D&D 4E The math behind power attack and why it needs to be redone in 4e

[Orryn Emrys]

Conclusions have been repeated several times in the thread: power attack is great for combatants who have very high attack bonuses and low damages, and who don't use a lot of iterative attacks (it's better for combatants wth high mobility). In 3.0 it was great for finesse fighters using rapiers and such. However, it isn't so great for fighters with large damage outputs. In fact, if your average damage is already 19 or more (say, flaming greatsword, Str 20 and Weapon specialization is enough) then on average you shouldn't power attack unless, after the PA penalty, you still hit on a 2+.

It has always bothered me a little bit that the broadest statistical advantage of a feat that seems intended to reflect somewhat-wilder, more powerful swings would most effectively be employed by characters wielding smaller, lighter weapons that don't deal a significant amount of damage without it. Interestingly enough, I can't think of a single player I've had in the last seven years who wielded knives and rapiers and such who turned to Power Attack as a low- to mid-level feat selection. They tend to go right to the feats that seem to more appropriately represent their specific weapon selections, while the greatsword and greataxe wielders have always been quick to snatch it up.

As for the effectiveness of the feat... predictably, mileage has certainly varied.

 

[BryonD]

Fine. Drop the word "optimal" out of my post, and its still true. You're the one getting hung up here.

You can't calculate whether power attacking is a good idea or not without knowing what your average damage per hit is. The ONLY time you can know whether power attacking is a good idea without knowing average damage per hit is when a roll of a natural one would still reach or exceed the target's armor class, and even then, you ONLY know that its a good idea to power attack until you hit on a natural roll of two.

For any attack other than that, you must know your average damage per hit to know whether power attack is a wise choice.

Editted to add: Actually, there's one other case where you don't need to know average damage per hit. If you only hit on a natural 20, you might as well power attack for all you're worth. You'll still be hitting on a 20.

I'm not sure where "every case" is better than not using power attack is "getting hung up".

And it isn't avg damage per hit, it is avg damage per swing, because the increased misses are the critical part of PA. Obviously in any case, not matter how poor, the avg damage per "hit" goes up with PA.

In my case the avg per swing is just under 19 with no power attack. +28 attack for 1d12+10 against AC 26 and a crit of 19/X3.
For PA ranging from -1 to -14 the avg damage ranges from 21.1 to 27.4. No matter what number is selected, even the most suboptimal choice of all (-1 in this case) using PA is always better than not using it.

Just look at the MM and you will see this is a very standard case. I have one player who flat out loves playing PA type characters and since 3E was released I could have grabbed a character for whom this general analysis was correct at just about any time. At low levels the bonus it pretty minimal, but that is the worst you can say for.

You are arguing that in only works in certain cases. The problem is that the certain cases exist a solid majority of the time and the exceptions are easy to spot.

 

[BryonD]

I still don't get how Byron thinks damage reduction works into this math either. Someone help me with that one

Edit 2: Didn't get the sarcasm
(and its Bryon)

Edit: In my case the average damage increase per swing if you assume completely random PA is 6.4 points. This is true with or without DR. Without PA the avg damage is 18.81. So against a creature without DR a completely random PA will increase damage by 34%. With DR 10 the avg hit without PA now deals 8.81 but the random PA increases remains 6.4. So the total damage output is improved by over 72% in the case with DR.

 

[BryonD]

Massive hit is an optional rule.

I didn't say massive damage, I said massive hit. Scoring a hit for over 100 points is fun in and of itself.

 

[Wolfspider]

I really didn't mean to come off that rude, I was trying to be playful. I had a long day at work, so give me a break. We can still call it najo's awesome attack though if you like.

We can start a thread about another feat and name it after you, do you want cleave or sunder?

Oooh! Tough decision. Let me think about it a while. (Heh. Considering my answer to your quesiton, maybe Dodge is in order...).

And it's ok. I was probably overly critical myself. It's the last week of classes, after all, and the papers are piling up....

 

[BryonD]

I got that part. I was, yet again, being playful and sarcastic (hence the ) and it back fired Still, the damage reduction doesn't play in to the mathematics of the feat until the final damage is determined, at that point it still doesn't matter, because we are showing how power attack effects average damage generated before applying it to the monster. Whether or not there is DR doesn't change those odds.

Ok, sorry, I missed this post.
Now, can you explain to me why you would ignore the applying it to the monster part?
Thats like saying pizza doesn't taste good because we are only cooking it and leaving out the eating it part. The whole point of PA is to apply more damage to the monster.

 

[Alaxk Knight of Galt]

Najo, your numbers look correct for -1 to hit gives +1 to damage. However, please consider the -1 to hit +2 to damage numbers

First Example
+hit hit % Ave Hit Damage per Round
15 50% 4.5+3 3.75
14 45% 4.5+5 4.275
13 40% 4.5+7 4.6
12 35% 4.5+9 4.725
11 30% 4.5+11 4.65
10 25% 4.5+13 4.375
9 20% 4.5+15 3.9
8 15% 4.5+17 3.225
7 10% 4.5+19 2.35

A gain of .975 Damage per Round (or DPR if you will). Not great, but better then nothing.

Second Example
+hit hit % Ave Hit Damage per Round
15 95% 4.5+3 7.125
14 90% 4.5+5 8.55
13 85% 4.5+7 9.775
12 80% 4.5+9 10.8
11 75% 4.5+11 11.625
10 70% 4.5+13 12.25
9 65% 4.5+15 12.675
8 60% 4.5+17 12.9
7 55% 4.5+19 12.925
6 50% 4.5+21 12.75

A gain of 5.8 DPR.

Obviously, neither of these take into account a 2nd (or third) attack. I like math, but not that much

Edit: Obviously, I do like math that much.
For a 2nd hit, just knock off 25% to hit on your selected hit percentage.


For the first example
Non-Powered attack
1st hit - 50% hit chance with 4.5+3: 3.75 average damage
2nd hit - 25% hit chance with 4.5+3: .9375 average damage
Total DPR: 4.6875

Power Attack
1st hit - 35% hit chance with 4.5+9: 4.725 average damage
2nd hit - 10% hit chance with 4.5+9: 1.35 average damage
Total DPR: 6.075

Second Example
Non-Power Attack
1st hit - 95% hit chance with 4.5+3: 7.125 average damage
2nd hit - 70% hit chance with 4.5+3: 5.25 average damage
Total DPR: 12.375

Power Attack
1st hit - 55% hit chance with 4.5+19: 12.925 average damage
2nd hit - 30% hit chance with 4.5+19: 7.05 average damage
Total DPR: 19.975

-------------------------------
I really hope my ability to do basic math is correct today

 

[Orryn Emrys]

Still, the damage reduction doesn't play in to the mathematics of the feat until the final damage is determined, at that point it still doesn't matter, because we are showing how power attack effects average damage generated before applying it to the monster. Whether or not there is DR doesn't change those odds.

Though it does, I feel, figure into the utility of the feat. The most enterprising aspect of Power Attack is the variable scale and application that is left up to the end user. It's up to the attacker to determine its utility in any given scenario. When confronted by an adversary that seems to soak up damage with little noticeable side effect, i.e. an enemy with Damage Reduction that allows it to shrug off a noticeable portion of any given attack, it behooves the player to consider the advantages of Power Attack in subsequent attacks. I suppose this makes it's utility extremely circumstantial, but it does demonstrate a subset of not-too-terribly-unusual exceptions to the cited averages.

Just a thought.

 

[Wolfwood2]

I got that part. I was, yet again, being playful and sarcastic (hence the ) and it back fired Still, the damage reduction doesn't play in to the mathematics of the feat until the final damage is determined, at that point it still doesn't matter, because we are showing how power attack effects average damage generated before applying it to the monster. Whether or not there is DR doesn't change those odds.

Completely wrong.

The less average damage you do per hit compared to your attack bonus, the better Power Attack is for you. That's already been established in the thread. DR effectively lowers your average damage.

So a character who normally does 1d12+7 damage when going up against a DR 5 creature should calculate his Power Attack as if he were doing 1d12+2 damage.

Speaking of power attack in general, I think "average damage" is overrated as a way of determining effectiveness. How quickly you drop an enemy is of vast importance, and the mere chance of dropping them quicker can be beneficial.

Consider a case where you can normal attack and drop an enemy in three rounds of hitting or power attack and probably drop an enemy in two rounds of hitting. The "average damage" is the same, because of the reduced odds of hitting when power attacking. However, it's to your benefit to try the power attack, because that gives you a shot at doing the two rounds drop in the first two rounds. You might not succeed, but if you do then that's one round of attacks the enemy doesn't get on you. Normal attacking will mean you take three rounds of attacks no matter what.

 

[Lord Zardoz]

A probable flaw in your reasoning

There is one big flaw with considering things purely from an average amount of HP per round that a user can inflict.

If the goal were to "run up the highest possible damage total in the shortest amount of time", then your approach is good.

But the goal is actually to "Kill the target I am attacking before it kills me".

So lets assume you are up against an opponent and using a 2 handed weapon doing 1d10 damage, x2 on a crit threat of 19, and have 18 Str, and a Bab of 6. Your target has AC 18 and 40 Hp.

You can either go for 1d10+6 damage at +10 to hit, or 1d10+18 at +4 to hit.

In one case, your hitting on a roll of 8 or higher (60% of the time) and doing an average of 11.5 damage. You need to hit about 4 times to score a kill, and a crit will basically count as a free hit. You need 6 or 7 attacks (0.6 * 6 = 3.6, round up) to be sure of hitting 4 times and getting a kill.

In the other case, you only hit on a 14 or higher (35% of the time) and doing an average of 23.5 damage. You only need to hit twice to end the fight. Once if you can crit. Rounding things up, you need about 5 or 6 (0.35*5=1.75, round up) attacks to be sure of a kill.

In any event, there are a whole hell of a lot of things worth considering in D&D combat, and many ways to interpret the data. And simply because attempting to use a given ability may allow for a suboptimal choice from time to time is hardly a reason to consider the mechanic broken. And attempts to consider a given feat in isolation are also fairly futile. Too many things can change from one fight to the next. Is power attack more or less optimal when you are being attacked by 3 creatures? A Rogue who can consistently trip and feint you and score sneak attack damage? Against a creature with Fast Healing?

END COMMUNICATION