Nutshell: Use a two-handed (and ideally reach) weapon, look for multipliers, and remember that Leap Attack and the Frenzied Berserker's boosts were erratad to be additive. Also, it really, really helps to build a quick spreadsheet for assorted Power Attack values to account for enemy AC, miss chances, and other things that lower your chances to hit and thus your expected damage. Max PA may not always be the best course of action.
Max PA may not always be the best course of action.
That's incorrect. There is never any circumstance where not power attacking for full is the best option. Your attack bonus should be high enough at any given time to hit enough.
@OP: The simple answer is yes, and here it is. In order don't, always max it, and really, really don't.
Max PA may not always be the best course of action.
That's incorrect. There is never any circumstance where not power attacking for full is the best option. Your attack bonus should be high enough at any given time to hit enough.
@OP: The simple answer is yes, and here it is. In order don't, always max it, and really, really don't.
Elusive target would disagree with you on that.
Elusive target would disagree with you on that.
A niche one, but an important one. Also, when attacking a target with attack negation (eg. Wall of Blades style effects) or when fighting multiple foes in a full attack, you will find attacks aimed at targets at base attack bonuses lower than your maximum. While it's not difficult to get your best attack to basically autohit, your subsequent ones are an uphill battle, and if you need them to hit, a lower PA can result in higher expected damage. There are also builds that rely on attack rolls that will not benefit from power attack (although this is splitting hairs, it does mean "max PA may not always be the best course of action" and that "there is never any circumstance where not power attacking for full is the best option" is false) - one example would be Edge of the Light from my sig, on the opening round (he's taking max penalty on Combat Expertise there to feed huge numbers forward for offense, but the numbers get bigger if the attacks hit at all - the optimal PA in that situation is 0).
Against single targets without Elusive Target or attack negation, a full PA is a great default. But it's not the only option. Be aware of your build and your situation.
Elusive target would disagree with you on that.
No, it doesn't. Because nobody takes it.
A niche one, but an important one. Also, when attacking a target with attack negation (eg. Wall of Blades style effects)
Then hit them twice. They only have one immediate action.
or when fighting multiple foes in a full attack, you will find attacks aimed at targets at base attack bonuses lower than your maximum. While it's not difficult to get your best attack to basically autohit, your subsequent ones are an uphill battle, and if you need them to hit, a lower PA can result in higher expected damage.
Except if your best attack is an autohit, at least one more attack should hit. And that means two dead opponents. Better than three or four injured ones.
There are also builds that rely on attack rolls that will not benefit from power attack (although this is splitting hairs, it does mean "max PA may not always be the best course of action" and that "there is never any circumstance where not power attacking for full is the best option" is false) - one example would be Edge of the Light from my sig, on the opening round (he's taking max penalty on Combat Expertise there to feed huge numbers forward for offense, but the numbers get bigger if the attacks hit at all - the optimal PA in that situation is 0).
Fine, I concede that, if you build specifically to avoid always max power attack, then power attacking isn't the best move. Whoda thunk, eh?
Against single targets without Elusive Target or attack negation, a full PA is a great default. But it's not the only option. Be aware of your build and your situation.
Not really. Nobody actually takes Elusive Target. It's like comparing something to a Matrix sequal; sure, it's nice and all, but it would work better if there was such a thing as a Matrix sequel.
Thats a pretty big assertion that no one ever takes elusive target.
No, not really. It's a downright garbage feat, with even worse prereqs.
I've seen the feat used several times in a couple of different groups (particularly in some rather nice twf builds).
First off, I hardly see how some poorly-optimized builds are relevent. Second off, good and TWF are mutually exclusive. Simple mathematics.
The long and short of it is if your group expects you to use beef cake smack down tactics all the time, odds are your going to see the feat in use at one point or another.
No, you won't. Because it sucks. Pathetically minor protection against a specific target? Maybe if it said "Any attack where the attacker is using power attack automatically misses," it might be worth it. But it doesn't. "Not getting killed" isn't a worthwhile feat investment. Killing the enemies is infinitely more efficient.
I've seen the feat used several times in a couple of different groups (particularly in some rather nice twf builds).
I hardly see how some poorly-optimized builds are relevent.
You said no build uses elusive target. He points out that at least one build has used elusive target. This proves your statement wrong. It's an issue of basic logic.
Likely, you meant that no well-optimized build uses elusive target. However, since you couldn't be bothered to say what you meant, you instead posted a demonstrably false claim that made you look foolish.
Fine, I concede that, if you build specifically to avoid always max power attack, then power attacking isn't the best move. Whoda thunk, eh?
As you imply, it's quite obvious that some niche situation can be contrived when power attacking for your full base attack bonus isn't the best choice. Far from being a defense, this only highlights how foolish it was for you to claim the opposite.
Likely, you meant that power attacking for your full base attack bonus is usually the best choice. You may have intended the stronger assertion that it's virtually always the best choice. No one but you knows, because you couldn't be bothered to accurately put your thoughts into words.
You said no build uses elusive target. He points out that at least one build has used elusive target. This proves your statement wrong. It's an issue of basic logic.Likely, you meant that no well-optimized build uses elusive target. However, since you couldn't be bothered to say what you meant, you instead posted a demonstrably false claim that made you look foolish.
Considering Cyclone_Joker is on record referring to our Evasion Tank as a "great build", and the Evasion Tank makes use of Elusive Target (awful prereqs and all), and CJ's identical rhetorical style to EA, one would assume that even that second statement isn't completely true, either.
I say this while agreeing that competent warriors should probably default to full Power Attack.
I've seen the feat used several times in a couple of different groups (particularly in some rather nice twf builds).
I hardly see how some poorly-optimized builds are relevent.
You said no build uses elusive target. He points out that at least one build has used elusive target. This proves your statement wrong. It's an issue of basic logic.
No, it isn't. Because in an online discussion of optimization, poorly optimized or unoptimized builds just don't matter. Just because one can take a build involving several levels of commoner, reaping mauler, and and take nothing but prereqs, weapon focus, and Skill Focus(Speak Language), it doesn't matter.
Likely, you meant that no well-optimized build uses elusive target. However, since you couldn't be bothered to say what you meant, you instead posted a demonstrably false claim that made you look foolish.
In a discussion off optimization, the two are synonymous.
Fine, I concede that, if you build specifically to avoid always max power attack, then power attacking isn't the best move. Whoda thunk, eh?
As you imply, it's quite obvious that some niche situation can be contrived when power attacking for your full base attack bonus isn't the best choice. Far from being a defense, this only highlights how foolish it was for you to claim the opposite.
No, not really. Coming up with bizarre niche situations is not a valid rebuttal.
Likely, you meant that power attacking for your full base attack bonus is usually the best choice. You may have intended the stronger assertion that it's virtually always the best choice. No one but you knows, because you couldn't be bothered to accurately put your thoughts into words.
Again, nope.
Considering Cyclone_Joker is on record referring to our Evasion Tank as a "great build", and the Evasion Tank makes use of Elusive Target (awful prereqs and all), and CJ's identical rhetorical style to EA, one would assume that even that second statement isn't completely true, either.
To be fair to CJ, "great build" doesn't mean "flawless build."
Considering Cyclone_Joker is on record referring to our Evasion Tank as a "great build", and the Evasion Tank makes use of Elusive Target (awful prereqs and all), and CJ's identical rhetorical style to EA, one would assume that even that second statement isn't completely true, either.
To be fair to CJ, "great build" doesn't mean "flawless build."
That's entirely possible, though we'd really have to hear from CJ to know his opinion on the matter.
But then, "flawless" isn't necessary, since a "great" build is a valid enough description for "well-optimized" in the context of optimization.
That's entirely possible, though we'd really have to hear from CJ to know his opinion on the matter.
Fair enough.
But then, "flawless" isn't necessary, since a "great" build is a valid enough description for "well-optimized" in the context of optimization.
I dunno. I'm having trouble seeing any justification for such a waste in the build. And if there isn't, then, by definition, it's poorly optimized, or at least not well-optimized.
Thenagain, that sort of choice is frequent in almost all of AR's more recent showcase builds(No offense meant, AR, I just believe your earlier builds were much more elegant and well-crafted).
Max PA may not always be the best course of action.
That's incorrect. There is never any circumstance where not power attacking for full is the best option. Your attack bonus should be high enough at any given time to hit enough.
As a point of information, you effectively just saying that using PA is the best course of action when using PA is the best course of action. Maybe your AB should be high enough, but if it's not, you should be thinking about not using PA. Plus you have those iterative attacks you might want to hit, too.
Anyway, this wasn't meant to be a discussion about PA, only that I know there are plenty of things like Leap Attack and Cobalt Power which work with it and wondering if anyone had ever put together a list. TBH I'm more interested in grappling but I was curious about others. I'll check out the Black Blood Cultist Handbook.
I've seen the feat used several times in a couple of different groups (particularly in some rather nice twf builds).
I hardly see how some poorly-optimized builds are relevent.
You said no build uses elusive target. He points out that at least one build has used elusive target. This proves your statement wrong. It's an issue of basic logic.
No, it isn't. Because in an online discussion of optimization, poorly optimized or unoptimized builds just don't matter. Just because one can take a build involving several levels of commoner, reaping mauler, and and take nothing but prereqs, weapon focus, and Skill Focus(Speak Language), it doesn't matter.
The simple matter is you made the assertion that no one takes the feat ever. Your clearly wrong on that. Even if you want to go back and make the amendment that no good build would make use of it you would still be wrong just on the basses of the evasion tank alone, not to mention any other possible progressions most anyone could come up with with a little thought on the matter. On that note, even if the evasion tank isn't "flawless", I compel you to find a build that is save for pun pun or the Omniscificer.
To say that the feat is bad because it doesn't kill the enemy is also a bad argument since living long enough to do anything is key in any kind of combat role. After all it's kind of hard to kill someone when you yourself are dead from their full power leap attack with the full serving of uber charger sauce. By investing in a few feats and using a reach weapon you can effectively shut down the primary damage source for most melee specialists at any time you take an action by designating them as your dodge. I honestly think you seriously undervalue the ability of this feat when used properly.
The simple matter is you made the assertion that no one takes the feat ever. Your clearly wrong on that. Even if you want to go back and make the amendment that no good build would make use of it you would still be wrong just on the basses of the evasion tank alone, not to mention any other possible progressions most anyone could come up with with a little thought on the matter. On that note, even if the evasion tank isn't "flawless", I compel you to find a build that is save for pun pun or the Omniscificer.
Wizard 5/Mindbender 1/Incantatrix 10/Divine Oracle 3/Sacred Exorcist 1.
Bard 7/Abjurant Champion 2/Divine Crusader 1/Sublime Chord 1/Mystic Theurge 9.
I can think of several more builds if you want. There's actually a great Bard one with Iniate of Mili stored around somewhere.
But flawlessness aside, no. You're wrong. Because in a discussion of optimized builds, Skill Focus(Speak Language) doesn't exist.
To say that the feat is bad because it doesn't kill the enemy is also a bad argument since living long enough to do anything is key in any kind of combat role. After all it's kind of hard to kill someone when you yourself are dead from their full power leap attack with the full serving of uber charger sauce. By investing in a few feats and using a reach weapon you can effectively shut down the primary damage source for most melee specialists at any time you take an action by designating them as your dodge. I honestly think you seriously undervalue the ability of this feat when used properly.
And this is how I know you aren't used to people knowing their stuff. Immediate action teleport cloak from DotU. Never leave home without it. Basic protections are something you should have with WBL, not by wasting feats.
You are wrong. It's a truly awful feat.Worse than SF(Speak Language) under most circumstances.
Awesome smite
brutal strike
brute fighting
combat brute
cobalt power
dragonfire assault
favored powerattack
flay
knockback
leap attack
ragewild fighting
shock trooper
there are probably more and this should be not rating on them to be good
The simple matter is you made the assertion that no one takes the feat ever. Your clearly wrong on that. Even if you want to go back and make the amendment that no good build would make use of it you would still be wrong just on the basses of the evasion tank alone, not to mention any other possible progressions most anyone could come up with with a little thought on the matter. On that note, even if the evasion tank isn't "flawless", I compel you to find a build that is save for pun pun or the Omniscificer.
Wizard 5/Mindbender 1/Incantatrix 10/Divine Oracle 3/Sacred Exorcist 1.
Bard 7/Abjurant Champion 2/Divine Crusader 1/Sublime Chord 1/Mystic Theurge 9.
I can think of several more builds if you want. There's actually a great Bard one with Iniate of Mili stored around somewhere.
But flawlessness aside, no. You're wrong. Because in a discussion of optimized builds, Skill Focus(Speak Language) doesn't exist.
Antimagic field. Next. Casters are strong, yes, flawless, no. Theoretically they are I suppose, but in practice they hardly stack up to flawless. Just because your a teir 1 doesn't mean you don't have a weakness. I could get into the how's and why's of this, but I don't feel like writing a thesis right now, and that would severely detract from the point of the thread (even more so than Im doing know really).
To say that the feat is bad because it doesn't kill the enemy is also a bad argument since living long enough to do anything is key in any kind of combat role. After all it's kind of hard to kill someone when you yourself are dead from their full power leap attack with the full serving of uber charger sauce. By investing in a few feats and using a reach weapon you can effectively shut down the primary damage source for most melee specialists at any time you take an action by designating them as your dodge. I honestly think you seriously undervalue the ability of this feat when used properly.
And this is how I know you aren't used to people knowing their stuff. Immediate action teleport cloak from DotU. Never leave home without it. Basic protections are something you should have with WBL, not by wasting feats.
You are wrong. It's a truly awful feat.Worse than SF(Speak Language) under most circumstances.
So your reasoning for why a feat you have all the tie is bad is a use per day magic item? That's kind of week. Not even counting that it has 3 charges per day, it's 10 feat of movement only, whats your plan for dealing with an enemy with exceptional reach like a dragon? Heaven forbid your dealing with multiple chargers. I can see it now.
"The Orc charges you and swings with his falchion"
"I teleport away as an immediate action!"
"Okay, the second orc charges you with his falchion. He deals 300 damage."
"I'm dead."
To add on top of all of that, you cant take immediate actions until you've acted in the initiative order. So if you rely on this item to get out of the way of a minotaur bearing down on you, you better hope you beat him in the turn order otherwise your going spalt. Personally the only thing I think I like about this item is the fact that it gives a deflection bonus which can be increased with the options in magic item compendium, freeing up a ring slot. Otherwise, I'd take the feat if i wanted to consistently negate power attack in a reliable fashion.
Anyway, I feel like I'm keeping this off of topic, so I think I'm done perusing this for now.
Antimagic field. Next. Casters are strong, yes, flawless, no.
Theoretically they are I suppose, but in practice they hardly stack up to flawless. Just because your a teir 1 doesn't mean you don't have a weakness. I could get into the how's and why's of this, but I don't feel like writing a thesis right now, and that would severely detract from the point of the thread (even more so than Im doing know really).
Wrong. No weaknesses is precisely waht tier 1 means.
So your reasoning for why a feat you have all the tie is bad is a use per day magic item? That's kind of week.
No. My point was such a pathetically weak feat is worthless, and if you really need it, a basic defensive item replaces it for a few thousand GP.
Not even counting that it has 3 charges per day, it's 10 feat of movement only, whats your plan for dealing with an enemy with exceptional reach like a dragon?
Because dragons always move directly adjacent to you? And because you won't have layered protection anyways? lolno.
Heaven forbid your dealing with multiple chargers. I can see it now.
Deny them a line. Also, the multiple chargers line totally invalidates your argument for Elusive Target anyways.
"The Orc charges you and swings with his falchion"
"I teleport away as an immediate action!"
"Okay, the second orc charges you with his falchion because you were idiotic enough to leave him a line. He deals 300 damage."
"I'm dead and I totally deserved it. Why didn't I pack several miss chances? Or even Ironguard?"
To add on top of all of that, you cant take immediate actions until you've acted in the initiative order. So if you rely on this item to get out of the way of a minotaur bearing down on you, you better hope you beat him in the turn order otherwise your going spalt. Personally the only thing I think I like about this item is the fact that it gives a deflection bonus which can be increased with the options in magic item compendium, freeing up a ring slot. Otherwise, I'd take the feat if i wanted to consistently negate power attack in a reliable fashion.
And if you can't beat a minotaur in initiative, you have some serious problems. With the gauntlet of I'm Going First, a Belt of Battle, and the Vagabond Sandle's you're looking at +11 before improved initiative or any other modifiers.
Wrong. No weaknesses is precisely waht tier 1 means.
Every class has a weakness - it's called the DM. Any character you build, I can build one which is identical and just, like, slap another level on it. Bang, I've got something which can beat your "flawless" class.
You made a very broad, general statement, and are now arguing in specifics. Both your builds are L20. You're talking about an item from a not-so-mainstream book which you can't even afford by the book until L5, and other items. You're then saying that you should have packed extra defenses when citing this as the sole reason you don't need Evasive Reflexes. You're even saying a tinfoil hat protects you from AMF (and if your DM is half not dead, he'll put a stop to that sort of stuff). Besides, this whole discussion has gone far from the discussion of whether you should always use PA (and your argument seems to be "I don't need to know whether I should always use PA because I'm fairly sure I should!") which was never the topic anyway.
---
In an attempt to get back on topic, I have some notes on grappling but there's seems to be loads of it. Even just today I discovered the two Aberrant feats which grant bonuses; these aren't really mechanically superior but I didn't even know they existed. Hence, a more complete compilation would be pretty handy, I think.
Every class has a weakness - it's called the DM. Any character you build, I can build one which is identical and just, like, slap another level on it. Bang, I've got something which can beat your "flawless" class.
You made a very broad, general statement, and are now arguing in specifics.
No. I was quite specific from the beginning.
Both your builds are L20.
And are perfectly functional at every level up to that, too.
You're talking about an item from a not-so-mainstream book
So?
which you can't even afford by the book until L5, and other items.
It's a good thing charging doesn't come on line until level 6 at the earliest.
You're then saying that you should have packed extra defenses when citing this as the sole reason you don't need Evasive Reflexes.
First off, you should have these defenses anyways. Second, it's not the sole reason, which you know, but are too busy trolling to acknowledge. Third, I've said nothing about evasive reflexes.
You're even saying a tinfoil hat protects you from AMF (and if your DM is half not dead, he'll put a stop to that sort of stuff).
Irrelevent and completely wrong.
Besides, this whole discussion has gone far from the discussion of whether you should always use PA (and your argument seems to be "I don't need to know whether I should always use PA because I'm fairly sure I should!") which was never the topic anyway.
In order: I don't care, that's not what I said, and yes it was on topic, which is why reading is cool and something you should learn.
In an attempt to get back on topic, I have some notes on grappling but there's seems to be loads of it. Even just today I discovered the two Aberrant feats which grant bonuses; these aren't really mechanically superior but I didn't even know they existed. Hence, a more complete compilation would be pretty handy, I think.
How is a completely worthless feat relevent at all. Should we include Skill Focus(Speak Language) in a skill bonus guide?
How is a completely worthless feat relevent at all. Should we include Skill Focus(Speak Language) in a skill bonus guide?
Dude, do you know what a compilation is? In fact, do you know how guides and handbooks work? Clearly not. So, please cease your idle comments.
Now I am quite sure you're trolling. Please take it elsewhere, we were having a fine discussion here.
No, you weren't. You were derailing a thread which I started to ask a serious question. If you want to take your discussion to your own thread, feel free, but either way, give it a break.
EDIT: link fixed
EA, before you continue insulting the people who haven't read as much as you have, could you please share your thoughts on this XKCD? It doesn't look like the OP is getting a better answer anyway.
Link is broken, FYI.
But my opinion? Like many things, it depends on context. In this case, I already provided the information. So, in the beginning, that would have been a reasonable thing to post. However several posts and a hissy fit from the OP(At least, given him bringing up rule zero, I seriously hope that's what it is) later, we're far enough past it that it doesn't matter.
Because in an online discussion of optimization, poorly optimized or unoptimized builds just don't matter.
This seems to get at the crux of your error. Irrelevance is not the same as nonexistence. Your statement is still flat-out wrong even if the counterexamples will not come up in play.
Additionally, poor options actually do have relevance to a community primarily concerned with comparing the value of various options. It will typically not be useful to enumerate all the bad options, but their existence should at least occupy a place in the back of our minds.
No, not really. Coming up with bizarre niche situations is not a valid rebuttal.
You can review the Code here: www.wizards.com/Company/About.aspx?x=wz_...
Please keep your posts polite, on-topic, and refrain from making personal attacks.You are welcome to disagree with one another but please do so respectfully and constructively.
If you wish to report a post for Code of Conduct violation, click on the Report Post button above the post and this will submit your report to the moderators on duty.
In the context of discussing a build or set of builds (here, Power Attackers), I find it very important to point out what kind of kryptonite opponents may bring to bear.
About grappling: it's fairly easy to reach very high Grappling bonuses, but the restriction on size differences can be a problem (due to typical monsters increasing in size with CR) so that magic of some form is generally needed.
And the kryptonite here is naturally Freedom of Movement, and teleporting to a lesser extent.
EA, you seem knowledgeable but I'm puzzled by your assumption that an optimized build only needs to consider specific opponents (i.e. ones that are optimized following your criterions).
I'm not sure what you mean. One should only counter that which is threatening.
In the context of discussing a build or set of builds (here, Power Attackers), I find it very important to point out what kind of kryptonite opponents may bring to bear.
But the issue is there isn't any real "kryptonite," which is a term I hate to use due to both DC's inferiority and Supe's massive Mary Sue-ness. What does Elusive Target even do? Lets you designate one target, and makes it so you might be able to survive an attack routine if you're lucky and the attacker is unprepared. And even if you manage to save yourself from one attacker, which is questionable anyways, there's nothing preventing Power Attacker B from walking up and chunky salsa-ing you. It's pathetic when it comes to protection.
About grappling: it's fairly easy to reach very high Grappling bonuses, but the restriction on size differences can be a problem (due to typical monsters increasing in size with CR) so that magic of some form is generally needed.
And the kryptonite here is naturally Freedom of Movement, and teleporting to a lesser extent.
It's far too easy to shut down and really isn't worth the risks. It's a bad deal.
1. At low levels, before I was able to qualify for easy multipliers like valorous weapons and Leap Attack, as well as easy ways to mitigate the attack loss (my build uses Shock Trooper, whereas if I was a more standard gish it would be persistent Wraithstrike or something), running the PA calculator to find the ideal number against most ACs was a pretty big deal. In these circumstances, Tempest is absolutely correct - going max PA every time does not actually result in the highest damage.
2. Level 6 was the big deal for me - I picked up an iterative attack, Shock Trooper and Leap Attack all at the same time. Now, when I was able to get Shock Trooper off, this was basically a no-brainer: I charge, PA for full with Shock Trooper and Leap Attack, and turn my enemies into a fine mist. That being said...
3. ...on instances where I could NOT get Shock Trooper off (more on that later), that PA calculator suddenly becomes super important again. See, I picked up enough other damage bonuses from my weapon that the potential damage loss from missing on my iterative attack was not worth the extra gain I got from Power Attacking. So, I had various AC benchmarks where certain PA solutions were more valuable than others.
4. I'm level 15 now, and the above has more or less remained true. If I can charge with Shock Trooper, there's no real downside to going PA for full. If I can't, I've got to pull out the calculator.
5. OK, here's a kind of big deal. Most of the major damage multipliers (Battle Jump, Leap Attack, valorous, etc.) all rely on a charge. The most common penalty mitigator, Shock Trooper, also relies on a charge. The thing is, there WILL be circumstances where you are unable to charge. If this isn't true for you, then I suspect you have either never played a charger, or you have a DM that seems to think combat should take place in a weird featureless plain.
Now, I know that my character's damage output drops considerably if I can't charge, so I've dedicated some resources to being able to move. I've got Anklets of Translocation, the Sudden Leap maneuver, and I bought wands of Benign Transposition for the party. Even with all this, in at least one quarter of all my encounters, I run into a circumstance where I can't freely charge. Maybe I'm in a tunnel with restricted movement. Maybe I'm in a room with low ceilings and difficult terrain that I can't jump or fly over. Maybe there's an enemy with lockdown techniques barring me from charging any available targets, so I have to find a different approach.
I can still deal respectable damage without a valorous leap attacking shock trooper charge, but simply PAing for full is NOT, I repeat NOT the ideal circumstance in most situations. Don't believe me? Here are some actual numbers!
Without all my charging bonuses and Shock Trooper, I'm still a character with +14 BAB and +9 Strength. My non-charging attack routine is +28/+23/+18. My weapon, a large +1 valorous wrathful healing longspear that I always have GMW on for a +4 bonus, deals 2d6+17 points of damage per hit when I'm not charging, assuming that I don't have any other enchantments on it. (I usually do, by the way - but that actually makes these figures further lean toward dedicating LESS to PA, not more).
So, let's say I'm attacking an enemy with an AC of 15. That's a LOW AC for this ECL - I can't actually remember the last time I've gone after someone with an AC lower than 24 short of touch attack shenanigans.
Against AC 15, if I PA for full (-14, as I'm one points shy of full BAB), my average damage is 125.84. If I PA for 13, my average damage is 126.5. If I PA for 12, my average damage is 126.72. -12 is the peak in this particular instance - if I drop my attack bonus to -11, I go back to 126.5 damage, etc.
OK, so even against a super low AC of 15, my average damage is actually slightly better by power attacking for 12 rather than for 14. But whatever, that's not even by a full point, so it doesn't ACTUALLY make much of a difference, right? Well, hang on, let's take my last enemy, a weird mutated giant with an AC of 24. (Again, that's one of the lowest we've faced in a long time - we've been fighting dragons lately, and they've had ridiculous ACs).
Well DANG, if I can't get my full charging off, that's actually a significant drop. Power attacking for the full 14 points is an average of 51.48 damage. That's less average damage than if I had not used PA at all - without PA, I would average 69.96. The sweet spot in this instance is -6, where I average 83.16.
And sure, this is a far cry from the 500+ points of damage I usually do when I am fully buffed and get off a complete charge with all appropriate goodies. Talking about a difference of 31 points of damage seems silly in that case. But it makes a difference, and the fact that I've played around with a good PA calculator has probably, over the course of this game, been worth thousands of points of damage and a lot of dead guys. If you just assume that PAing for full is that right decision all the time, you're either only playing on paper under ideal circumstances, or you're cheating yourself.
About grappling: it's fairly easy to reach very high Grappling bonuses, but the restriction on size differences can be a problem (due to typical monsters increasing in size with CR) so that magic of some form is generally needed.
And the kryptonite here is naturally Freedom of Movement, and teleporting to a lesser extent.
Oh yeah, grappling is I think very much not a high-level tactic, and is hardly going to be a foolproof tactic (but again, a smart DM will occasionally throw you up against something your key tactic - whatever it is - doesn't work so well against anyway). I have on occasion wanted to build one, however (mostly just for the hell of it, perhaps precisely because it's not largely considered a strong option) in which case having a few notes on the subject is handy.
Also, since it seems as though there are no decent compilations anyone knows of for these things (love the Dispelling/Counterspelling one ;)) beyond the BBC guide, I might split this off into separate topics.
www.giantitp.com/forums/showthread.php?t...
This guy has some Handbooks that are quiet good.
Hope you find this useful
1. At low levels, before I was able to qualify for easy multipliers like valorous weapons and Leap Attack, as well as easy ways to mitigate the attack loss (my build uses Shock Trooper, whereas if I was a more standard gish it would be persistent Wraithstrike or something), running the PA calculator to find the ideal number against most ACs was a pretty big deal. In these circumstances, Tempest is absolutely correct - going max PA every time does not actually result in the highest damage.
It's not quite just the low levels, but the conditions where you consider less-than-max PA are much more common there.
Specifically, those conditions are at least one of:
1) Very high enemy ACs when you lack easy ways to bypass them, or
2) Lower absolute output from using Power Attack (here, lower base attack to feed into it and fewer multipliers).
A high-level meleeist can easily bypass the first one if he values his job, and the second comes part and parcel with Power Attack Specialization - but situations still emerge where one of those conditions might show up.
The earlier example in this thread was Elusive Target's Negate Power Attack (which flat-out removes the bonus damage against your dodge target), but you don't have to go that extreme. If you're using Power Attack without ubercharging, it's not going to be hard to run into situations where the second condition might not carry enough weight, even in the mid levels.
EDIT: I just made my Power Attack calculator compatible with multipliers, add-on boosts (like Leap Attack and Supreme Power Attack) and Shock Trooper. Figured why not put it out there in public (downloadable XLS spreadsheet). It already includes critical hits and natural 1s, and accounts for DR and miss chances. It does not yet account for extra attacks at your max bonus (i.e. Haste, Whirling Frenzy) but I'll get those in there soon.
According to Optimization By The Numbers, CR 8 foes average 20 AC (and I assumed no DR or miss chances). According to the DMG and MIC's guidelines, a level 8 character can expect to have a +2 weapon and a +2 ability booster. Using the standard PC array and putting the 15 in Strength, it becomes 17 by 8th (effectively 19 with buffs). I've filled out the sample sheet with numbers assuming such a character, armed with a +1 Valorous greatsword. The character has Power Attack (assumed by using the sheet) and Leap Attack (the +100% Bonus Power Attack) in this example. We'll also assume he has the option of using Pounce, and that he's charging (see the +2 Atk+ entry). This is a pretty basic charger character.
As you can see, in this scenario, the optimal Power Attack is not max. Rather, if you're attacking once, it's -7. If you're pouncing, it's actually lower, at -2. And if you're not charging at all (and thus lack Leap Attack - set that 100 bonus power attack to 0 - nor Valorous (set the damage multiplier to 1) and the bonus accuracy from a charge - set the Atk+ to 0), you'll see the optimal power attack drops to -3 or -2! And this is against an average opponent from your CR, and we're outside the lower levels.
Of course, if you're ubercharging and have Shock Trooper, this quickly changes (max it out no matter what). But it's still pretty easy to show that without that level of accuracy, under pretty baseline benchmark assumptions (even tailored slightly for a charger) scenarios will exist where max PA is not the optimal goal.
I still stand by "Max is the best default assumption", though - it's not tricky to reach that point and most meleeists should do so (or risk doing their jobs poorly). But "Max is the best default assumption" is a far cry from "Max is the only option period".
Except you're assuming you won't have an easy way to lower an enemy's AC at the drop of a hat easily, which is obviously false.
Download the sheet and let's work that in there.
As an example of dropping enemy AC easily, I'll use my character here. I'm playing a hexblade who is NOT an ubercharger (he's got Power Attack and that's about it - his role is numeric hosing, especially saving throws and action economy). His at-will debuff arsenal is typically Dark Companion (-2 AC) and Imperious Command (Cowering foes (-2 AC and loss of Dex); I have enough passive save/check penalties and a quicker-than-usual Intimidate, so it's rare that anything not immune to fear doesn't cower when I tell them to) - I can go much deeper down the debuff rabbit hole than that, but those two don't require any limited resource expenditure (just one move action), so they're used pretty often - and another party member has Inspire Courage up to +4 at the moment (and enough songs to pull it off every encounter, so it's also a similarly always-on effect). On the sheet, that's +4 Atk+, +4 Dmg+, and the AC effectively lowers by... let's say 5 (-4 from debuffs, -1 from loss of Dex - the truncated average difference between AC and flat-footed AC at CR 8).
Plugging these values in for level 8 on a NON-ubercharger (0% bonus PA, multiplier 1, using a +2 greatsword) and we find optimal normal PA is -6 and the optimal full-attack is -5, not -8.
Plugging these values in on the Ubercharger given earlier and, without Shock Trooper, they're also -6 and -5. If he can't charge, the optimal PA falls to -6 and -4.
Exactly what methods are you referring to that give you a bigger-than-9-point swing in effective AC drop? Using the numbers from the very first example (a non-shock-trooper ubercharger), you'd have to be doing a single strike against AC 12, or a full attack against AC 9, before max PA becomes the optimal example. (EDIT: If this seems a bit odd, because of the 8-point swing to hit AC 12 being smaller than the 9-point above, it's because of the increased bonus damage from Inspire Courage meaning less PA is actually needed for the same damage output.)
Don't take my word for it. Play with the math yourself. You'll see what I mean.
(EDIT: The difference between the optimal and max results is usually pretty small - on the order of 10% unless the multipliers get insane - so if you don't know the exact AC your enemy is using, a max PA is still a good enough baseline assumption to start.)
Tempest, you didn't know? In EAtopia, every melee character has at least minimal arcane casting so that they can ALWAYS use wand-chambered greatsword with an Eternal Wand of Wraithstrike.
And the sheet I linked can check that. Assume that the arcane casting was picked up through Magical Training or something for no loss of base attack. (Also assume that it's an actual wand of wraithstrike, as an eternal wand is a wondrous item and not something that can be plugged into a chamber, but that doesn't matter here - it's someone using Wraithstrike with reasonable consistency that matters.) Average CR 8 touch AC is 11. Take the very first example, change absolutely nothing except the effective AC, and you find that you do, in fact, get optimal PA of -8... for single attacks only. If you're pouncing, it's -7. Hmm. Look at that.
Of course, on the full attack -7 vs AC 11, the expected damage is 91.35. The max PA expected damage is 90.72. This is why I keep saying that max PA is still a good assumption in almost every case - it's less than one point of expected damage difference! (The biggest differences between max and optimal that I've seen using realistic numbers from Optimization By The Numbers are about 25% lower for maxing it out, with most of them being closer to 10%. The more steps you take to negate the two conditions described above, the smaller those margins get - which is why Wraithstrike on a minor ubercharger only differs by a 1 point between optimal and max.)
Basically, this entire exercise shows you how much it costs to not focus heavily on accuracy and multiplication on Power Attack. If you don't use them, your damage can be up to ~25% lower than it would be otherwise.
Edit: Okay, now it does haste-style extra attacks. If the effect comes with a penalty to attack (Whirling Frenzy, Flurry) include that in the "Atk+" box.
EDIT: While adding that in, I noticed a slight bug in how I implemented the damage multipliers (i.e. Valorous). Fixing that causes this example to work with -8/-8, assuming you've got a multiplier from somewhere (as in the above example). It wasn't a massive change, but it adds up at high PAs and on full attacks. ALthough the expected damage is much higher (around 130), the difference between -8 and -7 is only about three points. Same basic moral as above.
Be advised that if you can't charge, this example plummets: Even with Wraithstrike, optimal PA for a single hit is still -8, but for a full attack it's -6. The difference between optimal and max full attack expected damage is 0.66 damage - less than one point. Again, max PA is not optimal, but is still a great assumption and starting point most of the time.
Like, seriously people. Marbles are, like, 1sp. If you can't hit with such basic equipment, urdoinitwrong.
Sigh...
Like, seriously people. Marbles are, like, 1sp. If you can't hit with such basic equipment, urdoinitwrong.
2 SP, actually. They also...
- ...only cover a five-foot square,
- ...are useless against anyone with ranks in Balance,
- ...are useless against Huge or larger enemies,
- ...are useless against enemies with multiple legs,
- ...are useless against enemies that fly, float, etc.
If your DM lets such basic equipment completely screw over every monster, he'sdoinitwrong.
(Also, can someone tell me what kind of an action it is to spread marbles/caltrops? I've had some DMs tell me it's a free action like dropping an item, while others have said it's a standard action because you're not just dropping them in place, you're spreading them in a 5x5 area. I've checked the PHB, A&EG and 3.5 FAQ, and I can't seem to find anything in that regard.)
(Also, can someone tell me what kind of an action it is to spread marbles/caltrops? I've had some DMs tell me it's a free action like dropping an item, while others have said it's a standard action because you're not just dropping them in place, you're spreading them in a 5x5 area. I've checked the PHB, A&EG and 3.5 FAQ, and I can't seem to find anything in that regard.)
I vaguely recall the PHB section on caltrops requiring a move action.
If so, it's not listed on either the section about the item itself on page 126, or on the "Actions in Combat" table on page 141.
A move action could make sense, since "manipulate an item" falls under move action, but they don't really clarify:
"In most cases, moving or manipulating an item is a move action. This includes retrieving or putting away a stored item, moving a heavy object, and opening a door."
None of those are really much like spreading caltrops, although I guess it's not a bad compromise for the next time I DM. It just bothers me that it's not actually spelled out somewhere, and I keep hoping that I've just missed it somehow.
Sigh...
Like, seriously people. Marbles are, like, 1sp. If you can't hit with such basic equipment, urdoinitwrong.
2 SP, actually. They also...
- ...only cover a five-foot square,
- ...are useless against anyone with ranks in Balance,
- ...are useless against Huge or larger enemies,
- ...are useless against enemies with multiple legs,
- ...are useless against enemies that fly, float, etc.
If your DM lets such basic equipment completely screw over every monster, he'sdoinitwrong.
(Also, can someone tell me what kind of an action it is to spread marbles/caltrops? I've had some DMs tell me it's a free action like dropping an item, while others have said it's a standard action because you're not just dropping them in place, you're spreading them in a 5x5 area. I've checked the PHB, A&EG and 3.5 FAQ, and I can't seem to find anything in that regard.)
Sigh...
Like, seriously people. Marbles are, like, 1sp. If you can't hit with such basic equipment, urdoinitwrong.
2 SP, actually. They also...
- ...only cover a five-foot square,
- ...are useless against anyone with ranks in Balance,
- ...are useless against Huge or larger enemies,
- ...are useless against enemies with multiple legs,
- ...are useless against enemies that fly, float, etc.
If your DM lets such basic equipment completely screw over every monster, he'sdoinitwrong.
If they can't shut down enough, urdoinitwrong. Also, if you can't figure out any other pieces of cheap gear that shut down more encounters, urreallydoinitwrong.
Plus, if you're ever seriously threatened by anything large or larger that isn't some outsider or dragon, urreallyreallyreallydoinitwrong(Also, can someone tell me what kind of an action it is to spread marbles/caltrops? I've had some DMs tell me it's a free action like dropping an item, while others have said it's a standard action because you're not just dropping them in place, you're spreading them in a 5x5 area. I've checked the PHB, A&EG and 3.5 FAQ, and I can't seem to find anything in that regard.)
Dropping an item is a free action. But if you don't have a Collar of Perpetual Attendance, you seriously... I really don't know how to describe fail like that in words...
If they can't shut down enough, urdoinitwrong. Also, if you can't figure out any other pieces of cheap gear that shut down more encounters, urreallydoinitwrong.
What pieces of gear do you have in mind that don't have similar shortcomings? Marbles are what you've suggested so far, and they don't fulfill the intended function in a variety of circumstances.
Dropping an item is a free action. But if you don't have a Collar of Perpetual Attendance, you seriously... I really don't know how to describe fail like that in words...
Dropping a bag containing some marbles isn't the same as spreading the marbles over an area, especially since you're unlikely to be carrying it in your hand in the first place.
The collar is useful, but since an unseen servant can only obey one instruction at a time, you're either banking on a significant need to specifically spread marbles during your adventuring day, or spending an action directing the servant to change its activity.
If they can't shut down enough, urdoinitwrong. Also, if you can't figure out any other pieces of cheap gear that shut down more encounters, urreallydoinitwrong.
What pieces of gear do you have in mind that don't have similar shortcomings? Marbles are what you've suggested so far, and they don't fulfill the intended function in a variety of circumstances.
(Razor) Net and lasso. Tanglefoor bags. Shapesand. Oh, and do I need to mention the Dust of Sneezing and Choking?
Look, I really could go on, but it's tiresome.
Dropping a bag containing some marbles isn't the same as spreading the marbles over an area, especially since you're unlikely to be carrying it in your hand in the first place.
Why not?
The collar is useful, but since an unseen servant can only obey one instruction at a time, you're either banking on a significant need to specifically spread marbles during your adventuring day, or spending an action directing the servant to change its activity.
So? Standing orders work fine.
There is also one for pathfinder handbooks here: www.minmaxboards.com/index.php?topic=193....
(Razor) Net and lasso. Tanglefoor bags. Shapesand. Oh, and do I need to mention the Dust of Sneezing and Choking?
The dust is the most useful of those, mainly due to the specific kind of protection required, though even there it's not without holes in its effect. For example, a creature immune to stunning, or with no need to breathe will find it no obstacle at all.
For every tactic there is a countertactic, so you simply look at how and when each option is appropriate and effective, and when it is inappropriate or ineffective.
Dropping a bag containing some marbles isn't the same as spreading the marbles over an area, especially since you're unlikely to be carrying it in your hand in the first place.
Why not?
Partly because of the bag; even if you don't have the neck tied closed to prevent some getting lost during a rough day of adventuring, simply dropping an open bag containing marbles won't spread them appropriately over a given area (some will stay in the bag, depending on how you drop it, and others will roll away in the direction the opening was pointing). Even simply dropping a handful of marbles without the bag doesn't generally do that; you might get all the marbles stopping in approximately the right space, or you might end up with marbles disappearing in all directions.
If you instead pour them out more slowly closer to ground level instead of just dropping them, you can produce a relatively even spread without too many of them getting away.
The collar is useful, but since an unseen servant can only obey one instruction at a time, you're either banking on a significant need to specifically spread marbles during your adventuring day, or spending an action directing the servant to change its activity.
So? Standing orders work fine.
Each tactic has its conditions for effective use. For example, what happens if you've ordered the unseen servant to spread marbles on your word or gesture, but run into an opponent who ignores them, but would be vulnerable to your dust of sneezing and choking?
(Razor) Net and lasso. Tanglefoor bags. Shapesand. Oh, and do I need to mention the Dust of Sneezing and Choking?
The dust is the most useful of those, mainly due to the specific kind of protection required, though even there it's not without holes in its effect. For example, a creature immune to stunning, or with no need to breathe will find it no obstacle at all.
Then use a different one. It's not complicated.
Partly because of the bag; even if you don't have the neck tied closed to prevent some getting lost during a rough day of adventuring, simply dropping an open bag containing marbles won't spread them appropriately over a given area (some will stay in the bag, depending on how you drop it, and others will roll away in the direction the opening was pointing). Even simply dropping a handful of marbles without the bag doesn't generally do that; you might get all the marbles stopping in approximately the right space, or you might end up with marbles disappearing in all directions.
Again, so? I fail to see how this is relevant.
If you instead pour them out more slowly closer to ground level instead of just dropping them, you can produce a relatively even spread without too many of them getting away.
Again, so? You keep posting this incorrect, or unverified at absolute best, claims like they actually matter.
Each tactic has its conditions for effective use. For example, what happens if you've ordered the unseen servant to spread marbles on your word or gesture, but run into an opponent who ignores them, but would be vulnerable to your dust of sneezing and choking?
Use your glove of the master strategist then. It's not that complex.
So let's take a look at Optimization By The Numbers again and use flat-footed AC (good proxy for denied-Dex) instead of standard AC.
The example from before (8th level PC with full base attack, a +1 Valorous greatsword, Leap Attack, and +4 Strength mod) vs average flat-footed AC for that CR (18) says the optimal power attack is -7 for a single strike, or -6 for a pounce. Neither is max. If the target is stunned instead of simply denied dex (add another -2 AC), the optimal single strike PA is maxed - but the optimal pounce PA is -7. And if the character isn't charging (let's give him a stock +2 greatsword here instead), then the optimal PAs vs flat-footed are -4 (single) and -3 (full attack), or vs stunned are -5 (single) and -4 (full attack). All of this is a pretty average warrior against a pretty average enemy, which isn't exactly a corner-case situation.
Check my numbers if you don't believe me. You'll need more than just saying "marbles, lol" to argue for universal max-out PA. (That said, I repeat - it's still a good place to start.)
EDIT: I've updated the sheet; it now returns how much better you're doing with optimal compared to max power attack for the given configuration (minimum 0% if optimal is maximum), and also gives you the percentage of average monster HP at CR = your base attack bonus that you're dealing in expected damage. I do this for single hits and for full attacks. These are useful evaluation tools, and I'll expand on them later. There's also an AC reporter for the mean AC, touch AC, and flat-footed AC for all foes within 4 CRs of your base attack bonus - so in other words, its the average of everything you can reasonably be expected to be fighting at any given time. They're just there as a reference.)
Several very bad assumptions and general incorrectness
C'mon, dude. You're normally far better than this.
Let's take a couple of snapshots, 'kay?
Level 6: C'mon, your strength has no business being below 25 unless you're some pansy elf. With a +1 weapon and a charge, you really are going to be hitting, even without any dirt cheap penalties. If not, spend a GP or two on a net or lolmarbles or something.
Level 11: First off, let's ignore the fact that you should have Shock Trooper by now, just to save your ego from a little extra bruising. Well, you really should have had it by level 6, but that's really beside the point. Average AC is below 25, and the max is barely above it. Now if you mean to tell me you can't jack up attack bonuses and strength, then you really should just give it up now.
16: Look, if you can't be hitting 30 or thereabout with just WBL, then... I really cannot figure out a way to accurately describe the levels of ignorance involved. I suggest you switch to 4e, or some other "balanced" not-a-game.
And, SERIOUSLY? Why the hell would any melee character only have a +4 modifier at level 8? That's just... depressing.
25 is certainly possible though: 18 base +2 racial or Rage +2 level +2 item = 24(+7), which is functionally the same as 25. In this situation, regular attacks against AC 20 are optimal with PA -3 (for a single hit, this is about a 15% improvement over just going max) and -2 (for a full attack; this is about a 35% improvement over just going max.) Throw in Leap Attack and a charge with a Valorous weapon, and the optimal PAs change to -7 (single; about 1% better than max) and -5 (for a full attack pounce; about 7% better than max). Hose the guy's AC with marbles or something and we switch to flat-footed (~AC 18), and only in this situation (charging, Leap Attack, Valorous, 24 Strength, vs marbled opponent) does the 24-Strength character find that max PA is optimal - and even then, only for single strikes (full attacks still favor -6, or 2% better than going max).
And as for Shock Trooper, I pointed out that once that's in the picture, if you're charging, max PA is always optimal. (In other words, it's a trivial case and I'm not going to contest it at all, and never have in this thread) I was looking into other situations - including non-charging PA. Your argument was that max PA was the only PA to ever consider in any situation, not just in charging.
I think my argument here - namely, that max is a reasonable place to start, but not always the most optimal choice, particularly if you aren't charging - is pretty easy to verify. I find that yours frequently comes down to things like "marbles, lol" ignoring
what impact that has on actual ACs:
cr Dex Denial Impact
1/10 -1.50
1/8 -2.00
1/6 -1.75
1/4 -1.75
1/3 -2.00
1/2 -1.31
1 -1.48
2 -1.82
3 -1.50
4 -0.95
5 -1.43
6 -1.61
7 -1.63
8 -1.68
9 -2.06
10 -1.16
11 -2.33
12 -0.33
13 -1.33
14 -1.83
15 -0.25
16 -1.18
17 -1.14
18 -1.50
19 -0.80
20 -2.00
21 -1.38
22 -3.89
23 -4.64
24 -4.63
25 -4.36
26 -4.72
27 -6.50
28 -10.25
29 -8.00
cr Dex Denial Impact
1/10 -1.50
1/8 -2.00
1/6 -1.75
1/4 -1.75
1/3 -2.00
1/2 -1.31
1 -1.48
2 -1.82
3 -1.50
4 -0.95
5 -1.43
6 -1.61
7 -1.63
8 -1.68
9 -2.06
10 -1.16
11 -2.33
12 -0.33
13 -1.33
14 -1.83
15 -0.25
16 -1.18
17 -1.14
18 -1.50
19 -0.80
20 -2.00
21 -1.38
22 -3.89
23 -4.64
24 -4.63
25 -4.36
26 -4.72
27 -6.50
28 -10.25
29 -8.00
Pro tip, by the way: Snapshots never include rage. If you don't have some racial prereq, there is never, ever ANY reason not to be an orc.
Now, if we allow for rage, since you seemed perfectly willing to, at level 6, then we have a strength 29, or +9. Do note that this is without me even trying. I feel reasonably confident that if I had the time and inclination, I could top 30 with no problem. Now, if you're missing with a +9 strength modifier then, well, that's pretty sad. Of course, it's even more sad if you don't one-shot an enemy, but that's generally beside the point.
Plus, there's one thing you're really missing. If one has Shock Trooper, there is never any excuse not to charge. This, combined with how easy it is to get Shock Trooper kinda renders the whole thing moot.
Le Link: tvtropes.org/pmwiki/pmwiki.php/Main/NoTr...
Luckily for me, I no longer get presented with his absurd babble. 
Side note: Increasing Str increases damage too, by 1.5 with a two-hander, so depending on your base damage and to-hit chance it might actually (in theory) make PA a less attractive option.
Side note: Increasing Str increases damage too, by 1.5 with a two-hander, so depending on your base damage and to-hit chance it might actually (in theory) make PA a less attractive option.
I am both far too lazy and busy to actually read through that wall of text. So here's what I've gotten from a skim:
Do you want me to break it down into bite-sized digests, o erudite ape?
Pro tip, by the way: Snapshots never include rage. If you don't have some racial prereq, there is never, ever ANY reason not to be an orc.
The bonus feat on human and near-human isn't a racial prereq but is often damn good reason to play such a character. And occasionally you'll find kalashtar or azurin psychic warriors for their other abilities. But I'll concur here - I had forgotten Orc was +4 (for some reason I thought it was +2), and all of the other +4s I know of have level adjustments.
Now, if we allow for rage, since you seemed perfectly willing to, at level 6, then we have a strength 29, or +9. Do note that this is without me even trying. I feel reasonably confident that if I had the time and inclination, I could top 30 with no problem. Now, if you're missing with a +9 strength modifier then, well, that's pretty sad.
I "allowed for rage" because I thought orc was +2, not +4. I needed to make up a +6 gap in Strength to match your numbers, and the two options that led to that were +4 racial / +2 item (the one used here, but one I couldn't see working due to how I misremembered orc) and +2 racial/item +4 rage. But I'll run with it as you did, and I'll show my work.
You're going with level 6, so let's check this.
- Let's assume you have +9 Strength at level 6. (18 base +4 orc +1 level + 2 item* +4 rage = 29.)
- I say "assume" because the +2 gloves of strength are budgeted as a level 8 item in the MIC. You CAN get it at level 6, but it's rather pricey, particularly because melee has a lot of gold taxes. But we'll grant it, because it IS possible.
- Assuming a +1 greatsword at the time, your attack bonus is +16, or +18 on a charge.
- For context, the average CR 6 ACs are: Normal 18, Flat-Footed 17, Touch 11, Hosed 9.
- We will ignore Shock Trooper for here - if you have it, and you're charging, you'll always PA for max, so the question is less interesting.
- If you do not have Leap Attack and you charge, the optimal Power Attack is -5. If you have Pounce, it's -3.
- Without Leap Attack, you would need to be delivering a single charge attack against AC 16, or a pounce against AC 13, to have max PA be optimal.
- However, against the usual AC, your numbers are, respectively, only +0.75% and +3.5% greater than max PA.
- If you do have Leap Attack and you charge, the optimal Power Attack is -6 in both cases.
- If you are unable to charge, the optimal PA is -4, or -2 for a full attack.
- You would need to be delivering a single strike against AC 14, or a full attack against AC 11, for max PA to be optimal.
- However, against the usual AC, your numbers are, respectively, only +2.74% and +8.17% greater than max PA.
Does THAT make it easy enough to understand?
My argument is, and always has been, that max PA is a good starting point, but that it will not always be the most optimal solution. You go to extreme examples and belittle me - and then I demonstrate that, without Shock Trooper, my argument still holds in your extreme cases.
The strongest example above is a Pounce with Leap Attack (optimal PA = max). Let's be generous and say you've also managed to deny him his Dex bonus through one of the tricks mentioned earlier, so his AC is only 17. The expected damage from this is 62.56 - or 90.51% of the average CR 6 HP. Last I checked, 91% was not 100%, so that's not a one-shot. However, I admit this is splitting hairs, since you're almost never fighting alone (and if you can't kill him but can get him into cherry-tap range for someone else, that's probably good enough).
My argument is, and always has been, that max PA is a good starting point, but that it will not always be the most optimal solution. You go to extreme examples and belittle me - and then I demonstrate that, without Shock Trooper, my argument still holds in your extreme cases.
Of course, it's even more sad if you don't one-shot an enemy, but that's generally beside the point.
The strongest example above is a Pounce with Leap Attack (optimal PA = max). Let's be generous and say you've also managed to deny him his Dex bonus through one of the tricks mentioned earlier, so his AC is only 17. The expected damage from this is 62.56 - or 90.51% of the average CR 6 HP. Last I checked, 91% was not 100%, so that's not a one-shot. However, I admit this is splitting hairs, since you're almost never fighting alone (and if you can't kill him but can get him into cherry-tap range for someone else, that's probably good enough).
I find this argument hilarious. EA seems to be taking the stance that an opponent won't be immune to stunning, or that they can't fly. Or that there's never going to be any obstructions on a battlefield, even an aerial one. I belive the problem here is he's trying to pull a"No True Scotsman" rather than recanting his earlier statement and is getting pretty stupid sounding because of it.
It looks to me like you aren't very good at reading.
Y'see, any decent opponent that is both flying and immune to stun is probably a NPC with decently-optimized WBL. If that's the case, its AC will be garbage, because AC isn't wirth investing in.
Also, obstructions? Lolswiftactionteleport. Dirt cheap.
So, really, it would be you who's coming off as pretty damn stupid.
Side note: Increasing Str increases damage too, by 1.5 with a two-hander, so depending on your base damage and to-hit chance it might actually (in theory) make PA a less attractive option.
Go ahead and play with the numbers to see how much you'd need. In my experience the only thing that makes PA less attractive is +Damage that doesn't come with a roughly equal amount of +Attack, and even then the gain is minimal (or negative, if you've got sufficient PA enhancers).
Also, obstructions? Lolswiftactionteleport. Dirt cheap.
Only works for things like soft cover and difficult terrain (and even then only if some clear path exists in the area - while this isn't guaranteed, I know proper wealth or character options can easily ignore this, though). A mere corner, curtain or smokestick (or Silent Image, or Wall of Gloom, etc) won't be bypassed by a swift-action teleport, because:
If you don’t have line of sight to the opponent at the start of your turn, you can’t charge that opponent.
Blocking line of sight is one of the easiest ways to stop charging flat-out.
I am both far too lazy and busy to actually read through that wall of text. So here's what I've gotten from a skim:
Do you want me to break it down into bite-sized digests, o erudite ape?
...And you mean to tell me the "hot air" jokes never occured to you?
So, sure, whatever.
I "allowed for rage" because I thought orc was +2, not +4. I needed to make up a +6 gap in Strength to match your numbers, and the two options that led to that were +4 racial / +2 item (the one used here, but one I couldn't see working due to how I misremembered orc) and +2 racial/item +4 rage. But I'll run with it as you did, and I'll show my work.
You're going with level 6, so let's check this.
- Let's assume you have +9 Strength at level 6. (18 base +4 orc +1 level + 2 item* +4 rage = 29.)
- I say "assume" because the +2 gloves of strength are budgeted as a level 8 item in the MIC. You CAN get it at level 6, but it's rather pricey, particularly because melee has a lot of gold taxes. But we'll grant it, because it IS possible.
- Assuming a +1 greatsword at the time, your attack bonus is +16, or +18 on a charge.
- For context, the average CR 6 ACs are: Normal 18, Flat-Footed 17, Touch 11, Hosed 9.
- We will ignore Shock Trooper for here - if you have it, and you're charging, you'll always PA for max, so the question is less interesting.
- If you do not have Leap Attack and you charge, the optimal Power Attack is -5. If you have Pounce, it's -3.
- Without Leap Attack, you would need to be delivering a single charge attack against AC 16, or a pounce against AC 13, to have max PA be optimal.
- However, against the usual AC, your numbers are, respectively, only +0.75% and +3.5% greater than max PA.
- If you do have Leap Attack and you charge, the optimal Power Attack is -6 in both cases.
- If you are unable to charge, the optimal PA is -4, or -2 for a full attack.
- You would need to be delivering a single strike against AC 14, or a full attack against AC 11, for max PA to be optimal.
- However, against the usual AC, your numbers are, respectively, only +2.74% and +8.17% greater than max PA.
Does THAT make it easy enough to understand?
You seem to be under the absurd notion that length and complexity have anything to do with each other.My argument is, and always has been, that max PA is a good starting point, but that it will not always be the most optimal solution. You go to extreme examples and belittle me - and then I demonstrate that, without Shock Trooper, my argument still holds in your extreme cases.
Except it really doesn't. I'll explain below.The strongest example above is a Pounce with Leap Attack (optimal PA = max). Let's be generous and say you've also managed to deny him his Dex bonus through one of the tricks mentioned earlier, so his AC is only 17. The expected damage from this is 62.56 - or 90.51% of the average CR 6 HP. Last I checked, 91% was not 100%, so that's not a one-shot. However, I admit this is splitting hairs, since you're almost never fighting alone (and if you can't kill him but can get him into cherry-tap range for someone else, that's probably good enough).
Look, I've been trying to be both completely civil and not condescending, but you're seriously making this difficult. Choosing Leap Attack over Headlong Rush? The only conceivable response is "what."
Seriously, even with the very basic options, your attack should be 90((7+1+10+12)x3), which will one-shot all but the most extreme for CR enemies. And that's without dipping into Rage or the Bands of Blood Rage, or really any of the other common and easy damage boosters.
EA, can you explain how someone is supposed to do 8100 HP damage at level 6? That part has me confused. Partly the base mods (7+1+10+12), but mostly... How do you get a *90 multiplier?
That's not a multiplier, it's the breakdown.
7 Greatsword
1 Enhancement
10 Strength (18 base +4 racial +2 item = 24 (+7), Two Handed: 7*1.5 = 7+3.5 = 10)
12 Power Attack (with nothing specific modifying it)
x3 (Headlong Rush plus one other multiplier, although I'm not entirely sure which he's using)
---
30x3 = 90. This isn't weighted for chance to hit, though.
The reason I didn't use Headlong Rush in the above example is, again, that I forgot about it. I'm fighting off a cold and far from my top mental game. Most of the time I game in Eberron, and we disallow both Faerun-exclusives and 3.0 material except in rare circumstances. (Headlong Rush is from Races of Faerun and was not officially updated, although some other aspects of the book clearly show 3.5's then-upcoming influence: Jotunbrud is pretty obviously a Powerful Build ability, for instance. Then again, it's also the book with Southern Magician in it...). I think it's understandable if I forgot.
Regarding the other multiplier, I'm again not sure what he's doing for it. Valorous is one of the simpler ones, but +2 Gauntlets of Ogre Power and a +1 Valorous greatsword will run you 12000gp. WBL for 6th is 13000. Somehow EA is managing to pull off immediate-action teleports with great frequency using only 1000gp in total gear outside of his sword and gauntlets. More likely, he knows of a different multiplier at work here. Most such multipliers are charge-based, though, so I think "double damage on a charge" is a fair assumption for what it does.
If that's the case, then here's what different Power Attacks look like for this character (employing Headlong Rush and assuming that second multiplier is also charge-based), against a CR 6 foe (AC 18, 75 average HP for context).
- Charge Attacks:
- Single attack (no Pounce)
- Power Attack for -6: 56.73 expected damage.
- Power Attack for -4: 57.51 expected damage. This is the optimal Power Attack.
- Full attack (Pounce)
- Power Attack for -6: 89.82 expected damage.
- Power Attack for -3: 94.94 expected damage. This is the optimal Power Attack.
- Unable to Charge:
- Single attack:
- Power Attack for -6: 16.78 expected damage.
- Power Attack for -3: 17.52 expected damage. This is the optimal Power Attack.
- Full attack:
- Power Attack for -6: 25.16 expected damage.
- Power Attack for -2: 28.46 expected damage. This is the optimal Power Attack.
If you use something cheap (like marbles) to deny the target their Dexterity bonus (so AC 17 instead of AC 18 here), the optimal power attacks change, but in every situation they remain superior to full Power Attack. In fact, in those four situations, the amount of penalties you'd need to inflict on AC 18 to get max PA to be the best choice are: Single Charge -3, Pounce -5, Single Attack -5, Full Attack -7. If you can't get AC reduction effects that high, then you will find that max PA is not your optimal choice.
(You'll also notice that EA claimed that 90 damage "will one-shot all but the most extreme for CR enemies". Which it will - assuming it hits! Factoring in accuracy (including natural 1s and critical hits), the only way that this character will one-shot an average CR 6 foe is with a pounce, and he won't actually break 90 expected damage on that unless he doesn't Power Attack for max!)
My argument is, and always has been, that max PA is a good starting point, but it's easy to find situations where lower Power Attack values will produce better results. (I have also repeatedly said that such results may only be marginally better - as the above shows - but they are still better. That's kind of the definition of "optimal" in this case.) Each time EA has tried to shout me down by throwing out an example out that doesn't involve Shock Trooper (which I have repeatedly admitted makes maxing your best strategy, bar none, assuming you can charge), looking at the numbers shows that his examples are the very examples I am talking about, thus supporting my side of the argument.
If you consider it opaque how I'm getting the numbers,
here's how I'm doing it.
I'm determining total attack bonus and base damage exactly as EA is here.
I'm determining what number the d20 has to roll to make each attack hit, and converting that to a percentage.
"Expected damage" for a particular attack is simply the average damage of that attack multiplied by its chance to connect.
I do account for natural 1s (I don't allow the hit% chance to climb above 95).
I account for critical hits like this: All die results that lead to a hit but not a threat constitute the "normal damage chance". All die results that lead to a threat leads to a second operation: I multiply the chance for a threat by the chance for a hit, and get the "critical damage chance", then I take the residual chance of a threat (i.e. the chance for a threat * the chance for a miss, or the chance for a threat * (1-the chance for a hit)) and add it to the "normal damage chance".
Expected Damage is then just (Normal Damage Chance * average damage) + (Critical Damage Chance * critical hit damage.)
The sheet also handles other things that affect Expected Damage (miss chances, DR, flat modifiers to attack and damage, increased chances to confirm critical hits, bonus damage that doesn't multiply (i.e. Sneak Attack), and so on). These are pretty straightforward.
Finally, it repeats this for every legal value for Power Attack available to you, and accounts for things that modify PA (such as handedness, Leap Attack, Shock Trooper, and so on) along with all applicable multipliers.
I then simply look at which PA result produces the highest expected damage (either for the first hit on its own, or for the entire full attack / pounce routine), and report what I work out.
For the above example, here's what I mean. Let's look at the Charge (Single attack) abilitiy.
-Attack +16 (includes +2 from charging) vs AC 18: 90% base chance to hit.
-19-20 threat range: 10% chance for a threat. This is broken down into 10%*90% chance of a critical hit, and 10%*10% chance of dealing normal damage.
-This gives us a 10% chance of missing (natural 1-2), an 81% chance to deal normal damage, and a 9% chance to confirm a critical hit.
-At 0 PA, expected damage is 7+1Enhancement+10(Str*1.5)+0(PA) = 18.5 base damage. You have an x2 critical multiplier.
-Headlong Rush triples your base damage, so due to multiplier stacking that amounts to "normal" damage being triple your base, and "critical hit" damage being quadruple your base.
-Expected damage is therefore 10%*0 + 81%*(18.5*3) + 9%*(18.5*4) = 51.62.
If you PA for the optimal amount, -4, and repeat the above math, you'll get a 30% chance of missing, a 63% chance of dealing normal damage, and a 7% chance for a critical hit. Your base damage is 26.5. Your expected damage is therefore 30%*0 + 63%*(26.5*3) + 7%*(26.5*4) = 57.51.
I'm determining what number the d20 has to roll to make each attack hit, and converting that to a percentage.
"Expected damage" for a particular attack is simply the average damage of that attack multiplied by its chance to connect.
I do account for natural 1s (I don't allow the hit% chance to climb above 95).
I account for critical hits like this: All die results that lead to a hit but not a threat constitute the "normal damage chance". All die results that lead to a threat leads to a second operation: I multiply the chance for a threat by the chance for a hit, and get the "critical damage chance", then I take the residual chance of a threat (i.e. the chance for a threat * the chance for a miss, or the chance for a threat * (1-the chance for a hit)) and add it to the "normal damage chance".
Expected Damage is then just (Normal Damage Chance * average damage) + (Critical Damage Chance * critical hit damage.)
The sheet also handles other things that affect Expected Damage (miss chances, DR, flat modifiers to attack and damage, increased chances to confirm critical hits, bonus damage that doesn't multiply (i.e. Sneak Attack), and so on). These are pretty straightforward.
Finally, it repeats this for every legal value for Power Attack available to you, and accounts for things that modify PA (such as handedness, Leap Attack, Shock Trooper, and so on) along with all applicable multipliers.
I then simply look at which PA result produces the highest expected damage (either for the first hit on its own, or for the entire full attack / pounce routine), and report what I work out.
For the above example, here's what I mean. Let's look at the Charge (Single attack) abilitiy.
-Attack +16 (includes +2 from charging) vs AC 18: 90% base chance to hit.
-19-20 threat range: 10% chance for a threat. This is broken down into 10%*90% chance of a critical hit, and 10%*10% chance of dealing normal damage.
-This gives us a 10% chance of missing (natural 1-2), an 81% chance to deal normal damage, and a 9% chance to confirm a critical hit.
-At 0 PA, expected damage is 7+1Enhancement+10(Str*1.5)+0(PA) = 18.5 base damage. You have an x2 critical multiplier.
-Headlong Rush triples your base damage, so due to multiplier stacking that amounts to "normal" damage being triple your base, and "critical hit" damage being quadruple your base.
-Expected damage is therefore 10%*0 + 81%*(18.5*3) + 9%*(18.5*4) = 51.62.
If you PA for the optimal amount, -4, and repeat the above math, you'll get a 30% chance of missing, a 63% chance of dealing normal damage, and a 7% chance for a critical hit. Your base damage is 26.5. Your expected damage is therefore 30%*0 + 63%*(26.5*3) + 7%*(26.5*4) = 57.51.
EA, can you explain how someone is supposed to do 8100 HP damage at level 6? That part has me confused. Partly the base mods (7+1+10+12), but mostly... How do you get a *90 multiplier?
Typo. Supposed to be a space there.
Though 8100 is difficult by level 6, but by level 11 it's pathetically easy.
Anyone notice how he keeps on spouting about shock trooper even though it's clearly been established that it's the only thing that even makes his assertion half way correct (although half way to correct considering his initial argument is still wrong), and the math shows that without it he is completely wrong? The initial argument was that max power attack is always the optimal choice, not "max power attack is always the optimal choice because I have every necessary fringe item needed to always be in position for a charge all the time, plus a +1 great sword and a +2 strength item. Oh ya, by the way levels 1-5 don't count because you can't take shock trooper and afford all of this stuff till level 6 (if even)".
It's like you're seriously incapable on anything resembling higher brain functions.
Tempest, you're making a serious miscalculation. One a lot of smart guys make, which I have not been able to understand.
Look, you're seriously overvaluing hitting, when, when the only thing that matters is instasplatting. Any attack that does under... eighty, if I remember correctly, is completely worthless. Unless you memorize the enitre by the numbers table, and completely memorize every single description, and can completely predict any NPC or DM monster overhaul, more damage is always, completely and absolutely, better. A mostly dead enemy can still kill you.
All of these have the same expected damage. Since they're all strong enough to kill the monster, you'd go with reliability over raw damage here.
Now, what if they did 45 damage 100% of the time, 90 damage 50% of the time, and 4500 damage 1% of the time? Only the last is an instasplat, and that's so uncommon I wouldn't wager on it. I'd take #2, personally, unless I knew my teammates could deal 5 damage to it before its next turn without costing us the battle as a whole.
You see, max damage (which is a different issue than optimal power attack, mind you) is only optimal in certain contexts. You need two other pieces of information to set the context: how reliable the attack is, and how much impact it'll have relative to your target.
Tsuyoshikentsu said: having 300 when you need 100 once is not optimal. Having 100 every time you need 100 is.
If you instead pour them out more slowly closer to ground level instead of just dropping them, you can produce a relatively even spread without too many of them getting away.
Again, so? You keep posting this incorrect, or unverified at absolute best, claims like they actually matter.
You asked me about why a bag of marbles wouldn't just spread itself out as a result of being dropped, and this is simplythe logic of it.
If you want the mechanical information, since there's no description of how long it takes to cover a 5-foot area of ground with marbles, you can go with logic, your DM's judgement, or possibly a standard action if you can argue that as a default.
Each tactic has its conditions for effective use. For example, what happens if you've ordered the unseen servant to spread marbles on your word or gesture, but run into an opponent who ignores them, but would be vulnerable to your dust of sneezing and choking?
Use your glove of the master strategist then. It's not that complex.
The glove of the master strategist would be uncommon, since it's an item from Ghostwalk and takes up a hand slot.
Uh, dood? If you don't hit, you can't do damage. Logic 101, brosef.
Yes, but you can't kill an enemy without, you know, killing them.
You're fighting a monster in a CR band averaging 50 HP. Would you rather have an attack that does 50 damage 100% of the time, 100 damage 50% of the time, or 5000 damage 1% of the time? All of these have the same expected damage. Since they're all strong enough to kill the monster, you'd go with reliability over raw damage here. Now, what if they did 45 damage 100% of the time, 90 damage 50% of the time, and 4500 damage 1% of the time? Only the last is an instasplat, and that's so uncommon I wouldn't wager on it. I'd take #2, personally, unless I knew my teammates could deal 5 damage to it before its next turn without costing us the battle as a whole. You see, max damage (which is a different issue than optimal power attack, mind you) is only optimal in certain contexts. You need two other pieces of information to set the context: how reliable the attack is, and how much impact it'll have relative to your target. Tsuyoshikentsu said: having 300 when you need 100 once is not optimal. Having 100 every time you need 100 is.
HOLY NOT WHAT'S GOING ON, BATMAN!
Making up imaginary statistics and situations aren't cool. Especially since they don't match the discussion. A better one would be an attack that does 50 with an 80% chance of hitting, or a 100% chance of doing 45 against an enemy with 50 HP. The first option is by far better.
You asked me about why a bag of marbles wouldn't just spread itself out as a result of being dropped, and this is simplythe logic of it.
And where in the rules does it say that?
The glove of the master strategist would be uncommon, since it's an item from Ghostwalk and takes up a hand slot.
You mean a good and cheap item that everyone knows about? lolkay.
Making up imaginary statistics and situations aren't cool. Especially since they don't match the discussion.
I merely wanted to demonstrate that the rubric for selecting attacks cannot just be "more damage = more better". You have to consider accuracy, and your opponent. I used hyperbole there, I admit, but the point holds. In fact, you do as well, because you factor in accuracy and monster HP when you say:
A better one would be an attack that does 50 with an 80% chance of hitting, or a 100% chance of doing 45 against an enemy with 50 HP. The first option is by far better.
I concur!
And, oddly, I do see where you're coming from, I think. And I think I know why you view my perspective as a mistake.
I'm attempting to maximize the expected damage each turn. Mathematically (assuming you express everything related to killing the monster in terms of damage - run with me here) this means that the number of turns before the monster falls will be minimized.
You are attempting to end the monster in one hit, (almost) regardless of the accuracy of that hit.
Oddly, these are NOT at cross purposes to each other - presuming you're capable of getting the expected damage high enough.
(Simply put, it's the statistician's joke: We average 1.5 children per household, so say hello to our son and our half-a-son! In this case, damage is either dealt or not dealt on any individual attack roll; expected damage is a long-term over-time thing. One is tactical, the other strategic.)
Both of our positions agree: the utility of damage that removes the monster's last hit point is astronomically larger than the utility of any other amount of damage, even if it brings the monster to 1 HP from full. I wonder if we can actually quantify this somehow, and figure out not just "expected damage" but rather the "utility of the attack". An attack that deals insano-damage but doesn't kill the target is less useful than a cherry-tap that does kill him, but a cherry-tap that doesn't kill him is worse than an insano-damage that doesn't kill him.
But I'm getting ahead of myself.
Back on the initial point, since this has drifted a bit: If you are aiming for reliable damage, max PA is a good place to start, but there will be places where a lower PA will result in a higher expected damage. Consider your target (both accuracy in hitting the target and the target's current HP) when deciding which is more optimal.
In your above example (STR+7 with gear, BAB+6, +1 Valorous Greatsword, Headlong Rush, vs AC 18), max PA will deal 90 damage per hit at 60% and 35% hit rates, working out to 89.82 expected damage. The biggest PA is -3, which will deal 73 damage per hit at 75% and 50% hit rates, working out to 94.94 expected damage. Let's consider this in the context of four possible HP configurations.
-Average HP at CR 6 is about 70. In this case, any single hit from either of those results will kill him, meaning full PA has a 74% chance of killing him, while PA-3 has a 87.5% chance of killing him. Against this monster, PA-3 is optimal.
-If the monster had 90 HP instead, then a single hit at full PA will kill him. While the expected damage from PA-3 will kill him, any individual non-critical hit will not; both need to hit. This works out to a 74% chance of instasplat for full PA, or a 37.5% chance (plus some extra for the odds of a critical hit, I'm not working that in here) of instasplat for the lower PA. Against this monster, full PA is optimal.
-If the monster had the maximum CR 6 HP, 113, then neither group's expected damage is high enough to kill him. Two hits from either setting will do it, though. This is a 21% chance at full PA and a 37.5% chance for PA-3 (again, not accounting for critical hits). Agains this monster, PA-3 is optimal.
-Now let's say the monster has some temporary HP / DR / Share Pain / whatever buff up and needs, oh, let's say, 150 damage to be dealt to him to die. In this case, PA-3 can't deal enough damage on its own (without critical hits, which I'm ignoring for the purposes of this illustration) with both hits to kill him at all (as two hits will deal 146 damage). However, two hits from full PA will kill him - and this happens 21% of the time. Assumign you have to be the one to kill him, then against this monster, full PA is optimal. (I say "assuming" because if PA-3 can get him down to 4hp - which he will do 37.5% of the time - and another teammate acts before the monster does, the teammate can certainly (say, 95% of the time) deal the 4 remaining damage to him to kill him. This works out to a 35.625% chance of killing the monster before his turn, compared to the 21% of the full PA route. However, If the monster acts before your ally does, this is still a failure. Also, if your teammate's actions are better needed elsewhere - say the monster wasn't the last one standing - then this is also a less desirable strategy.)
And what if the monster's already critically wounded (single digits)? The optimal Power Attack there is 0 - any hit will kill it, even without the PA bonus, so the most accurate attack is the one with the greatest chance of killing him.
I hope that shows how our positions can be reconciled. It's all about context - and that context does not include just raw damage, but also accuracy and the monster's current status (and in some cases, the initiative order).
TLDR: On the long term (as you're building the character), expected damage is the statistic to maximize. Round-by-round, you should pick based off of kill rate.
You asked me about why a bag of marbles wouldn't just spread itself out as a result of being dropped, and this is simplythe logic of it.
And where in the rules does it say that?
Nowhere. According to the rules, there isn't any existing way to cover the area with marbles, just a consequence if you do it.
Since the rules don't explicitly permit you to cover an area with marbles, the real-world behaviour of marbles is the next port of call.
The glove of the master strategist would be uncommon, since it's an item from Ghostwalk and takes up a hand slot.
You mean a good and cheap item that everyone knows about? lolkay.
It's an item from a campaign book that many people haven't heard of, let alone have access to a copy.
It's not like everyone will have one.
Interestingly, if you aren't able to kill the target with a plausible number of attacks (i.e. if you deal 20 damage at a particular PA per hit, and the monster has 100 hit points remaining, and you only can attack four times, then you can't plausibly kill the monster this round at that PA regardless), then it appears to be in your best interests to use the "optimal" PA as I was describing it earlier, which maximizes expected damage. After the monster's current HP drop low enough to plausibly be killed by at least one attack routine, then this sheet will figure out the odds of that attack routine succeeding at each applicable PA, and then respond with the PA that maximizes your chances of killing the targets.
I say this is interesting in that it, too, isn't always maxed out. The sheet there is pre-loaded with the same level 6 example before (nothing special, +6 base attack, +1 greatsword, +6 Strength modifier), but now auto-loads average HP and AC from monsters within +/- 4 CRs of your BAB. (That way, it's the average of the mook, the challenge, and the boss. I haven't adjusted this to a weighted average yet though.) Against such a monster having suffered 50 damage (leaving it with 25 HP), the "Optimal" PAs from before are -3 (15.18) for a single strike and -2 (24.26) for a full attack. The kill rates, however, are optimized at PA -4 for a single strike (55%) and at 0 for a full attack (87.5%). While a full Power Attack would have worked as well, in this scenario it has a lower chance to kill (in a single strike, 45%, and as a full attack, 45%).
Why is max PA not ideal in this circumstance? Because a single hit at PA-4 is enough to incapacitate the monster. PA-4 is more accurate than PA-6. Likewise with a full attack.
There are situations where max PA produces higher kill rates. Specifically, here's how that monster (AC 22, 75 max) responds with assorted PAs, optimized for kill rate. As you might guess, the weaker the monster is, the lower the optimal Power Attack is to maximize your kill rate. The stronger the monster is the better your kill rate PA goes - but only to a point. In this case, if the monster has taken less than 17 damage, you have no chance of actually killing him even if all your attacks hit (your kill rate is 0 here - even though it goes way up after he takes that much damage).
My compromise appears to be this: If you can kill your target in one hit/fullattack/whatever, DO IT and take the PA optimized for kill rate. (Frequently this will be a high PA, as you're more likely to see partially-weakened monsters than seriously-weakend ones with lower PAs needed to kill). However, if you can't kill it up front, your optimial PA is the one that maximizes your expected damage. This provides the most reliable decrease in HP you can find - which is exactly what you are looking for when you're trying to get the monsters weak enough to kill. Switch from optimal expected damage to optimal kill rate as soon as that kill rate is nonzero.
(Now we just need peer review and publication ;))
Thanks for your work: this is quite enlightening.
(Now we just need peer review and publication ;))
I'm working on that. Looking forward to some feedback on the review front too. I'm positive I need to handle it a bit better (currently I do a first-order approximation for miss chance where, say, a 20% miss chance reduces the basic damage roll to 80%; now that I'm thinking in terms of binary hits I will need to go back into the code and multiply each individual attack roll hit chance by (1-miss chance). Not to mention stuff like multiplications and bonus attacks aren't worked in on the kill rate yet), and there's a lot that's messy (particularly because I did a fast-pass approach on combining attack probabilities rather than a smoother one - i.e. I really should have added together the probabilities of all the different ways the monster SURVIVES and then subtract that from 1) but I did make it clear this is preliminary.
One thing that did occur to me is just how much extra information you need to not only compute a kill rate, but to decide if a kill rate can possibly be nonzero. You need a hit point estimation, both current and max (including temporary HP), on top of all the information you need to figure out expected damage (which is largely a function of attack bonus, damage bonus, critical hit properties*, monster AC (and miss rate / DR estimates), and your range on Power Attack. Keeping that in mind, I think I have pretty strong justification for recommending expected damage optimization over kill rate optimization, particularly at the start of the battle.
I still suspect that there will be plenty of cases where optimal kill rate Power Attacks are not maxed out. We know that's the case for expected damage. For kill rate, if any result other than max PA is capable of killing the target, then you are open to the possibility that such a lower PA will have a higher chance of hitting. (Toy example with only one attack, for simplicity: If you can Power Attack for -10 (say, an 85% hit rate) and deal 60 damage, or power attack for -9 (say, a 90% hit rate)and deal 54 damage, and the monster has 53 hit points remaining, then it is optimal to Power Attack for -9, not -10.) (Note: While this looks like the lowest possible PA that can kill the target is optimal, that is only true for single strikes and if the monster doesn't just ignore it. Add in multiple attacks and a few other things and it gets a bit foggier. However, "least possible PA that can still kill the target will have a higher chance of hitting than max PA" is an important fact here - as is "If the monster isn't hurt enough to be killed in one turn, then the optimal PA maximizes expected damage" argument.)
*
Oh, balls, that's something else I'm going to need to work in sooner or later - fortification / immunity to crits. I'm not sure if I'll bother extending this to precision damage as well; as it stands I can probably do it by altering regular damage results (reduce the chance of critical hits by the chance of fortitication; on immunes this will mean no critical hit damage). Then again, most people interested in Power Attack aren't interested in Skirmish or Sneak Attack, so it might just be an audience thing.
Speaking of added features, it's vestigial on significant targets but might be important on tough regulars - I can include death from massive damage on my kill rate calculations.
Speaking of added features, it's vestigial on significant targets but might be important on tough regulars - I can include death from massive damage on my kill rate calculations.
(Obviously this has the lowest priority ever. The thought just popped into my head.)
Well, if you think you're not working enough on this, you could also do a PA optimizer under uncertainty: i.e. you know the enemy's AC is between 22 and 26, but not anything more. Assume every figure is as likely, and...
(Obviously this has the lowest priority ever. The thought just popped into my head.)
My PhD thesis is about the mechanisms behind Bayesian reasoning, i.e. reasoning under uncertainty, measuring degrees of confidence in beliefs, and updating how confident you are in a belief based on newly observed evidence. The military used methods based on this to aim artillery for ages (updating to account for maximum likelihood of their next shell hitting, given what they saw happen to the last one), so there's precedent. But at this point you're not so much working on a Power Attack calculator as an autopilot.
(I did just modify it a bit, but that's for our private game. Revenant Blade + massive TWF with double-handed Power Attack and Blood in the Water = this is kind of interesting to work with (i.e. if you're at 0 crits with Blood, and the marginal gain between an optimal PA full attack and a zero-PA max-accuracy full attack is minor, it is in your better long term interest to PA for less than optimal here, because once Blood charges up, it can steamroll things pretty quickly, and it charges up faster when you're hitting more often). Kill rate optimization took a backseat for a moment.)
You asked me about why a bag of marbles wouldn't just spread itself out as a result of being dropped, and this is simplythe logic of it.
Actually, I would think that dropping marbles from a bag being a free action is not too much of a stretch. You would arguably need to open the bag (making it a move action?) because if it wasn't open, they'd all come out in your pack, but as I said, not too much of a stretch. The only problem then is that even if they do spread out, well done, you just marbled your own square. There's also the question of what happens if the ground is soft, e.g. dirt or grass. I would think the marbles wouldn't actually work at all, as stepping on one would just push it into the ground, but even if they did they wouldn't roll or spread by themselves.
You're fighting a monster in a CR band averaging 50 HP. Would you rather have an attack that does 50 damage 100% of the time, 100 damage 50% of the time, or 5000 damage 1% of the time? All of these have the same expected damage.
Well, that's the thing - the Law of Large Numbers only applies when the numbers are, well, large. If downing a guy quickly is important, doing any kind of damage is better (given that you have a team to finish him off) than swinging and missing. I don't think it's enough to consider just the expectation, but also the variance.
Also, the standard monster stats by CR are immensely useful for this sort of thing, but don't paint the whole story. Full PA might even be the best tack to use for the average monster, but when a high-than-normal-AC one (or an NPC or the like) comes along the game changes. It's ridiculously easy to pump a creature's AC so assuming you can always hit is short-sighted to say the least.
Thanks for your work: this is quite enlightening.
(Now we just need peer review and publication ;))
Until I see it written up in LaTeX, I refuse!
Well, if you think you're not working enough on this, you could also do a PA optimizer under uncertainty: i.e. you know the enemy's AC is between 22 and 26, but not anything more. Assume every figure is as likely, and...
Hmm, couldn't you just calculate across the range - e.g. if -X PA is optimal for all values, go with that? It's whether you want to optimise PA for the best average damage across the range rather than just take average (of one sort or another) optimal PA across it, I guess, but given you're guessing anyway you could do so decently without too much trouble either way, I'd have thought.
Actually, I would think that dropping marbles from a bag being a free action is not too much of a stretch. You would arguably need to open the bag (making it a move action?) because if it wasn't open, they'd all come out in your pack, but as I said, not too much of a stretch. The only problem then is that even if they do spread out, well done, you just marbled your own square. There's also the question of what happens if the ground is soft, e.g. dirt or grass. I would think the marbles wouldn't actually work at all, as stepping on one would just push it into the ground, but even if they did they wouldn't roll or spread by themselves.
Manupulating and object, Move action. Players Handbook page 143.
Well, if you think you're not working enough on this, you could also do a PA optimizer under uncertainty: i.e. you know the enemy's AC is between 22 and 26, but not anything more. Assume every figure is as likely, and...
Hmm, couldn't you just calculate across the range - e.g. if -X PA is optimal for all values, go with that? It's whether you want to optimise PA for the best average damage across the range rather than just take average (of one sort or another) optimal PA across it, I guess, but given you're guessing anyway you could do so decently without too much trouble either way, I'd have thought.
You can't really optimize across a range - Power Attack has discrete returns for any given AC, and unless changing their AC won't significantly impact your hit chance, different Power Attack values are 'optimal'.
Then again, as I've said, the marginal improvement over full PA in many average cases is minor. The more information you have, the more you can fine-tune it. The less you have, the more likely "full PA" is going to produce reasonable results (whether you're looking for kill rate or expected damage). If you've got reason to believe the enemy has higher AC, though, the marginal improvement from full PA to the optimal value can be rather substantial, particularly over an entire full attack / pounce.
Hmm, couldn't you just calculate across the range - e.g. if -X PA is optimal for all values, go with that? It's whether you want to optimise PA for the best average damage across the range rather than just take average (of one sort or another) optimal PA across it, I guess, but given you're guessing anyway you could do so decently without too much trouble either way, I'd have thought.
Yes, if -X is optimal for each value then it's the one you pick, and yes, averaging is probably a decent approximation, but the current code doesn't do it. ;)
Actually, I would think that dropping marbles from a bag being a free action is not too much of a stretch. You would arguably need to open the bag (making it a move action?) because if it wasn't open, they'd all come out in your pack, but as I said, not too much of a stretch. The only problem then is that even if they do spread out, well done, you just marbled your own square. There's also the question of what happens if the ground is soft, e.g. dirt or grass. I would think the marbles wouldn't actually work at all, as stepping on one would just push it into the ground, but even if they did they wouldn't roll or spread by themselves.
Well, as Vortsukoto notes, they could well fall under general manipulation of an object, but the surface does make a lot of difference.
My own experience with marbles suggests that (on the practical level, not necessarily as covered by game rules) a soft surface is the only way to drop them and have them stay mostly within a 5-foot area. Dropping them on a hard surface results in the marbles making a sudden, unanimous, and generally omnidirectional bid for freedom. If you need to keep marbles in a close area, you usually have to do something like pour them out across the area at ground level.
There is, I think, a formula in Excell to make a table out of the values of one cell (or maybe set) and input from others.
There's a few that do things like that, and I'm using one like it to create my moving averages (i.e. "average HP from all monsters within +/- 4 CRs of your level" - I haven't figured out a good weighting system for that (you encounter fewer CR+4s, and CR-4s matter less; while CR= is valid, it's also not the whole story of what you'll face from mook to boss), but once I do I'll add that in as well (that's not that difficult to do either, even in a single cell).
The problem isn't with the tools so much as the algorithm.
(The_Fred: As much as I adore LaTeX, I'm working in cognitive science now, and APA style is somewhat lax in terms of what it will allow. I'm getting sloppy as a result.)
www.share-online.biz/dl/2SV4T9KM9Z
its a executable file for linux, i will make a windows version as soon, as i have access to a windows system.
Also added the sourcecode.
Pro tip, by the way: Snapshots never include rage. If you don't have some racial prereq, there is never, ever ANY reason not to be an orc.
The bonus feat on human and near-human isn't a racial prereq but is often damn good reason to play such a character. And occasionally you'll find kalashtar or azurin psychic warriors for their other abilities. But I'll concur here - I had forgotten Orc was +4 (for some reason I thought it was +2), and all of the other +4s I know of have level adjustments.
Why would you play an Orc if you're going Power Attacker/Charger? I thought you guys were optimisers.
It's funny, since this discussion got intellectual, a certain someone seems to have gone.
...
Wow. The fail in that section of the post alone is so extreme, I really can't think of anything that will be sufficiently mocking.
Why would you play an Orc if you're going Power Attacker/Charger? I thought you guys were optimisers.
*facepalm*
Just stop posting. You're embarassing yourself.
Perhaps I should rephrase. Why play an untemplated Orc? I'd say you still count as Orc so long as you still have the (Orc) subtype.
Was that solely for Headlong Rush, or do you like the warlord & Gray Orc?
Headlong Rush. That said, Orc Warlord is pretty cool. I think I managed to build the world record for number of followers using it a while ago. Lulz, Martial Monk is hilarious.
What, just for Headlong Rush?
Yep. What else matters?
+4 strength, +2 con, and a charge multiplier is better than any other race.
Perhaps I should rephrase. Why play an untemplated Orc? I'd say you still count as Orc so long as you still have the (Orc) subtype.
Because LA sucks? Because there aren't really any good strength-boosting LA0 templates and that's really all that matters? Or because if you go down that road, it leads to Tauric and the Emerald Legion? Take your pick.
I'm just gonna take a guess here, but maybe Fred was referring to Water Orc and/or Half-Orc?
There's such a thhing as an Orc that isn't a Water Orc? This is news to me.
I'm just gonna take a guess here, but maybe Fred was referring to Water Orc and/or Half-Orc?
I was thinking more along the lines of Incarnate Dustform Orc, actually (stupid as that may be).