit the target's AC - but when you need a 22 or higher to hit, then the swinging for the fences rule still applies. (Of course, that situation should almost never come up in a real game, unless your DM is a total jerk. :evillaugh )
This is why I wasn't too worried about not including the edge cases: It simply won't come up in a game. A simplified formula is good enough for any reasonable situation you'll face. Still, a more advanced (necessarily more complex) formula is better in all ways except simplicity (by its nature.)
