The problem of -5/+10

[Kryx]

Spawned out of Commonly Accepted House Rules (http://www.giantitp.com/forums/showthread.php?453290-Commonly-Accepted-House-Rules)

Here I'll present my argument for why -5/+10 is bad for the game.
I would ask that you try to be as respectful as possible. I expect many to be ok with the current situation and that's their choice.

Background:
The default game that doesn't allow feats has a baseline difference in damage between options like a Barbarian wielding a greatsword compared to a Barbarian wielding sword and board.
Adding feats is great - most groups do it. However not all feats are created equally. Great Weapon Master and Sharpshooter both include the option to increase your damage by 10 by decreasing your attack by 5. Under normal circumstances this can seem ok, but many common builds become extreme outliers as a result.

To show you the difference I used my DPR of Classes (https://docs.google.com/spreadsheets/d/1d-9xDdath8kX_v7Rpts9JFIJwIG3X0-dDUtfax14NT0/edit#gid=655309186) sheet. Forum math can handle some things, but it often ignores many factors.
Disclaimer: If you are like some and don't trust my sheet I've purposefully chosen simple classes that have no or very few assumptions.

Here is a very high level overview of graphs. Keeping everything the same except removing -5/+10 from all options:


-5/+10
http://i.imgur.com/PdPWwsx.png
without -5/+10
http://i.imgur.com/8MWXnRi.png





Now to add some granularity lets compare everything to a GWM Barbarian who isn't using frenzy. There is a lot of options but I'll narrow it down to Polearm+GWM vs other builds as that is the biggest outlier.
S&B classes use dueling if they have it as a class feature. Polearm+GWM guys use a halberd with -5/+10 and add the bonus attack/provoke at 8. Bladelock doesn't take -5/+10 until 17 (Foresight) because it costs him more DPR to take it earlier. Rend houserule for TWF = if you hit with both mainhand and offhand then you deal additional damage equal to proficiency modifier. Given at 11.

Level 5
Fighter Polearm+GWM: 22 DPR
Fighter Polearm+GWM w/o -5/+10: 22 DPR
Barb Polearm+GWM: 30 DPR
Barb Polearm+GWM w/o -5/+10: 22 DPR
Bladelock Polearm+GWM: 16 DPR
Bladelock Polearm+GWM w/o -5/+10: 16 DPR
Fighter TWF: 20 DPR
Fighter TWF w/ Rend Houserule: 20 DPR
Fighter S&B: 19 DPR
Barb S&B: 21 DPR
Rogue TWF: 16 DPR
Monk: 20 DPR

Level 11
Fighter Polearm+GWM: 40 DPR
Fighter Polearm+GWM w/o -5/+10: 34 DPR
Barb Polearm+GWM: 39 DPR
Barb Polearm+GWM w/o -5/+10: 31 DPR
Bladelock Polearm+GWM: 34 DPR
Bladelock Polearm+GWM w/o -5/+10: 34 DPR
Fighter TWF: 32 DPR
Fighter TWF w/ Rend Houserule: 35 DPR
Barb S&B: 23 DPR
Rogue TWF: 27 DPR
Monk: 27 DPR

Level 17
Fighter Polearm+GWM: 54 DPR
Fighter Polearm+GWM w/o -5/+10: 44 DPR
Barb Polearm+GWM: 57 DPR
Barb Polearm+GWM w/o -5/+10: 48 DPR
Bladelock Polearm+GWM: 62 DPR
Bladelock Polearm+GWM w/o -5/+10: 52 DPR
Fighter TWF: 36 DPR
Fighter TWF w/ Rend Houserule: 41 DPR
Barb S&B: 36 DPR
Rogue TWF: 36 DPR
Monk: 35 DPR

Level 20
Fighter Polearm+GWM: 68 DPR
Fighter Polearm+GWM w/o -5/+10: 55 DPR
Barb Polearm+GWM: 73 DPR
Barb Polearm+GWM w/o -5/+10: 57 DPR
Bladelock Polearm+GWM: 64 DPR
Bladelock Polearm+GWM w/o -5/+10: 55 DPR
Fighter TWF: 45 DPR
Fighter TWF w/ Rend Houserule: 50 DPR
Barb S&B: 42 DPR
Rogue TWF: 39 DPR
Monk: 35 DPR

Now that we're past the gluttony of over information let me simplify it:
Allowing -5/+10 means that certain builds become extreme outliers. Looking at level 11 for example:
A TWF Rogue can do 27 DPR. That's pretty comparable to a Barb's Polearm+GWM 31 or a Fighter's Polearm+GWM 34. In this case Rogue does 79-87% the damage that a Fighter or Barbarian does. Ok. Now if we add in -5/+10 then a Barb becomes 39 and a Fighter becomes 40. Now you're dealing 67-69%. In my opinion that is too little.
Many classes are in the same situation. Monk, Ranger, Valor Bard, Melee casters like Cleric or Sorcerer, etc. GWM by itself is similar, but to a lesser degree as it has less attacks.
It only gets worse at 20.
At 20 the rogue does 68-71% of a Fighter/Barb. Or 53-57% with -5/+10.

By choosing to allow -5/+10 Polearm+GWM becomes the best melee build choice by far. GWM and Polearm itself as now lesser options (at all tiers). Classes that previously played other roles (skills, CC, stunning) that used to put up fairly good damage (about 75-90% depending on level and class) now put up ~55-70%.

TL;DR: Allowing -5/+10 means most classes lose competitiveness with several build options: Barbarian, Battlemaster Fighter, or Paladin using GWM or Polearm+GWM. Without -5/+10 those classes still have a significant advantage, but it is not overwhelming.

[JNAProductions]

Here's what I have to say about it-numbers don't make the game run. I play D&D to play characters, to be a mighty hero slaying dastardly villains (or a dastardly villain slaying mighty heros, or in one case, a coffee shop owner helping a foreign invasion for fun and profit). I don't care if my numbers are the best they can be.

Mathematically, it might be the best. Thematically, it's sometimes the best. But the fact that there's a best option isn't a big deal when every other option is still viable and fun.

Edit: That being said, nice math there, Kryx.

[Kryx]

every other option is still viable and fun.
The whole crux of my argument is the opposite of this. Other classes become far less viable with the outliers.

[JNAProductions]

Less effective does not mean non-viable.

[JoeJ]

Three questions:

1) How do differences in survival play into this? The different combinations have different AC, hit points, and healing as well as DPR, and a dead or incapacitated character does no damage at all. Some builds, therefore, might be doing DPR for fewer rounds per day than others.

2) How much of the difference in DPR is wasted because the damage exceeds the enemy's remaining hit points? Some builds might come significantly closer to their theoretical DPR than others.

3) Most importantly, have you tested this model against actual play? Theory is nice, but if you're going to use it as a guide it needs empirical grounding.

[sophontteks]

I think a lot of people have already found this out by experience, but its always cool seeing the math. I mean, ya gotta give props to a guy who makes a freaking chart on it :smallwink:. Considering all the work you put into it, were you looking for anything beyond confirmation? Like alternatives or contradictory opinions?
I've personally already ruled it out of my games entirely. I think things like advantage would make the disparity even worse. I mean really, +10 damage at the loss of only 5 attack? That's crazy.

[Malifice]

TL;DR: Allowing -5/+10 means most classes lose competitiveness with several build options: Barbarian, Battlemaster Fighter, or Paladin using GWM or Polearm+GWM. Without -5/+10 those classes still have a significant advantage, but it is not overwhelming.

I showed you the Math in the other thread that shows you how a baseline Champion fighter 4 (who invested in +2 Strength at 4th instead of selecting GWM) gets more DPR per round than his identical twin who instead took GWM at 4th level vs 'expected' CR 4 AC of 14 from the +5/-10 alone.

You're also leaving opportunity cost out of many of your calculations. Characters that select GWM are not increasing Str by +2, so will not only have lower DPR than if they had have just increased Strength (barring outliers like Str Barbarians taking advantage of reckless attack, but even that carries its own opportunity cost) but will also be weaker in other areas aside from DPR (Str saves, Str DC's for manouvers, Athletics checks and combat manouvers like grapple and shove, hit and damage bonus with thrown weapons and other melee weapons, carrying and lifting capacity, or simply the benefits from taking a different feat than GWM).

GWM's real benefit is the bonus action extra attack.

[Submortimer]

Frankly, I don't see a problem with this. Fighters and Barbarians have a very specific spot now: out in front, taking big hits, dealing big damage. They still have to deal with mobility issues, and tun a higher risk of getting murdered in the face.

By contrast, Sharpshooter is MUCH worse: ranged doesn't have to deal with mobility, only positioning, archers are at far lower risk of getting killed, and they have the archery fighting style to mitigate the -5 to hit.

Still, though, this is okay. Imbalance in certain areas is not only desired, but required. I WANT the level 20 fighter to deal 100 damage with his polearm or bow in a round, just like I want the cleric to be able to bring me back to life and the wizard to fling 40d6 fire spells at an army. The imbalance is what makes these things FUN.

[Kryx]

Less effective does not mean non-viable.
The point of the argument is that 5e's baseline expects characters to contribute to damage fairly well even if they aren't the top damage dealer. It seems to be around 75-90% of the top class.

If you find 55-70% acceptable then that's your choice, but my whole argument is that is not what 5e's baseline expects.

I, personally, would find 55-70% nonviable.




1) How do differences in survival play into this? The different combinations have different AC, hit points, and healing as well as DPR, and a dead or incapacitated character does no damage at all. Some builds, therefore, might be doing DPR for fewer rounds per day than others.

2) How much of the difference in DPR is wasted because the damage exceeds the enemy's remaining hit points? Some builds might come significantly closer to their theoretical DPR than others.

3) Most importantly, have you tested this model against actual play? Theory is nice, but if you're going to use it as a guide it needs empirical grounding.
1. The survival numbers are the same as the baseline game. The baseline game has differences in DPR based on those numbers. My argument is that this feat distorts those baseline differences significantly.
2. No model that I know of calculates excess damage. Though based on the average monster HP and gameplay I expect that to be a very minimal difference.
3. I have experience this as a problem in play, yes. Many others have also reported seeing the exact same thing in game.




I think a lot of people have already found this out by experience, but its always cool seeing the math. I mean, ya gotta give props to a guy who makes a freaking chart on it :smallwink:. Considering all the work you put into it, were you looking for anything beyond confirmation? Like alternatives or contradictory opinions?
I've personally already ruled it out of my games entirely. I think things like advantage would make the disparity even worse. I mean really, +10 damage at the loss of only 5 attack? That's crazy.
I made this chart in February of this year and have incrementally improved upon it. I made it so that I could test my opinions and houserules against the actual math. As a result of making the sheet and seeing the math I have changed many of my opinions and houserules. -5/+10 was one of the main things I wanted to test, but also TWF, sharpshooter removing cover from ranged, ranged vs melee rogue, hand crossbow vs Heavy Crossbow, etc.




I showed you the Math in the other thread
You showed me some math. Some call that kind of math "napkin math". I find it to be misinformation in the best circumstances.
Champion Fighter is an awful build mathematically. It's literally the worst Fighter DPR possible. It is not really worth mentioning in a DPR discussion. I never claimed -5/+10 is good for all builds. I specifically said there are several outliers and those outliers are some of the most common builds.


You're also leaving opportunity cost out of many of your calculations.
Increasing Strength by 2 does nothing comparable to GWM for a Barbarian or Fighter. Simple calc at level 5:
Barbarian GWM (with 3 str mod): 33 DPR
Barbarian Greatsword (with 4 str mod): 25 DPR


weaker in other areas aside from DPR (Str saves, Str DC's for manouvers, Athletics checks and combat manouvers like grapple and shove, hit and damage bonus with thrown weapons and other melee weapons, carrying and lifting capacity, or simply the benefits from taking a different feat than GWM).
1 weaker in a bounded accuracy system isn't much. 8 DPR is huuuge. At level 12+ this 1 difference goes away.


GWM's real benefit is the bonus action extra attack.
Mathematically, it's not. Polearm is significantly better than it in that regard anyways.




By contrast, Sharpshooter is MUCH worse: ranged doesn't have to deal with mobility, only positioning, archers are at far lower risk of getting killed, and they have the archery fighting style to mitigate the -5 to hit.
Agreed that ranged is worse.


Still, though, this is okay. Imbalance in certain areas is not only desired, but required. I WANT the level 20 fighter to deal 100 damage with his polearm or bow in a round, just like I want the cleric to be able to bring me back to life and the wizard to fling 40d6 fire spells at an army. The imbalance is what makes these things FUN.
Fun for everyone who is imbalanced. Not fun for everyone like a Rogue, Monk, Ranger, Valor Bard, Melee casters like Cleric or Sorcerer, etc.
Even then the GM has to alter encounters quite significantly to deal with this boost of DPR. What was once hard is now medium so the GM has to add more to make it actually a challenge. It only serves to make those who are weak even weaker as the outliers can shine as much as normal and the other classes shine less than normal.
I prefer more viable options, not less.

[DanyBallon]

The major problem I see is that you are assuming that Combat is the only thing that matter in D&D, and that every classes needs good DPR. But D&D ain't just a combat game. While combat has always been a big part of the game (even more during the 3.P,/4e era), it still only part of the game. Rogue, Bard and to an extent, Ranger, are skill monkey and should shine out of combat.

But I agree with you that in a game where fighting is happening more often than anything else, the classes that can benefits from -5/+10 have a bigger advantage.

[This be Richard]

I'm inclined to accept that the -5/+10 situation is a serious problem, but I'm not entirely sure what I want to do about it.

I don't want to take away the feats entirely, because I like the idea of the feats in theory. It's just this particular bit of it that I'm not comfortable with. Are there any easy fixes that brings the DPR boost down to something more reasonable? Increasing the accuracy penalty or reducing the damage on a hit, maybe? I saw someone in the common house rules thread mention just applying a flat boost to damage, but I'd rather not go that direction if I can avoid it. Likewise, I'd rather not dissolve the feats into their component parts and award them through other means (such as applying cleave to great weapons).

I'm comfortable with a nerf, but I'd rather not completely undo the way the feat works.

[Kryx]

The major problem I see is that you are assuming that Combat is the only thing that matter in D&D, and that every classes needs good DPR. But D&D ain't just a combat game. While combat has always been a big part of the game (even more during the 3.P,/4e era), it still only part of the game. Rogue, Bard and to an extent, Ranger, are skill monkey and should shine out of combat.
I believe you may have read too much into my focus on DPR. D&D is about a lot more, but the crux of my issue focuses on DPR because that's where the issue is.
I wrote this above which summarizes it:

The point of the argument is that 5e's baseline expects characters to contribute to damage fairly well even if they aren't the top damage dealer. It seems to be around 75-90% of the top class.

If you find 55-70% acceptable then that's your choice, but my whole argument is that is not what 5e's baseline expects.

I, personally, would find 55-70% nonviable.




I don't want to take away the feats entirely, because I like the idea of the feats in theory. It's just this particular bit of it that I'm not comfortable with. Are there any easy fixes that brings the DPR boost down to something more reasonable?
Most replace the -5/+10 part with a +1 stat increase. That would bring both GWM and Sharpshooter more in line with normal feats.

[This be Richard]

Most replace the -5/+10 part with a +1 stat increase. That would bring both GWM and Sharpshooter more in line with normal feats.

Excellent. Thank you!

[Finieous]

The whole crux of my argument is the opposite of this.

In my opinion, your math is very strong but your argument is weak. Or to put it less provocatively, the math is objective but the judgment on it is subjective (as you recognized in your OP). For the most part, when folks argue with you about it, I believe they're arguing with the judgment rather than the math.

[Kryx]

the math is objective but the judgment on it is subjective (as you recognized in your OP).
Yup, it is subjective and as you saw I pointed that out in the OP. People are free to disagree with my subjective opinion. I'm just sharing it because I know many others who feel the same way. :)

[Crusher]

The whole crux of my argument is the opposite of this. Other classes become far less viable with the outliers.

Yes and no. GWF's value is largely depend on having a huge boss monster with a gigantic pool of HP to chop through plus a not terribly high AC, as well as a lack of differentiation in the value of damage done. For example, GWF is a *hindrance* if you're trying to finish off an injured enemy spellcaster before they get a chance to cast another spell, because accuracy is paramount. Now obviously you can simply choose not to use GWF in that situation, but then you're modestly worse off from having purchased a feat you're not using right then (perhaps you're at -1 hit and -1 damage from not getting +2 STR instead).

Additionally, while your damage output is higher, it is also much more erratic. Lets say you have a +7 to hit, 2 attacks/round and you're facing a moderately armored opponent that's AC 16. Instead of hitting 60% of the time, you're hitting 35% of the time. On average you're doing more damage, but its *far* spikier. 42% of the time you're going to miss both attacks and do zero damage. Without GWF that's only going to happen 16% of the time. Over short battles lasting a few rounds, the distribution of your expected results have *far* bigger tails.

Against a stone golem or adult dragon or something else with a huge HP total that's fine. Missing this round and clobbering it next round isn't really a problem. But if you're facing a group of, say, Hobgoblins, this could be a problem. The party's tactical planning is constrained, because when they say "Ok, Grondor, you go take two those over there" there's less certainty around the outcome. Maybe you'll get lucky and kill them both in the first round, but maybe you'll get unlucky and take 3 or even 4 rounds to kill them. Getting the bonus action attack after killing one does help (brings your chance of missing entirely that round down to ~27%), but its still more erratic.

Also, from a game-play standpoint, those erratic results are *frustrating*. When the dice are hot and you're rolling 16s and 17s all the time, you're an engine of destruction and it feels awesome. But when you're rolling 6s and 7s you feel useless. And in the situation I listed above, you've got an ~18% chance of missing 4 attacks in a row which is just god-awful. You're *more* likely to miss every attack for 2 straight rounds, than the non-GWF player is to miss both their attacks in a single round (and that's not even including any benefit from having spent the feat on something else).

Fighting a big group of small monsters, or even a smaller pack of medium-difficulty monsters is worse when you go GWF + Polearm. GWF's equalizing value against smaller creatures comes from the bonus action attack you get after killing something. But when you have Polearm as well, that value is greatly reduced because Polearm is using it already and GWF is just going to periodically make it a little better.

So, I'm not at all questioning that GWF and especially GWF + Polearm is a monster in average damage done in a vacuum (getting that 3rd attack as a bonus action does smooth things out quite a bit), but most of the time fights aren't done in a vacuum and some damage is more valuable than others. Hacking through mooks faster is nice, but is less valuable than putting down the enemy Sorcerer *NOW* before he nukes the party and GWF isn't really so good at that.

[Ruslan]

Now that we're past the gluttony of over information let me simplify it:
Allowing -5/+10 means that certain builds become extreme outliers. Looking at level 11 for example:
A TWF Rogue can do 27 DPR. That's pretty comparable to a Barb's Polearm+GWM 31 or a Fighter's Polearm+GWM 34. In this case Rogue does 79-87% the damage that a Fighter or Barbarian does. Ok. Now if we add in -5/+10 then a Barb becomes 39 and a Fighter becomes 40. Now you're dealing 67-69%. In my opinion that is too little.
Many classes are in the same situation. Monk, Ranger, Valor Bard, Melee casters like Cleric or Sorcerer, etc. GWM by itself is similar, but to a lesser degree as it has less attacks.
It only gets worse at 20.
At 20 the rogue does 68-71% of a Fighter/Barb. Or 53-57% with -5/+10.I disagree that DPR is the end-all criteria for build evaluation. Let's look at the same Rogue and Fighter out of combat. The Rogue has much more noncombat skills. He may have taken Expertise in Persuasion and Deception to be a social powerhouse, or Stealth and Perception to be an uber-scout. He has Thieves' Tools to solve a bunch of problems (the the Wizard admittedly can also solve, but it doesn't cost a spell slot to the Rogue). Reliable Talent also helps. Meanwhile, what can the Fighter do out of combat? Use an Intimidate skill, I guess. With no sizeable charisma bonus or Expertise.

Since the Rogue is so much better out of combat, it would have been totally unfair if the two classes were equal in combat. Rogue would have been the strictly superior class. Let the fighter and Barbarian be better than the rogue in combat. Let them have their DPR. It's not a bug. It's a feature.

[DanyBallon]

I believe you may have read too much into my focus on DPR. D&D is about a lot more, but the crux of my issue focuses on DPR because that's where the issue is.


But if you consider the other aspects of the game, then a rogue or a bard dealing less than 50% of the DPR of a fighter or barbarian with GWM, doesn't matter much as these classes will shine while finding traps, stealthing into a dragon lair, persuading an important NPC, etc.

Ruslan post is a perfect illustration, of what I was trying to explain :smallbiggrin:

[JoeJ]

1. The survival numbers are the same as the baseline game. The baseline game has differences in DPR based on those numbers. My argument is that this feat distorts those baseline differences significantly.

2. No model that I know of calculates excess damage. Though based on the average monster HP and gameplay I expect that to be a very minimal difference.

3. I have experience this as a problem in play, yes. Many others have also reported seeing the exact same thing in game.

1. But what are the survival numbers? That is, how many rounds per day is each build actively doing damage? Also, why are they the same as the baseline? Shouldn't they change if, for example, a character who passes up GWM opts for Heavy Armor Master or Magic Initiate for the Shield spell?

2. Expectations can be wrong. I'd really like to see some numbers for this. And I'd like to know how it changes if the number of encounters per day is changed within the same xp budget. (That is, more but easier fights vs. fewer but harder ones.)

3. Can you quantify the result? How much damage/day are the various characters doing, compared to what the model predicts? A subjective evaluation of whether or not the difference is enough to be a problem doesn't give me much to go on.

I appreciate that you've done a lot of fine work on your model. Taking it a step further would be even more helpful, although you're at the point where further complexity might make active simulation (the "Monte Carlo" approach) a better choice than using static averages as inputs. Particularly for examining the effects of different monster groupings, and of synergies between different characters working together. (For example, a wolf totem barbarian should be significantly less effective alone than paired with another melee character.)

[Kryx]

Since the Rogue is so much better out of combat, it would have been totally unfair if the two classes were equal in combat. Rogue would have been the strictly superior class. Let the fighter and Barbarian be better than the rogue in combat. Let them have their DPR.

But if you consider the other aspects of the game, then a rogue or a bard dealing less than 50% of the DPR of a fighter or barbarian with GWM, doesn't matter much as these classes will shine while finding traps, stealthing into a dragon lair, persuading an important NPC, etc.
The default game considers all those factors and has the Rogue mostly competitive in damage. I think that's a good balance level. Others prefer a Rogue (or a Ranger, or a Monk) to do much less. I personally can't see how that is justifiable, but my intent isn't to change their minds as that would be a fruitless argument.
My intent is to show the math for everyone to make up their own mind.

For those of you who think it's a feature not a bug: I ask you to look not only at the comparison between other classes, but also the same class as presented in my initial argument. S&B and TWF make it out the worst, but even GWM and Polearm suffer vs Polearm+GWM

[Theodoxus]

Last session our DM used a Shartshooter/Xbow Expert dragonriding minion against us. It was devastating. He wanted to show how he felt, facing against my GWM PAM barbarian. Point made.

I still think the best compromise for those two feats is to change it to the old Power Attack: Minus Proficiency to hit, Plus two times Proficiency to damage. This keeps the power curve on par without being absolutely encounter ending at low level on a lucky hit and still viable later on.

Yes, it's fiddly with the "simplicity on all" that 5E desires. But -5/+10 already jumps the shark on that regard. Having it scale isn't pushing the rock any further up the hill. I get why the feats don't simply impose Disad (since it's so easy to counter), that a static negative works better...

I'd be interested in seeing the DPR on the scaling though... maybe it doesn't solve the problem either.

[Kryx]

1. But what are the survival numbers? That is, how many rounds per day is each build actively doing damage? Also, why are they the same as the baseline? Shouldn't they change if, for example, a character who passes up GWM opts for Heavy Armor Master or Magic Initiate for the Shield spell?

2. Expectations can be wrong. I'd really like to see some numbers for this. And I'd like to know how it changes if the number of encounters per day is changed within the same xp budget. (That is, more but easier fights vs. fewer but harder ones.)

3. Can you quantify the result? How much damage/day are the various characters doing, compared to what the model predicts? A subjective evaluation of whether or not the difference is enough to be a problem doesn't give me much to go on.

I appreciate that you've done a lot of fine work on your model. Taking it a step further would be even more helpful, although you're at the point where further complexity might make active simulation (the "Monte Carlo" approach) a better choice than using static averages as inputs. Particularly for examining the effects of different monster groupings, and of synergies between different characters working together. (For example, a wolf totem barbarian should be significantly less effective alone than paired with another melee character.)
1. If you want to compare all feats like GWM vs Heavy Armor or GWM vs Magic Initiate then by all means go for it. I compared it directly to other DPR options and found it to be an outlier.

2. You've brought this topic up often now. If you want numbers around this then I suggest doing a proof of concept yourself. It would be a lot of work to think about how to do this and then make it. I'm not convinced there is a good way to model it, nor is it a large factor unless you're fighting many mooks. If you are then go ahead and do it. I'd be curious.

3. I did not run a full playtest comparing Polearm+GWM vs Polearm vs GWM. Even then you'd have to run it thousands of times under different circumstances to see actual usable results. The rest is just conjecture and feeling. This is why I do not value playtesting much - I find it to be based mostly on feeling.




I still think the best compromise for those two feats is to change it to the old Power Attack: Minus Proficiency to hit, Plus two times Proficiency to damage. This keeps the power curve on par without being absolutely encounter ending at low level on a lucky hit and still viable later on.

Yes, it's fiddly with the "simplicity on all" that 5E desires. But -5/+10 already jumps the shark on that regard. Having it scale isn't pushing the rock any further up the hill. I get why the feats don't simply impose Disad (since it's so easy to counter), that a static negative works better...

I'd be interested in seeing the DPR on the scaling though... maybe it doesn't solve the problem either.
I thought about this and -3/+6. I don't have the numbers on me currently, but the numbers have the same effect on the math.

[locke411]

I'll need to agree that -5/+10 does invite some issues, especially with bounded accuracy and limited AC growth in 5e. Polearm wielder getting an offhand with -5/+10 is definitely nice, and one of the best builds. TWF ends up dragging much too much. I like the idea of finding/coming up with feats to help the other builds, though.

[DanyBallon]

The default game considers all those factors and has the Rogue mostly competitive in damage. I think that's a good balance level. Others prefer a Rogue (or a Ranger, or a Monk) to do much less. I personally can't see how that is justifiable, but my intent isn't to change their minds as that would be a fruitless argument.
My intent is to show the math for everyone to make up their own mind.

For those of you who think it's a feature not a bug: I ask you to look not only at the comparison between other classes, but also the same class as presented in my initial argument. S&B and TWF make it out the worst, but even GWM and Polearm suffer vs Polearm+GWM

Just for the sake of discussion; anyone have considered that it may be the rogue that is too powerful in a no feat game, and that the issue is sneak attack, which was the solution to keep the rogue competitive in the combat heavy 3.P edition, where as in previous edition he was truly a skill monkey?

[Kryx]

Just for the sake of discussion; anyone have considered that it may be the rogue that is too powerful in a no feat game, and that the issue is sneak attack, which was the solution to keep the rogue competitive in the combat heavy 3.P edition, where as in previous edition he was truly a skill monkey?
Well you'd have to ask the same for most other classes who have the same level of damage (Ranger, Monk being the prime candidates that are in a similar role, albeit slightly less out of combat).

But your opinion is equally as possible - maybe some do indeed think a rogue too strong.
I, personally, don't think the best melee rogue build doing 79-87% the damage of a Fighter build to be a problem. I think that's the sweet spot. Plus the damage would be reduced by a fair amount if the rogue doesn't use TWF. More like 65-70%

[ad_hoc]

I am on board with the people saying that there is more to the game than combat and that it doesn't completely invalidate other characters.

The thing is, what does it add to the game?

Without GWM you can still wield a 2-handed weapon
Without Sharpshooter you can still fire ranged weapons


If it unbalances combat and doesn't add anything to the game, why keep it? I like that feats are in the book. I like that they are specified as being optional. The game is letting us know that not all circumstances are accounted for and you may not want to allow some or all of them.

Personally I allow feats on a feat by feat basis.

[Crusher]

My question is: Is there a reduced level at which the damage output becomes appropriate? For example, someone above mentioned possibly making GWF (and Sharpshooter, for that matter) a flat +2 damage. That seems reasonable to me. What if GWF was tweaked to, say, -3/+6? It would maintain the original flavor of the feat, but would it be enough of a nerf to bring the damage output into line?

[DanyBallon]

Well you'd have to ask the same for most other classes who have the same level of damage (Ranger, Monk being the prime candidates that are in a similar role, albeit slightly less out of combat).

But your opinion is equally as possible - maybe some do indeed think a rogue too strong.
I, personally, don't think the best melee rogue build doing 79-87% the damage of a Fighter build to be a problem. I think that's the sweet spot. Plus the damage would be reduced by a fair amount if the rogue doesn't use TWF. More like 65-70%

Ranger are quite decent and get access to Sharpshooter if they go the ranged route. And where they shine the most is during the exploration part of the game.

Monks I do agree, that they don't have much outside of combat, except for acrobatics check.

As far as rogue are concerned, I prefer rogues to be more about trapfinding, lockpicking, breaking in where access is prohibited, than being a effective fighter. But this may be me being nostalgic of my 2e days :smallbiggrin:

[Kryx]

My question is: Is there a reduced level at which the damage output becomes appropriate? For example, someone above mentioned possibly making GWF (and Sharpshooter, for that matter) a flat +2 damage. That seems reasonable to me. What if GWF was tweaked to, say, -3/+6? It would maintain the original flavor of the feat, but would it be enough of a nerf to bring the damage output into line?
This was discussed above. -3/+6 is nearly the same as -5/+10. Somewhat less, but not by much.
Flat 2 damage is quite strong. That would still be a large boost. Not sure how large.



Ranger are quite decent and get access to Sharpshooter if they go the ranged route. And where they shine the most is during the exploration part of the game.

Monks I do agree, that they don't have much outside of combat, except for acrobatics check.
I probably should't have brought up Ranger as there are so many opinions on it, but a melee Ranger is not good damage wise. Though even if that's fixed they still fall well short of a GWM build due to not having access to -5/+10.

[Saggo]

I disagree that DPR is the end-all criteria for build evaluation. Let's look at the same Rogue and Fighter out of combat. The Rogue has much more noncombat skills. He may have taken Expertise in Persuasion and Deception to be a social powerhouse, or Stealth and Perception to be an uber-scout. He has Thieves' Tools to solve a bunch of problems (the the Wizard admittedly can also solve, but it doesn't cost a spell slot to the Rogue). Reliable Talent also helps. Meanwhile, what can the Fighter do out of combat? Use an Intimidate skill, I guess. With no sizeable charisma bonus or Expertise.

Since the Rogue is so much better out of combat, it would have been totally unfair if the two classes were equal in combat. Rogue would have been the strictly superior class. Let the fighter and Barbarian be better than the rogue in combat. Let them have their DPR. It's not a bug. It's a feature.
You're arguing that DPR reduction is a balanced cost for out of combat utility, and I'm willing to bet everyone here will agree. My question is at what cost is the DPR reduction no longer balanced, considering that out of combat utility (especially social skills) is often balanced around requiring just Proficiency since most checks are static DCs or rolling against another d20.

If you accept the model (I personally do, I can't find anything catastrophically wrong with the math), then you've effectively argued that a few skills with Expertise, Jack of all Trades, or optimized Charisma mods (which does add to DPR for the classes commonly using it) is worth doing as low as half the DPR of an -5/+10 build, a build that can still pick out of combat proficiencies. Of that, I disagree. That disparity can occur as early as level 5. And while combat is not always the proportional majority of table time spent, it is the primary focus of the mechanics and the most complex aspect of 5e.

What I do appreciate is that 5e combat is still playable and enjoyable, even if classes are widely disparate.

[Finieous]

But your opinion is equally as possible - maybe some do indeed think a rogue too strong.
I, personally, don't think the best melee rogue build doing 79-87% the damage of a Fighter build to be a problem.

I don't think rogues are too strong given that fighters and barbarians can deal a lot more damage if they choose, but I think they would be without those heavy-weapon options. Specifically, 79-87% is too strong compared to the highest-damage fighter build that spends the day using Trip Attack to maximize raw DPR. Given the rogue's advantage in consistency, damage distribution, rest-independence, ranged combat, skills, utility, and damage mitigation, I can't really imagine why anyone would choose to play the fighter. Same for the ranger. Monks? I'd feel bad for monks if I didn't hate monks. I admit my biases. :smallbiggrin:

[Malifice]

Increasing Strength by 2 does nothing comparable to GWM for a Barbarian or Fighter. Simple calc at level 5:
Barbarian GWM (with 3 str mod): 33 DPR
Barbarian Greatsword (with 4 str mod): 25 DPR

He claims, while using the one outlier (due to easy access to advantage via reckless attack).

A Champion fighter with the same options at 5th level (GWM v +2 Strength) loses out on DPR by selecting GWM. As do Eldritch Knights, Rangers (both sorts) Monks (assuming they want to use heavy weapons), all the full spellcasters, all Druids, Bards and Clerics (barring war clerics a few times per long rest) all Warlocks, Paladins (barring vengance paladins), and all Rogues.

Advantage via Reckless skews the numbers. The bonus action attack is far more useful.

At 5th level vs expected AC's, +1 to hit and +1 to damage place the DPR in favor of the non GWM warrior (discounting the bonus action extra attack).

Then look at the other stuff (aside from raw DPR) that a GWM fighter misses out on. The opportunity cost of picking up DPR instead of increasing a stat or learning another feat. Its totally balanced as is (barring the Barbarian, and who cares? That's his fluff).

[Theodoxus]

GWM+PAM already allows a bonus attack with the polearm that would deal 1d4+Str+10... this is only slightly worse (and probably balanced around) a 1d6+str (or dex) + 10 mainhand, and 1d6+str (or dex) + 10 offhand in a TWF situation. (Provided you had the fighting style to support it - although, I don't see why you couldn't make an improved TWF feat that rolled that into it - would allow fighters/rangers to pick up a different style while still being competitive with TWF.

We know that GWM+PAM is the strongest combination, in part because of that bonus attack. It's a cascading decision tree. Did you crit/fell a guy and have a secondary target? Cleave with GWM. Neither of those happen? Bonus attack with PAM. You're looking at a max of 1d10+Str+10 per attack and the same with the bonus if it triggers, or at worse, a 1d4+Str+10.

With this new TWF Feat, you're looking at 1d8+(attack mod)+10 per attack and the same with the bonus. A d8 is what, 2.5 points better than a d4? But a d8 is one point less than d10? So, at 4th level, vhuman with GWM+PAM against a single target, would be doing 1d10+13 (using SPBI) and 1d4+13 on a bonus, or (just pure damage, not taking to-hit into consideration (19+15), for 34 total)) Where a TWF with the TWF Feat and Improved (theoretical feat I made up) TWF would be doing 1d8+13 and 1d8+13 (18+18) for 36 total.

For a two feat investment, that makes TWF on par with the similar two feat investment with GWM+PAM; sans reach and reaction attack of PAM (arguably better, imo).

[Finieous]

He claims, while using the one outlier (due to easy access to advantage via reckless attack).

A Champion fighter with the same options at 5th level (GWM v +2 Strength) loses out on DPR by selecting GWM. As do Eldritch Knights, Rangers (both sorts) Monks (assuming they want to use heavy weapons), all the full spellcasters, all Druids, Bards and Clerics (barring war clerics a few times per long rest) all Warlocks, Paladins (barring vengance paladins), and all Rogues.


Not just the one outlier. He uses battlemaster to trip prone for advantage.

[JNAProductions]

Champion Fighters, Valor Bards, Ancient/Devotion Paladins, Casting Rangers...

There's a lot of people who have no guaranteed advantage.

[Kryx]

I don't think rogues are too strong given that fighters and barbarians can deal a lot more damage if they choose, but I think they would be without those heavy-weapon options. Specifically, 79-87% is too strong compared to the highest-damage fighter build that spends the day using Trip Attack to maximize raw DPR. Given the rogue's advantage in consistency, damage distribution, rest-independence, ranged combat, skills, utility, and damage mitigation, I can't really imagine why anyone would choose to play the fighter. Same for the ranger. Monks? I'd feel bad for monks if I didn't hate monks. I admit my biases. :smallbiggrin:
So you think the base rogue is too strong. I don't agree, but that's a fine opinion to take, but Rogue was just an example that represents every class that cannot take -5/+10. TWF (even in a fixed state), S&B, and the Monk and Ranger already mentioned.




He claims, while using the one outlier
Please leave if you're simply here to be combative. See the OP:

I would ask that you try to be as respectful as possible.

Disagreement is fine. Rudeness is uncalled for and I'll simply ignore any post with it.




1d6+str (or dex) + 10 mainhand, and 1d6+str (or dex) + 10 offhand in a TWF situation. (Provided you had the fighting style to support it - although, I don't see why you couldn't make an improved TWF feat that rolled that into it - would allow fighters/rangers to pick up a different style while still being competitive with TWF..
You could indeed allow all classes to take -5/+10, though that just bloats the math vs the MM and DMG averages. I thought about it for a long time, but rejected it.




Not just the one outlier. He uses battlemaster to trip prone for advantage.
Many classes can get boosts to attack. Paladin for example can run Bless. OoV Paladin can get advantage.




Champion Fighters, Valor Bards, Ancient/Devotion Paladins, Casting Rangers...

There's a lot of people who have no guaranteed advantage.
There are indeed. That's the whole argument! All of those options become much less viable in comparison to a -5/+10 build.

[Malifice]

Not just the one outlier. He uses battlemaster to trip prone for advantage.

Yeah; which is why I left them and Vegance paladins in there as well. Noted martial strikers who are designed to deal damage with big heavy weapons who have easy access to advantage.

For the other two fighters, all the rogues, the other two paladins, the two rangers, monks, all barring one cleric domain a few times per day, the druids, warlocks, bards and spellcasters its a suboptinal choice for two handers vs expected critter AC at 5th level (accounting for the +5/-10 in isolation - the bonus action attack is where the feats real power kicks in).

It only becomes worth taking it for those classes once you have your strength maxed out and against certain AC's - and then its a simple increase in already high DPR vs (lucky, alert, resilient, HAM etc).

[Kryx]

It only becomes worth taking it for those classes once you have your strength maxed out and against certain AC's - and then its a simple increase in already high DPR vs (lucky, alert, resilient, HAM etc).
Those very common classes are indeed the issue. Plus normal Paladin who can run Bless for himself and his party which minimizes it for all.
They can also take lucky, alert, resilient, HAM, etc as well.

-5/+10 is a huge boost of DPR. On top of SS or GWM it's worth at least 1.5 feats, likely 2 in pure DPR value.

[Malifice]

Please leave if you're simply here to be combative. See the OP:

I'm not being combatative man. Im just saying that youre presenting the outlier to prove a point.

90 percent of the classes at 5th level are better off bumping strength by +2 vs expected average AC. Including EK's and Champions, Devotion and Ancients paladins, all Bards, Rogues, Rangers, Monks and spellcasters.

GWM only becomes worth it if you have a ready source of advantage, and even then only against a range of AC's, and even then it costs you opportunity cost elsewhere.

[Kryx]

Im just saying that youre presenting the outlier to prove a point.
These classes are literally the whole point of the thread. Let me quote the OP:


Great Weapon Master and Sharpshooter both include the option to increase your damage by 10 by decreasing your attack by 5. Under normal circumstances this can seem ok, but many common builds become extreme outliers as a result.

[Finieous]

So you think the base rogue is too strong. I don't agree, but that's a fine opinion to take, but Rogue was just an example that represents every class that cannot take -5/+10. TWF (even in a fixed state), S&B, and the Monk and Ranger already mentioned.


Yeah, it would be a concern in a no-feats game if I were running one. In a feats game, S&B is entirely dependent on party composition -- with other melee PCs to benefit from Shield Master, along with the improved defense, it's fine. Ranger is fine. Monk I admitted I don't care about. For the record, I'd be fine with a feat or fix that boosts TWFing a bit, particularly 11+, but I still wouldn't want it to actually match great weapon damage: consistency and distribution need to count for something, though I don't know how to quantify that value. The only reason it doesn't bother me more is that TWFing is kinda goofy outside of dueling and gladiatorial performance. But it's become a D&D staple and it shouldn't be as poor as it is at higher levels. Something like your Rend feat would be pretty cool.

[Malifice]

Those very common classes are indeed the issue.

Yeah man. And V paladins, Barbs and BM Fighters are designed to be the kings of DPR, and are pretty much designed to use great weapons (especially so in the case of the Barb).

Theyre high damage tanks all three of them.

Damage isnt the be all and end all of the game, let alone for combat. Look at Monks.


Plus normal Paladin who can run Bless for himself and his part which minimizes it for all.

Yeah of course he can run a bless spell mate. We can factor in Haste, the smite spells, hunters mark or a swathe of melee buffs.


They can also take lucky, alert, resilient, HAM, etc as well.

But they didnt. They devoted a level to GWM, and those three or four martial paths that can benefit from it, should be able to get a benefit from it for the cost.


-5/+10 is a huge boost of DPR. On top of SS or GWM it's worth at least 1.5 feats, likely 2 in pure DPR value.

Its fantastic for Barbs, BM Fighters with tripping attack (and sup dice remaining), Vengance Paladins or someone rocking around with a Wolf totem barbarian in the party.

Its not the greatest for everyone else.

[Kryx]

Its fantastic for Barbs, BM Fighters with tripping attack (and sup dice remaining), Vengance Paladins or someone rocking around with a Wolf totem barbarian in the party.
Its not the greatest for everyone else.
Repeating last post as it seemingly wasn't read:


These classes are literally the whole point of the thread. Let me quote the OP:


Great Weapon Master and Sharpshooter both include the option to increase your damage by 10 by decreasing your attack by 5. Under normal circumstances this can seem ok, but many common builds become extreme outliers as a result.


We disagree on the appropriate divergence of class DPR. That's fine. No need to continuously argue it.

[Pex]

My paladin uses a shield. Wearing full plate his AC is 20. He also has heavy armor master. Last session the mooks kept missing him. Only the BBEG could hit him regularly, and 3 damage was shaved off each hit. I'm not crying over the less damage I do for not going the great weapon master route. I'm in a party. I don't need to solo kill a monster, though I have done it once in a while. Smites have been helpful fighting against undead. My character has durability.

The tortoise did win the race.

[Malifice]

These classes are literally the whole point of the thread. Let me quote the OP:

Have you considered that those classes are intended to deal more damage?

Even taking GWM (or even all feats) out of the picture they get a substantial percentage increase in DPR when obtaining advantage, blowing sup dice, recklessly attacking, and channel divinity/ hunters mark/ bless. Without feats, they are the leading tanks in the game.

Thats what they're designed to do. BM's lack the spells of an EK (haste, fireball, mirror image, counterspell, shield) the all day versatility of the Champion, the exploration pillar stuff of the ranger, the skills of the Rogue, stunning fist, mobility and defences of the Monk etc.

All BM's, Valadins and Barbs can do is hit stuff, hard. They may come to the party from time to time with a few extra bells and whisles (some 'ribbons' from the path abilities, sizing up your opponent as a BM etc) but they are sorely lacking in this area.

[KorvinStarmast]

TL;DR: Allowing -5/+10 means most classes lose competitiveness with several build options: Unless the party is playing to find out who does the most damage, and competes with one another to see who does on a given raid or campaign, I am at a loss to see where this is a problem. It appears to me that the -5/+10 in intended to reflect high level warriors being the masters of combat and perilous foes. It appears, from your numbers, to do that.

I won't argue whether or not this was done as a mitigation to linear fighter quadratic magic user syndrome, but if so ... fine.

Rogue: if what you wanted to be as the damage dealing king of battle, then next time be a fighter.

[hymer]

My mathematical intuition is a little rattled here. Am I right in this: The reason it is particularly valuable to get advantage while -5/+10ing is that the assumption is that you'll more likely hit than miss; that it pulls the likelihood you hit closer to 50% And that the closer to 50-50, the more is advantage worth? So you may lose 5 points to-hit, but you pick up another two or so extra with advantage?

[This be Richard]

Have you considered that those classes are intended to deal more damage?
Even taking GWM (or even all feats) out of the picture they get a substantial percentage increase in DPR when obtaining advantage, blowing sup dice, recklessly attacking, and channel divinity/ hunters mark/ bless. Without feats, they are the leading tanks in the game.

They might hit harder even without feats, but not this much harder. I suspect that, if they really were intended to hit stuff that much harder, we would probably see comparable differences in DPR even in the absence of feats. But this particular gap not only requires feats to happen, it even requires very specific feats that seem not to match up with the rest. That's why I have the feeling that it wasn't intended to happen to the extent we're seeing here.

[Kryx]

My mathematical intuition is a little rattled here. Am I right in this: The reason it is particularly valuable to get advantage while -5/+10ing is that the assumption is that you'll more likely hit than miss; that it pulls the likelihood you hit closer to 50% And that the closer to 50-50, the more is advantage worth? So you may lose 5 points to-hit, but you pick up another two or so extra with advantage?
There is some of that, but 65% is +15% over the normal while 40% is 10% below the normal so it's pretty minimal.
http://onlinedungeonmaster.com/wp-content/uploads/2012/05/Advantage_Disadvantage_Probabilities.png
In the 65% case you're gaining 22.75%
In the 40% case you're gaining 24%
Not significant, really.

[hymer]

There is some of that [...] Not significant, really.

I see. Thanks!

[Z3ro]

Also, from a game-play standpoint, those erratic results are *frustrating*.

This is the reason the -5/+10 feats aren't a problem at my table. Yes, hits are bigger when they land. But missing three or four attacks in a row sucks something fierce. Myself, and most of the people I play with, stay away from these feats because we prefer consistent damage and hitting more. It's the same reason I like TWF and monks; hitting things is fun! One of my players loves the big hits; he plays a half-orc barbarian and uses GWM whatever the situation because he likes doing more damage. Haven't noticed him killing significantly more things than my monk, due to the misses.

[hymer]

This is the reason the -5/+10 feats aren't a problem at my table. Yes, hits are bigger when they land. But missing three or four attacks in a row sucks something fierce. Myself, and most of the people I play with, stay away from these feats because we prefer consistent damage and hitting more. It's the same reason I like TWF and monks; hitting things is fun! One of my players loves the big hits; he plays a half-orc barbarian and uses GWM whatever the situation because he likes doing more damage. Haven't noticed him killing significantly more things than my monk, due to the misses.

Perceptions are powerful, indeed. I have a fellow player who is barbarian. He hasn't used reckless attack in three sessions. I'm not sure I want to encourage him to use it more. :smallredface:
Anyway, I believe this discussion is meant to deal with the times when -5/+10 is a problem. They may have no problems with snowdrifts in Nairobi, but it's still a problem worth discussing for people living closer to the poles. :smallsmile: For the rest of us following the thread, it may simply be an interesting diversion. Or preparation in case we should encounter the problem at a later date.

[Gwendol]

Let me ask a question here: if a player picks a feat and class specifically to deal plenty of melee damage, why is that a problem?

[Crusher]

Alright, I've done some number crunching and I have some observations.

- Kryx, the thoroughness of your spreadsheet is a little alarming. I have some minor quibbles over some of your estimates (Why does Fighter Crit% spike so sharply once you get into Prone targets? I fear you might be double counting advantage. Also, I wonder if you're slightly over-estimating Prone%, seeing as how if you kill something you have to start attacking a target that's no longer Prone. Its possible you accounted for this and I missed it, but I didn't see it.), but, seriously, these are *minor* quibbles and that spreadsheet is a work of art. If you want to know how damage averages out by class, that spreadsheet is the thing to look at. Bring a pith helmet.

- As intuition would suggest, -5/+10 isn't a widespread problem. GWM is only barely a positive for Paladins. A level 20 Paladin stuck with a normal weapon might actually get more value out of Savage Attacks than GWM, though with even a +1 weapon GWM edges back ahead, though only slightly.

- Fiddling with the numbers (such as tying it to Proficiency) doesn't change things very much, because -5/+10 is overkill for most classes. Paladins get maximum value at -2/+4 and actually get worse after that, while Fighters peak at -3/+6 and then basically plateau after that. Bear in mind that this is with normal weapons, so magic weapons will indeed make things a little better as you go up.

- Barbarians are the real culprit. Yes, BM fighters get good mileage out of GWM, but not like Barbarians do. For Barbarians, mostly thanks to Reckless attacks, GWM is an unalloyed plus that just keeps on giving. Because they get perma-Advantage, they actually get a little less mileage out of magic weapons than other classes.

- BM Fighters at level 20 are just ridiculous. GWM is the cherry on top, but even without it they're already a hot fudge sundae of awesome all on their own.

Their basic per round damage isn't that scary, but Action Surges and Maneuvers give you a lot if/then decisions. So they cruise along doing moderate damage, but when things break right they can burst impressively. 4 attacks/round + 2 surges (per short rest) + 6 chances to trip people (per short rest) = massive destruction. If you successfully trip your target, you do normal damage (plus a superiority die), you then surge and get 7 attacks with advantage, and if any of them crit (~53% chance of yes, including the trip itself) you get a final 8th attack with advantage. That is one hell of a lot of potential damage. And with that many chances to trip someone, why waste an Action Surge on someone who isn't tripped?

GWM adds to the power (though not nearly by as much as the Barbarian) but the ability to combine abilities and nova makes the Fighter arguably too strong already, and GWM just arguably pushes fighters even farther past where they "ought" to be.

- Melee classes scale really, really weirdly. They scale up incredibly fast at low levels, and then just coast to level 20. Barbarians and Paladins both do ~55% as much damage at level 5 as they'll do at level 20 (not accounting for improved magic items). A Barbarian gains more raw damage output from going from level 3 to level 5 than they do from level 5 to level 19. And if you assume the Barb goes from a normal weapon at level 3 to a +1 weapon at level 5 to a +2 weapon at level 19 its even worse (though a +3 weapon at 19 kicks the advantage back the other way).

Fighters are really the only class that continue scaling up in a halfway decent kind of way. Their abilities aren't quite as front-loaded as the other classes thanks to continued increased attacks, action surges, and superiority dice. So they're actually a littler weaker than the other melee at level 5, but by level 20 are juggernauts.

After crunching it all, my suggestion is:

1) Shift GMW to -6/+10, it makes a big difference
2) Bump down Fighters' superiority dice from 4 to 3 to start, then increase by 1 at levels 7 and 15 as before
3) Don't worry too much about level 20 balance. Everyone at level 20 will be a juggernaut in their own way but they should also all be within days of enforced retirement.

The GWM classes still do *way* more damage than everyone else, but the gap is modestly reduced without changing the flavor of the talent. This also reduces the advantage that Fighters have over Barbs and Paladins during mid-levels (roughly 7-17) to a more manageable size (its ok that Fighters do more damage than Paladins, but the gap at level 11 is ~35% which seems high thanks to the 3rd attack and that seems high).

Honestly, Barbarians and Paladins should probably be reworked in some fashion to be less front-loaded (Barbs especially), but that's beyond my pay-grade.

[hymer]

Let me ask a question here: if a player picks a feat and class specifically to deal plenty of melee damage, why is that a problem?

Ask them in Kamchatka, my Kenyan friend. :smallwink:

[Gwendol]

The point being: you can make up plenty of similarly worded questions. It is only a problem if there is assumption that the classes should be closer to equal by some metric (DPR in the current case).

[hymer]

The point being: you can make up plenty of similarly worded questions. It is only a problem if there is assumption that the classes should be closer to equal by some metric (DPR in the current case).

But isn't this a thread that has that very assumption as its premise? It's a little like going to art class, ask everyone's attention, and say 'Well, isn't this whole art thing kinda pointless?' :smallsmile: I need to do that sometime, by the way.

[Kane0]

My group has considered a cpuple alternatives to -5/+10:
- half feat (+1 str/dex)
- add half prof bonus to weapon damage when used in two hands
- add full prof bonus to weapon damage when used in two hands
- add +2 to weapon damage when used in two hands

The half feat solution is boring, so of the other three which pne do you think would be closer balanced? I'm leaning towards prof bonus myself. Not asking for a full math and chart response mind you, just the opinion pf one more mathematically inclined than myself.

[Finieous]

I wanted to add Fighter S&B back in at 11 and 17 and 20, since it seems to have dropped after Level 5. It's way too interesting and relevant for the discussion to leave out.


Level 11
Fighter Polearm+GWM: 40 DPR
Fighter S&B: 36 DPR
Fighter Polearm+GWM w/o -5/+10: 34 DPR
Barb Polearm+GWM: 39 DPR
Barb Polearm+GWM w/o -5/+10: 31 DPR
Bladelock Polearm+GWM: 34 DPR
Bladelock Polearm+GWM w/o -5/+10: 34 DPR
Fighter TWF: 32 DPR
Fighter TWF w/ Rend Houserule: 35 DPR
Barb S&B: 23 DPR
Rogue TWF: 27 DPR
Monk: 27 DPR

Level 17
Fighter Polearm+GWM: 54 DPR
Fighter S&B: 50 DPR
Fighter Polearm+GWM w/o -5/+10: 44 DPR
Barb Polearm+GWM: 57 DPR
Barb Polearm+GWM w/o -5/+10: 48 DPR
Bladelock Polearm+GWM: 62 DPR
Bladelock Polearm+GWM w/o -5/+10: 52 DPR
Fighter TWF: 36 DPR
Fighter TWF w/ Rend Houserule: 41 DPR
Barb S&B: 36 DPR
Rogue TWF: 36 DPR
Monk: 35 DPR

Level 20
Fighter Polearm+GWM: 68 DPR
Fighter S&B: 65 DPR
Fighter Polearm+GWM w/o -5/+10: 55 DPR
Barb Polearm+GWM: 73 DPR
Barb Polearm+GWM w/o -5/+10: 57 DPR
Bladelock Polearm+GWM: 64 DPR
Bladelock Polearm+GWM w/o -5/+10: 55 DPR
Fighter TWF: 45 DPR
Fighter TWF w/ Rend Houserule: 50 DPR
Barb S&B: 42 DPR
Rogue TWF: 39 DPR
Monk: 35 DPR[/SPOILER]

[Corran]

Completely agree with the OP. Let us discuss a hypothetical. Say I wanna play a barbarian. And because I watched a lot of that vikings series or whatever, I want my barbarian to fight with an axe and a shield. Taking out the GWM feat, I can go on and play my SnB barbarian without looking back at when I created this character with remorse. He can go on and have his glorious moments, and share the spotlight equally with all other members of the group. Everyone had fun during the encounter, balanced pcs, end of story. Now lets allow the GWM feat. Reaching 4th level (or from 1st level if human variant), GWM is an option. As I went with SnB, I will not select it obviously. I can go on and play my SnB barbarian as before, but since I am a math guy I will know that my barbarian would be considerably better if he was just wielding a two hander and had the GWM feat. By which I mean, the trade off of having a shield and a feat/ASI to spend wherever I want to, will not make up for all that extra damage my character would do if he was using a two-hander and the GWM feat while attacking recklessly. I know people will say that one should not look at it the way I just presented it, but knowing some basic probability will make it a hard choice for me to select the fighting style I intended for, as it is very sub-optimal. I want to be able to choose a build for my character that suits my preferences, without having to spend too much time thinking if I should do so because it is way worse (and hence not as fun to play - take away balance, hurts fun in some way) than another build. Bottom line, I want equally (at least as much as possible) supported options. And -5/+10 takes that away for some specific cases, as the OP mentioned.

[steppedonad4]

It really annoyed me when I saw that this mechanic was added. It was said from the very start that penalties and bonuses to attacks wouldn't be added in this manner and then they put them in anyway. If they had stuck to the methodology of the system, then the feats would've used disadvantage instead (which is mathematically almost the same) and this would've also eliminated the issues that the feat has as they primarily come from having advantage on the attacks and thus negating the -5 penalty.

[Finieous]

You either want to play Rollo with his Dane-axe, or your axe-and-shield Viking wants to be a fighter. :smallwink:

[Kane0]

It really annoyed me when I saw that this mechanic was added. It was said from the very start that it penalties and bonuses wouldn't be added in this manner and then they put them in anyway. If they had stuck to the methodology of the system, then the feats would've used disadvantage instead (which is mathematically almost the same) and this would've also eliminated the issues that the feat has as they primarily come from having advantage on the attacks and thus negating the -5 penalty.

Can someone so a quick math check on this please? It sounds solid enough to look at.

[Lollerabe]

So let's say I wanted to play a GW paladin in a GWMless game.

I really wanted to play a GW paladin, but seeing how much it sucked compared to SB + duelist + shieldmaster I just couldn't do it.

How is that a point? You can play SB barb all you want even with GWM still there - if you somehow can't enjoy that because it's an inferior choice that's on you. The whole notion that you ADD enjoyment and build diversity by REMOVING options is assbackwards to me.

At the end of the day Kryx said it pretty clearly in the original post. If you don't think that GW barbs, figthers and paladins should be the kingpins of melee DPR by a large margin that's your opinion - and that's fine, alot of us just don't agree with that.

I also believe that alot of us who think GWM is fine as is, consider the dps chart a bit pointless when used in pratice. Because as mentioned barbs,figthers and paladins are less versatile out of combat, and a barb isn't always gonna be reckless attacking, nor is a BM always tripping and a OOV paladin only got 1 channel pr short rest.

That's without mentioning that all the 3 classes have to get close to the enemy and actually survive the trade of blows to 'abuse' GWM.

Anyway im just repeating myself, rule as u see fit.

[Theodoxus]

It really annoyed me when I saw that this mechanic was added. It was said from the very start that it penalties and bonuses wouldn't be added in this manner and then they put them in anyway. If they had stuck to the methodology of the system, then the feats would've used disadvantage instead (which is mathematically almost the same) and this would've also eliminated the issues that the feat has as they primarily come from having advantage on the attacks and thus negating the -5 penalty.

GWM with disadvantage instead of -5? It would be the go to feat for every barbarian - Vuman's would dominate the barbarian landscape. Getting a normal attack, every attack with a +10 rider? Yes please.

The only way I could see it be viable and still play by the original intent, is if GWM imposed a special kind of disad that couldn't be negated by advantage. Of course, that brings reckless attack down to about as pointless as Frenzy (especially Frenzy coupled with GWM+PAM) - so it would require a lot of reworking of mechanics just to bring the one feat in line with the ad/disad mechanic. (Not to mention there isn't anything like RA for archery - the closest is the Archery fighting style, negating 2 points of the 5 point penalty - which would be horrendously not worth it if every shot were made at disadvantage. Great for fighting against, horribly frustrating if you're playing with it.

[steppedonad4]

GWM with disadvantage instead of -5? It would be the go to feat for every barbarian - Vuman's would dominate the barbarian landscape. Getting a normal attack, every attack with a +10 rider? Yes please.

Barbarians already negate the -5 with Reckless Attack, however under the current system they also have a second chance at a critical which, statistically, is a very significant DPR boost. If it were disadvantage instead of the -5, it would just be a normal attack. Ergo, the barbarian is worse off with it as disadvantage, not better.

[Kryx]

- Kryx, the thoroughness of your spreadsheet is a little alarming. I have some minor quibbles over some of your estimates (Why does Fighter Crit% spike so sharply once you get into Prone targets? I fear you might be double counting advantage. Also, I wonder if you're slightly over-estimating Prone%, seeing as how if you kill something you have to start attacking a target that's no longer Prone. Its possible you accounted for this and I missed it, but I didn't see it.), but, seriously, these are *minor* quibbles and that spreadsheet is a work of art. If you want to know how damage averages out by class, that spreadsheet is the thing to look at. Bring a pith helmet.
Thank you so much! Honestly! Hearing this makes the work much more enjoyable. I'm open to constructive criticisms with how I can improve.

Fighter Crit: Are you talking about overall crit chance? That's just for calculating the GWM bonus attack chance. At level 20 on a Fighter GWM Trip that is 23%. That is because he as 1 attack at 5% crit, and then 3 attacks at ~6.57% crit. To calculate the chance to crit overall you take 1-crit chance to the power of the number of attacks. You can see the formula. I believe that is correct and some others have looked somewhat heavily through some of the formula. Please do let me know if this is wrong.

The Fighter's attack block assumes he attacks something that is standing. He has to hit (65% chance normally) and then beat them on an athletics check which equals out to ~33% chance to prone the enemy. Are you saying that is too high? Or that you may kill them with the 1st or 2nd attack while they are prone and the 3rd won't have the target be prone? If so that is indeed possible, but truthfully his total DPR is ~13-19% of an average monster's HP so I think that isn't of great concern.


GWM is only barely a positive for Paladins.
Paladin does get a lot from Improved Divine Smite, so -5/+10 is naturally worth less. However I do not take into account bless because it is an action spell. I haven't been able to model action based spells as they'd depend on pre-combat buffing which is quite iffy.
If Paladin took into account bless -5/+10 would be much better for him.


- Barbarians are the real culprit. Yes, BM fighters get good mileage out of GWM, but not like Barbarians do. For Barbarians, mostly thanks to Reckless attacks, GWM is an unalloyed plus that just keeps on giving. Because they get perma-Advantage, they actually get a little less mileage out of magic weapons than other classes.
Barbarians are awesome. Resistance to most damage, huge dmg from reckless attacks. They have been very impactful in my games. I'm not sure I agree that they are OP though.


- Melee classes scale really, really weirdly. They scale up incredibly fast at low levels, and then just coast to level 20. Barbarians and Paladins both do ~55% as much damage at level 5 as they'll do at level 20 (not accounting for improved magic items). A Barbarian gains more raw damage output from going from level 3 to level 5 than they do from level 5 to level 19. And if you assume the Barb goes from a normal weapon at level 3 to a +1 weapon at level 5 to a +2 weapon at level 19 its even worse (though a +3 weapon at 19 kicks the advantage back the other way).
This is greatly diminished w/o -5/+10. My Houserules for example scale differently.


2) Bump down Fighters' superiority dice from 4 to 3 to start, then increase by 1 at levels 7 and 15 as before
Agreed. The dip potential is really high. I do the following: Battlemaster maneuvers: gain 2 dice at level 3, 2 more at 5. Learn 2 maneuvers, and an additional 1 at 5.


3) Don't worry too much about level 20 balance. Everyone at level 20 will be a juggernaut in their own way but they should also all be within days of enforced retirement.
Mostly agreed. A lot of people say "High levels everything is OP", but I find that silly. There should still be some balance.
I include multiple tiers (5,11,17,20) when I do comparisons to try to give a better impression.


Honestly, Barbarians and Paladins should probably be reworked in some fashion to be less front-loaded (Barbs especially), but that's beyond my pay-grade.
Without -5/+10 Paladins are actually incredibly strong. Especially OoV Paladins. They have burst and sustained damage via IDS. It's a bit too much for a class that also has heals and amazing auras imo.




My group has considered a cpuple alternatives to -5/+10:
- half feat (+1 str/dex)
- add half prof bonus to weapon damage when used in two hands
- add full prof bonus to weapon damage when used in two hands
- add +2 to weapon damage when used in two hands

The half feat solution is boring, so of the other three which pne do you think would be closer balanced? I'm leaning towards prof bonus myself. Not asking for a full math and chart response mind you, just the opinion pf one more mathematically inclined than myself.
Of those I believe the half feat to be the most balanced. If you must go with another I'd suggest half prof to weapon damage. Full prof would be way too much.




Bottom line, I want equally (at least as much as possible) supported options. And -5/+10 takes that away for some specific cases, as the OP mentioned.
This is the core of the issue to me. It limits options for those who care about the math.




At the end of the day Kryx said it pretty clearly in the original post. If you don't think that GW barbs, figthers and paladins should be the kingpins of melee DPR by a large margin that's your opinion - and that's fine, alot of us just don't agree with that.
Removing -5/+10 doesn't stop GWM Barbs, Fighters, and Paladins from being the kingpins. They still do significantly more damage than other builds.
Phrasing the argument that way biases it.

[djreynolds]

Mr Kryx, I always like your threads. Very cool.

You stat can never go above 20, with exception of barbarian and archery style granting + 13. Once the game gets to higher levels gwm and sharpshooter can be outclassed by magical armor, at least for the bosses you face.

So I find sometimes accuracy is more important in favor of 10 extra points of damage.

If the players are churning through baddies like butter, up their AC to a point where its a choice but not out of reach for gwm. Put baddie bards in their ranks, and give the warriors the magic adept feat for shield, or defensive duelist, etc.

This may mitigate that damage gwm can inflict but not take away from player concepts. Those paladin's of vengeance may find now that bless is better used than hunter's mark.

[Lollerabe]

So a lvl 5 paladin with 18 strength, a maul and GWF deal alot more damg than a lvl 5 paladin with 16 strength, SB, duelist and shieldmaster?

I really don't try to rephrase you in a bias way, by a large margin I meant the one represented by GWM, maybe I should've made that clear.

Can I ask you though Kryx, does anyone actually use GWs at your table besides polearms? Do they still pick up the (IMO) lackluster +1strength bonusaction on KO's/crits feat?

Edit: And how does it limit options for those who value math? Again more options dosen't add more limits. If anything it goes against the instinct of those who care about math, and min/maxing damage. To outright claim it limits you're options is just not true.

I feel like you got to stick to your guns here, if saying "GWM limits player options" isn't silly then saying "every single full caster becomes OP" isn't silly either.

[Kryx]

Kryx, I always like your threads. Very cool.
Thanks man! :D


So I find sometimes accuracy is more important in favor of 10 extra points of damage.
Sometimes that is the case. Based on the averages presented in the DMG using -5/+10 will be better than not using it on the average.

Your other tactics are ways of mitigating the issue. I would prefer that the issue not exist in the first place instead of me designing encounters specifically the mitigate the issue. That seems silly to me.

[Kryx]

So a lvl 5 paladin with 18 strength, a maul and GWF deal alot more damg than a lvl 5 paladin with 16 strength, SB, duelist and shieldmaster?
GWF = Great Weapon Fighting which is the Fighting Style
GWM = Great Weapon Master which is the feat

I assume you mean the feat since we aren't discussing GWF? I'll also assume 16 str since he's taking a feat.

Paladin actually loses DPR at most levels from -5/+10 without bless. Paladin is an edge case that my model can't handle to well currently as bless is an action spell which would require pre-buffing

By my numbers:
5:
GWM Paladin: 21 DPR
GWM Paladin w/o -5/+10: 20 DPR
GWM Paladin OoV: 29 DPR
GWM Paladin OoV w/o -5/+10: 27 DPR
S&B Paladin: 17 DPR
S&B Paladin OoV: 22 DPR


I really don't try to rephrase you in a bias way, by a large margin I meant the one represented by GWM, maybe I should've made that clear.
I posted the numbers in the OP. You can see how much a Barbarian -5/+10 build beats other Barbarian builds. It's also true for Fighter to a lesser extent.


Can I ask you though Kryx, does anyone actually use GWs at your table besides polearms?
GWs? Great Weapons? Currently I have the following classes:
Hunter Ranger w/ Heavy Crossbow
Moon Druid
Swashbuckler Rogue
Awful melee Wizard
Bladelock doing S&B

Not having any Greatswords or Polearms isn't a result of my changes though. I had one Barbarian earlier who was using a Greatsword, but he had to stop playing. In the test phase I had a Polearm wielder, but he swapped to a cleric who lost his captaincy and then retired the character who now plays the moon Druid. The Awful melee Wiz used to play a EK Greatsword build, but got marooned after burning up a tower the party had worked hard to build. He then built a Greatsword bladelock who was awesome until he killed himself because he didn't like the fluff of the warlock control (lame).
I'm very open with my players and let them have the final say in a lot of houserules. In the -5/+10 case I actually recently tried to reimplement it due to the math in some cases, but was told it wasn't a fun option and they don't want it.


Do they still pick up the (IMO) lackluster +1strength bonusaction on KO's/crits feat?
Cleave is a great option. By my calculations it'll happen ~30% of the time, depending on the build. It's a pretty damn good DPR boost. Though I've turned it into a default weapon property for greatsword just recently. You can see my full houserules (https://docs.google.com/document/d/112evwX4-QFfkLlAEq8UDKLREazWlCpvAg6Ek_16752A/edit#heading=h.uo2lldwx4402). The purpose of turning some feats into weapon properties is multipurpose, but one of the main ones is to allow weapon swapping. So picking up a polearm or greatsword or shield depending on the situation is an option. Another is to remove somewhat of a "tax" which every decent offensive build must take. Therefore they can focus more on ASIs or other feats. That part of the goal is also helped by having split feats.

[Lollerabe]

Thanks for the indepth reply :)

No I meant GWF as the style, as in "since GWM is not an option how much more damg would the GWF paladin do than the SB pala"

Since you claimed that even without GWM, GW wielders are still the kingpins by a large margin.

But I assume the 27 DPR of the non -5/+10 still includes the crit/ko bonus action then?

So my previous post was without errors :)

[djreynolds]

Your stuff is great. Always is.

I find upping the ante is just preferred. My kid plays a wood elf ranger with sharpshooter. Crazy at low levels. But if I take that away. Tears.

But she doesn't know that the bad guys hired mercenaries. They have an Archer as well or a cleric or a cave troll. Throw in a damage sponge. Goblin barbarian.

It is like putting your fingers in a damm. Agreed. But players will always try to find an advantage. Otherwise they are un-invested in the game. I want the players to earn victory.

[Kryx]

GWM is still an option, just without -5/+10.

My claim is only substantiated on Barb and to a lesser extent fighter. It likely exists on Paladin (besides OoV) as well due to bless, but I don't have the numbers to back that up.
Bless can also help a whole party when cast by a cleric.

[Lollerabe]

And the prone condition from shieldmaster can also benefit a whole party, but that's neither here nor there.

Alright if the numbers aren't there they aren't there.

Thanks for all the work you have done regardless. Much like the thread on GWF rerolls on additives I have not changed my mind, but really do appriciate your dedication and input, cheers.

[Corran]

Edit: And how does it limit options for those who value math? Again more options dosen't add more limits. If anything it goes against the instinct of those who care about math, and min/maxing damage. To outright claim it limits you're options is just not true.
Say you have option A and option B, and let those two options be equally good and balanced in term to what they offer (even if they offer completely different things). If option A can be improved significantly by applying feat X, and there is no equivalent feat/feature/etc that can do the same for option B, then immediatelly it follows that option B is no longer viable. You can always go ahead and pick option B, so technically feat X is not limiting your options, but the fact that option B is now significantly lacking compared to option A, practically limits your options. And I am not talking about sub-optimal characters or minmaxing here, I am talking about pure imbalance. If you dont accept the ''limits your options'' phrase, then read it as ''creates for non equally-supported fighting styles''.

In the case of the recklessly attacking barbarian and GWM, the SnB barbarian essentially becomes a joke. Well, that was perhaps a bit too harsh. To be more specific, it becomes a less well-supported option for the class. And that is something I am at odds with.

Someone with a fairly basic knowledge of probability can see that before playing the character. Anyone else will come to understand that after a couple of sessions playing the character. No matter which is the case, the problem still remains the same.

[Soular]

Completely agree with the OP. Let us discuss a hypothetical. Say I wanna play a barbarian. And because I watched a lot of that vikings series or whatever, I want my barbarian to fight with an axe and a shield. Taking out the GWM feat, I can go on and play my SnB barbarian without looking back at when I created this character with remorse. He can go on and have his glorious moments, and share the spotlight equally with all other members of the group. Everyone had fun during the encounter, balanced pcs, end of story. Now lets allow the GWM feat. Reaching 4th level (or from 1st level if human variant), GWM is an option. As I went with SnB, I will not select it obviously. I can go on and play my SnB barbarian as before, but since I am a math guy I will know that my barbarian would be considerably better if he was just wielding a two hander and had the GWM feat. By which I mean, the trade off of having a shield and a feat/ASI to spend wherever I want to, will not make up for all that extra damage my character would do if he was using a two-hander and the GWM feat while attacking recklessly. I know people will say that one should not look at it the way I just presented it, but knowing some basic probability will make it a hard choice for me to select the fighting style I intended for, as it is very sub-optimal. I want to be able to choose a build for my character that suits my preferences, without having to spend too much time thinking if I should do so because it is way worse (and hence not as fun to play - take away balance, hurts fun in some way) than another build. Bottom line, I want equally (at least as much as possible) supported options. And -5/+10 takes that away for some specific cases, as the OP mentioned.

As a S&B Fighter lover myself, I fail to see the problem. Yes you will be down on damage, but that is the price you pay for consistency and resilience. When fighting minions, some of the GWM awesomesauce will be spilled on the ground and wasted when dispatching a mook with just a few hitpoints. Other times, those gnarly swings could result in some equally gnarly whiffs at the wrong moment.

By comparison, the ax and shield barbie will deal damage as regularly as a metronome, enjoy the added AC of a shield, and swap weapons quickly should the type of enemy warrant it.

This part is a bit meta-gamy, but I would like to point out that magical one-handed weapons are probably more attainable than heavy weapons. And in a story driven game it's quite possible to lose your weapon somehow, and have to utilize whatever is on hand.

Finally, last Saturday we were playing 2E. Our group of four adventurers was challenged to pick a champion to fight a challenger in single combat; a trial by combat. We have two martial characters that can be absolutely devastating in combat, much like the GWM, but it was the S&B Fighter's superior damage avoidance and reliable DPM that we trusted our lives to. The same principal applies to 5E.

The tortoise does often win the race.

[Lollerabe]

Yes Corran I get that, but in your very specific case GWM isn't really the main culprit. It's the fact that anything but heavy weapons is meh for barbs, shieldmaster is lackluster as the major draw - chance of advantage on attacks as a bonus action, holds little to no merit when you got reckless.

A 2ac bump is also 'less' important when you have resistance to physical/all damage.

And brutal critical works better with higher damage dice.

So again SB barbs is just not good, GWM or not. Hell remove GWM and polearm master is still waay better for a barb than SB.

Unless you also remove that aswell, in which case you are trying to force an inferior build to be on equal footing with every other build.

To me it's the same as wanting smite reworked because it dosen't work on ranged attacks, and I want to play a bowadin!

You are trying to force a playstyle that dosen't suit the class, and then you try to justify that by removing options to make that inferior build less inferior.

[DanyBallon]

Say you have option A and option B, and let those two options be equally good and balanced in term to what they offer (even if they offer completely different things). If option A can be improved significantly by applying feat X, and there is no equivalent feat/feature/etc that can do the same for option B, then immediatelly it follows that option B is no longer viable. You can always go ahead and pick option B, so technically feat X is not limiting your options, but the fact that option B is now significantly lacking compared to option A, practically limits your options. And I am not talking about sub-optimal characters or minmaxing here, I am talking about pure imbalance. If you dont accept the ''limits your options'' phrase, then read it as ''creates for non equally-supported fighting styles''.

In the case of the recklessly attacking barbarian and GWM, the SnB barbarian essentially becomes a joke. Well, that was perhaps a bit too harsh. To be more specific, it becomes a less well-supported option for the class. And that is something I am at odds with.

Someone with a fairly basic knowledge of probability can see that before playing the character. Anyone else will come to understand that after a couple of sessions playing the character. No matter which is the case, the problem still remains the same.

Is option A and B are equally fun concept to play whatever the math is? Then it's all that matter. Fun should be the most important aspect of the game.

Nothing prevent you from having the most fum playing the most effective damage dealer though

[Corran]

Yes Corran I get that, but in your very specific case GWM isn't really the main culprit. It's the fact that anything but heavy weapons is meh for barbs, shieldmaster is lackluster as the major draw - chance of advantage on attacks as a bonus action, holds little to no merit when you got reckless.

A 2ac bump is also 'less' important when you have resistance to physical/all damage.

And brutal critical works better with higher damage dice.

So again SB barbs is just not good, GWM or not. Hell remove GWM and polearm master is still waay better for a barb than SB.

Unless you also remove that aswell, in which case you are trying to force an inferior build to be on equal footing with every other build.

To me it's the same as wanting smite reworked because it dosen't work on ranged attacks, and I want to play a bowadin!

You are trying to force a playstyle that dosen't suit the class, and then you try to justify that by removing options to make that inferior build less inferior.
Ok, you are right. Actually, it was a mistake on my part to make it into a contest of barbarian playstyles, as that was not what the OP was aiming at, but rather the difference in dpr between different dpr-intended (use that term loosely) classes.

[Kryx]

Much like the thread on GWF rerolls on additives I have not changed my mind, but really do appriciate your dedication and input, cheers.
That thread actually made me change my mind. I did the comparative math and it's actually necessary to apply on additive rolls to be worth it in comparison to Defensive fighting style.




Ok, you are right. Actually, it was a mistake on my part to make it into a contest of barbarian playstyles, as that was not what the OP was aiming at, but rather the difference in dpr between different dpr-intended (use that term loosely) classes.
Actually, I entirely agree with what you're saying and it was one of my points. Barbarian is the best case to look at -5/+10 to see the differences.
Barbarian S&B doesn't compare at all to GWM. Mainly because S&B's benefit is proning for advantage, which a Barbarian already has.

But that difference is largely diminished w/o -5/+10.

5:
Barb S&B: 21 DPR
Barb GWM: 33 DPR
Barb GWM w/o -5/+10: 26 DPR

11:
Barb S&B: 23 DPR
Barb GWM: 36 DPR
Barb GWM w/o -5/+10: 30 DPR

17:
Barb S&B: 36 DPR
Barb GWM: 47 DPR
Barb GWM w/o -5/+10: 41 DPR

20:
Barb S&B: 42 DPR
Barb GWM: 58 DPR
Barb GWM w/o -5/+10: 47 DPR

S&B is still quite behind GWM, but at least it's competitive.

In the end without -5/+10 the comparison between all the options is much more inline with the default DPR. I think that's a good thing. Still some large differences, but those differences are more inline with an ASI.

[octavius_maximu]

I've had too many combats where turn 1, my Wizard misses a Cantrip and then the Fighter kills half the enemies himself before I get another turn because of the combination of Battlemasters being able to take a D8 after the fact, killing + Extra attack when you killl something, more attacks because fighter and easily dispatching smaller enemies easily because of the +10 damage.

[Lollerabe]

So let me get this straight - at lvl 20 a barb with the nerfed GWM feat deals a measly 5 more DPR than a SB barb with no feat used? That's what you consider fair or in line?

This is the class that's completely flavored to favor heavy weapons, with all due respect that seems horrible.

Hell even with the standard GWM dps of 58 compared to the 42 of the SB barb im not seeing a problem.

Again - this is the class that uses GWs the absolute best.

I know you changed your mind in the GWF reroll thread, im saying I didn't - and frankly with those numbers you just showed I won't now either :)

But hey again - this seems to be more about me and you just disagreeing on how big of an egde GWs should have over other options, on the one class the can 'abuse' GWs the most.

[Sredni Vashtar]

What about just adding in -5/+10 options for the other combat styles? It's the Incredibles "if everyone is special, then nobody is" thing.

[Corran]

+5 damage compaired to +2 AC, having in mind several class features that you mentioned above (like the fact that +2 to AC is not great considering the resistance to damage is your main source of defense, or that brutal critical works better with a two hander) looks fairly balanced to me. But for a moment let me accept that a great weapon wielding barbarian should do much better than a SnB barbarian, just because that's what this class is all about (ie heavy weapons). I dont like that to be the case, I would like for the mechanics of the game to allow for an equally good SnB or TWF barbarian, but let us assume that it is logical for a barbarian wielding a two hander to outshine every other barbarian build by default. Now look at the difference in dpr between said great weapon wielding barb with the GWM feat and some other classes. Take for example a melee rogue. A rogue cannot tank like a barbarian, it is not even close (nor should it be). But where does that leave the rogue when a fight takes place. If he cannot at least try to match the barbarian's dpr, what is his purpose in combat? Yes, a rogue can do numerous other things a barbarian cannot do outside combat (even during combat in some extremelly rare circumstances, like having to disarm a trap or unlock a door during combat, etc, but that will hardly be the case in a fight). But is this what we want for the roles the classes fulfill? To have some classes that are combat oriented and others that are not? I personally would like to see every single class contribute more or less up to the same effect during combat. That way everyone can be happy during encounters. I guess my point is, let the player who wants to play a concept choose the class that better encompasses this concept, without having to double-check if his choice would make encounters for him boring. So my point is this. If a GWM fighter/barbarian/OoV Paly dominates the combat in term of hurting the monsters, what is left for a rogue/monk/other squishy dprs that cannot benefit from a -5/+10 feat to do during combat?

[Saggo]

Is option A and B are equally fun concept to play whatever the math is? Then it's all that matter.
Arguably, no, fun is not all that matters. Math is the only language we can agree on (and even then everyone still manages to argue about it), and balance achieved through math is an accepted goal for multiplayer games. This game is math. Even roleplaying scenarios are often resolved through the use of mathematical probability.


To clarify, I'm not saying anyone gets to tell you what is or isn't fun for you or your table, nor should anyone get to tell me. But we can't go table to table. We need a medium here. For a case example, I've always enjoyed TWF and Ranger builds, ever since my wee days as a 2e and 3.5e player. It frustrates me that I have to consciously choose to lower my DPR to build a TWF Ranger just to play my personal iconic build. I could come here and say I don't like TWF Ranger because it doesn't feel as fun where we just trade anecdotes, or come here and say it does 32-50% less DPR over Sharpshooter Rangers for little to no additional utility where everyone can understand a baseline for discussion.

[Finieous]

As far as the rogue is concerned, I don't think Kryx's builds or strategies have been updated or optimized to the same degree as the heavy-weapon builds. For example, the TWF rogue has a 78% sneak attack chance across the board. The swashbuckler will pretty much always be able to sneak attack, unless he's facing three or more opponents alone (facing two alone, his offhand attack may not be eligible for sneak attack unless he eats an AoO; he attacks #2 without sneak attack eligibility to freely move out of #2's reach). Change the sneak attack chance to 95%, he's up to 45 DPR at 20 compared to a GWM barbarian at 58. I'm not sure how you land on an assumption for reaction attacks with Sentinel, but give him that feat for double sneak-attack opportunities and you're in business. Given the rogue's options in the other two pillars of the game, that seems fair to me. Granted, he still won't match the DPR of a barbarian who has both GWM and PM in the end-game. If that's still a concern for you, I'd suggest your problem is really with the combination rather than with -5/+10 or polearms by themselves.

[DanyBallon]

Arguably, no, fun is not all that matters. Math is the only language we can agree on (and even then everyone still manages to argue about it), and balance achieved through math is an accepted goal for multiplayer games. This game is math. Even roleplaying scenarios are often resolved through the use of mathematical probability.


To clarify, I'm not saying anyone gets to tell you what is or isn't fun for you or your table, nor should anyone get to tell me. But we can't go table to table. We need a medium here. For a case example, I've always enjoyed TWF and Ranger builds, ever since my wee days as a 2e and 3.5e player. It frustrates me that I have to consciously choose to lower my DPR to build a TWF Ranger just to play my personal iconic build. I could come here and say I don't like TWF Ranger because it doesn't feel as fun where we just trade anecdotes, or come here and say it does 32-50% less DPR over Sharpshooter Rangers for little to no additional utility where everyone can understand a baseline for discussion.

Math is only important for theorization, it doesn't stand in a home game. If you're having fun playing a twf ranger, then who cares that it does less damage per round that an sharpshooter ranger, for the simple reason you are not playing a twf ranger against a sharpshooter ranged ranger, you are playing a twf ranger in a party that have different strength and weakness.
If D&D was about playing against (or with) thousands of character like in a MMORPG, then you're right playing a less efficient built may/will result in less fun, but usally a typical D&D game rarely reunite more than a handful of players at a time, so you don't need to be uber effective to have fun playing your character concept.
As an example, I'm playing a EK in the stater set with friends of mine. My character is mostly a meat shield for my party while our ranger (who don't have the Sharpshooter feat yet), is our long range master. While I can do decent damage with my sword or cantrip, our Bladelock is dealing far more damage than I do, but if it weren't of my character to take most damage, our warlock, wouldn't last long. In the mean time, we have a bard that can do a bit of everything (and fill in in arcana and investigation checks as I have a tendency to roll 3 or less on skill checks, as oppose to high roll on attack...:smallbiggrin:) and our CN tempest cleric is some kind of walking boomstick. No one is playing an optimized character, but since we complete each other, we are an effective group in the end, and mostly, we all have fun playing our characters.

[Kane0]

What i'd like to see is the new cantrips (Booming Blade/GFB) when used with a rogue/bard/warlock on that chart.

[Finieous]

What i'd like to see is the new cantrips (Booming Blade/GFB) when used with a rogue/bard/warlock on that chart.

The killer for the rogue is that he loses his bonus action attack with the spells, which means he loses his second chance to land sneak attack every round. Otherwise, booming blade would be a thing of beauty for a swashbuckler rogue.

[Shaofoo]

Arguably, no, fun is not all that matters. Math is the only language we can agree on (and even then everyone still manages to argue about it), and balance achieved through math is an accepted goal for multiplayer games. This game is math. Even roleplaying scenarios are often resolved through the use of mathematical probability.

Actually the DM has final say in what happens, even when dice are rolled DMs can fudge the results around or even blatantly ignore them or change goal posts after the roll so to speak.

Math isn't the end all be all, if the DM allows then math can express itself but the DM likewise can close the door on math if he wishes on certain things. Maybe he wants the final blow to happen to the BBEG so he declares an auto hit instead of rolling dice.

And the problem is that you can't balance the entire game through math, at least when you consider all facets of the game. How can you compare increased skill effectiveness over combat effectiveness? Can you mathematically chart the value of a Linguist or Actor feat in all points? Math is important but it is far from the end all be all. D&D is much more complex than what simple math can do, starting with the fact that the DM is the biggest variable in the game. A skill heavy and combat light game renders all DPR calculations moot when it comes to effectiveness in the game.

You can't math all of D&D, to try to math D&D you should math DMs, can you mathematically chart a human being?

[Malifice]

Completely agree with the OP. Let us discuss a hypothetical. Say I wanna play a barbarian. And because I watched a lot of that vikings series or whatever, I want my barbarian to fight with an axe and a shield. Taking out the GWM feat, I can go on and play my SnB barbarian without looking back at when I created this character with remorse. He can go on and have his glorious moments, and share the spotlight equally with all other members of the group. Everyone had fun during the encounter, balanced pcs, end of story. Now lets allow the GWM feat. Reaching 4th level (or from 1st level if human variant), GWM is an option. As I went with SnB, I will not select it obviously. I can go on and play my SnB barbarian as before, but since I am a math guy I will know that my barbarian would be considerably better if he was just wielding a two hander and had the GWM feat. By which I mean, the trade off of having a shield and a feat/ASI to spend wherever I want to, will not make up for all that extra damage my character would do if he was using a two-hander and the GWM feat while attacking recklessly. I know people will say that one should not look at it the way I just presented it, but knowing some basic probability will make it a hard choice for me to select the fighting style I intended for, as it is very sub-optimal. I want to be able to choose a build for my character that suits my preferences, without having to spend too much time thinking if I should do so because it is way worse (and hence not as fun to play - take away balance, hurts fun in some way) than another build. Bottom line, I want equally (at least as much as possible) supported options. And -5/+10 takes that away for some specific cases, as the OP mentioned.

No your S+B barbarian isn't worse off at 4th. Assume starting stats of 16. He's worse off in DPR in some corner cases than your GWM barbarian (v low ACs) but the Shield barb has +2 AC over the GWM barbarian to compensate. At 4th when the GWM barb took the GWM feat, the board barbarian bumps Con.

In a one on one fight at 5th level, the axe and board barbarian wins. +6 to hit v AC 16 = 55% of 19 = 10.45 DPR. +1 to hit v AC 19 = 15% of 43 = 6.45 DPR.

Taking advantage into account helps the GWM fighter better (and brings the DPR to about even) but the numbers are still on the side of the 18 Con S+B Barbarian 5 in a one on one fight (he has more HP for starters).

[Saggo]

Math is only important for theorization, it doesn't stand in a home game. If you're having fun playing a twf ranger, then who cares that it does less damage per round that an sharpshooter ranger, for the simple reason you are not playing a twf ranger against a sharpshooter ranged ranger, you are playing a twf ranger in a party that have different strength and weakness.
If D&D was about playing against (or with) thousands of character like in a MMORPG, then you're right playing a less efficient built will result in less fun, but usally a typical D&D game rarely reunite more than a handful of players at a time, so you don't need to be uber effective to have fun playing your character concept.
That theorization is where the game mechanics were born from. I'm not at your table, and you're not at my table. If not the only way, math is the most objective way to discuss this game with other players that isn't anecdotal stories.


Actually the DM has final say in what happens, even when dice are rolled DMs can fudge the results around or even blatantly ignore them or change goal posts after the roll so to speak.

Math isn't the end all be all, if the DM allows then math can express itself but the DM likewise can close the door on math if he wishes on certain things. Maybe he wants the final blow to happen to the BBEG so he declares an auto hit instead of rolling dice.

And the problem is that you can't balance the entire game through math, at least when you consider all facets of the game. How can you compare increased skill effectiveness over combat effectiveness? Can you mathematically chart the value of a Linguist or Actor feat in all points? Math is important but it is far from the end all be all. D&D is much more complex than what simple math can do, starting with the fact that the DM is the biggest variable in the game. A skill heavy and combat light game renders all DPR calculations moot when it comes to effectiveness in the game.

You can't math all of D&D, to try to math D&D you should math DMs, can you mathematically chart a human being?
I'm well aware of what the DM does. We can reasonably assume the DM, as I said, often resolves scenarios with dice rolls.

Linguist and Actor feats are mechanically defined by their ability score increases, their advantage rolls, and their skill checks. These are demonstrably measurable with numbers. You can correlate their successes against advantage and skill throws. I can't find a single feat that doesn't include a measurable mechanical improvement.

Not that that matters. I never said "Math is the end all be all." I said it was a medium. Objective and necessary.

[Mrglee]

Honestly, if it wasn't for PM, GWM wouldn't be quite as problematic. Like, looking at the DPR doc, GWM Barb is 58 DPR, PM Barb is 57, and both is 73. Like, two feats that normally gives 5 to 10 more DPR, give over 20 when combined together, and really make other options just seem so much weaker.

[DanyBallon]

That theorization is where the game mechanics were born from. I'm not at your table, and you're not at my table. If not the only way, math is the most objective way to discuss this game with other players that isn't anecdotal stories.


I'm well aware of what the DM does. We can reasonably assume the DM, as I said, often resolves scenarios with dice rolls.

Linguist and Actor feats are mechanically defined by their ability score increases, their advantage rolls, and their skill checks. These are demonstrably measurable with numbers. You can correlate their successes against advantage and skill throws. I can't find a single feat that doesn't include a measurable mechanical improvement.

Not that that matters. I never said "Math is the end all be all." I said it was a medium. Objective and necessary.

Theorization may be where the mechanics come from, but they were refined by actual playtesting.
Don't blame the mechanics (that play just well in most home game) if DPR disparancies between classes creates a problem on a internet discussion forum. Instead, blame the over exposure of theorization, that create virtual problems, that just don't exist in normal play.

Sincerely, tell me how playing a sharpshooter ranger over a twf ranger (or GWM barb vs S+B barb, etc.) will give you more fun, if you are already have the best DPR in your party?

[Corran]

Sincerely, tell me how playing a sharpshooter ranger over a twf ranger (or GWM barb vs S+B barb, etc.) will give you more fun, if you are already have the best DPR in your party?
If the twf ranger doesnt get to kill enemies and instead gets knocked unconscious very often, I guess the player playing that character wont enjoy the encounters a lot. Dpr and generally combat isnt everything in a dnd campaign, but having a character that is way worse than everyone else isnt fun either. Remember, getting some kills is always fun and makes your character feel important, especially if you went about creating a character with the hope of killing things once in a while (eg twf ranger).

[Saggo]

Theorization may be where the mechanics come from, but they were refined by actual playtesting.
Don't overblow theory more than what it is. D&D is applied statistics (math) and playtesting is cataloged statistics.


Don't blame the mechanics (that play just well in most home game) if DPR disparancies between classes creates a problem on a internet discussion forum.Not a thing I was doing.


Sincerely, tell me how playing a sharpshooter ranger over a twf ranger (or GWM barb vs S+B barb, etc.) will give you more fun, if you are already have the best DPR in your party?
It's a case study, of a well established problem with 5e, as an example of using math to initiate discussion with a common understanding. Not a review of how fun a particular fighting style and feat are.

Fun is anecdotal, math is not. It adequately fulfills several needs, including objective baselines for discussion.

But killing things is fun, killing things harder and faster can be more fun. Surprisingly, some people don't think ranged attacking is fun, some people think melee and ranged are equally fun. Whatever the case, cause it will vary wildly from table to table, people that choose a particular class and choosing a fighting style it supports will be punished with less DPR for little to no gain and math is a way of discussing that.

To wit, math is a way of discussing that, it means the same thing at every table.

[numerek]

No your S+B barbarian isn't worse off at 4th. Assume starting stats of 16. He's worse off in DPR in some corner cases than your GWM barbarian (v low ACs) but the Shield barb has +2 AC over the GWM barbarian to compensate. At 4th when the GWM barb took the GWM feat, the board barbarian bumps Con.

In a one on one fight at 5th level, the axe and board barbarian wins. +6 to hit v AC 16 = 55% of 19 = 10.45 DPR. +1 to hit v AC 19 = 15% of 43 = 6.45 DPR.

Taking advantage into account helps the GWM fighter better (and brings the DPR to about even) but the numbers are still on the side of the 18 Con S+B Barbarian 5 in a one on one fight (he has more HP for starters).

You seem to be assuming 16 str 16 dex 16 con to start, the only race that can pull that off is standard human.

but ok we are standard humans did hit points using flat increase per level as listed in the book

with advantage
the GWM chooses not to activate -5/+10 55 hit points
64%*(7+3+2)*2+9.75%*7*2+19%*(64%*(7+3+2)+9.75%*7)= 19.90275 halved due to resistance(not going to try to figure out how rounding down affects) 9.951375
dies in 6.8621334996880848409232688708671 rounds

axe+board 60 hit points
79.75%*(4.5+3+2)*2+9.75%*4.5*2=16.03 halved 8.015
dies in 6.0293175566190601801258620039944 rounds

GWM lives for 4/5ths more rounds than the other guy

without advantage
the GWM chooses not to activate -5/+10 55 hit points
40%*(7+3+2)*2+5%*7*2+9.75%*(40%*(7+3+2)+5%*7)=11.3 0425 halved due to resistance(not going to try to figure out how rounding down affects) 5.652125
dies in 10.091743119266055045871559633028 rounds

axe+board 60 hit points
55%*(4.5+3+2)*2+5%*4.5*2=10.9 halved 5.45
dies in 10.61547648008492381184067939049510.09174311926605 5045871559633028 rounds

So without advantage the axe+board does survive an extra half round
With advantage against just one enemy the GWM feat is adding 9.5% more dpr without using the -5/+10, when the bonus action happens more often from killing something it adds even more.

Also the GWM build could have a 17 ac with medium armor and then wouldn't need a 16 in dexterity and wouldn't be constrained to the standard human race.

edit: realized I got the dies in swapped for without advantage, doesn't really matter because as long as one participant wants advantage they both get it so the one that wins that comparison wins the fight.

[Malifice]

You seem to be assuming 16 str 16 dex 16 con to start, the only race that can pull that off is standard human.

but ok we are standard humans did hit points using flat increase per level as listed in the book

with advantage
the GWM chooses not to activate -5/+10 55 hit points
64%*(7+3+2)*2+9.75%*7*2+19%*(64%*(7+3+2)+9.75%*7)= 19.90275 halved due to resistance(not going to try to figure out how rounding down affects) 9.951375
dies in 6.8621334996880848409232688708671 rounds

axe+board 60 hit points
79.75%*(4.5+3+2)*2+9.75%*4.5*2=16.03 halved 8.015
dies in 6.0293175566190601801258620039944 rounds

GWM lives for 4/5ths more rounds than the other guy

without advantage
the GWM chooses not to activate -5/+10 55 hit points
40%*(7+3+2)*2+5%*7*2+9.75%*(40%*(7+3+2)+5%*7)=11.3 0425 halved due to resistance(not going to try to figure out how rounding down affects) 5.652125
dies in 10.615476480084923811840679390495 rounds

axe+board 60 hit points
55%*(4.5+3+2)*2+5%*4.5*2=10.9 halved 5.45
dies in 10.091743119266055045871559633028 rounds

So without disadvantage the GWM still survives an extra half round
With advantage against just one enemy the GWM feat is adding 9.5% more dpr without using the -5/+10, when the bonus action happens more often from killing something it adds even more.

Also the GWM build could have a 17 ac with medium armor and then wouldn't need a 16 in dexterity and wouldn't be constrained to the standard human race.

Hang on mate. So to prove -5/+10 is better... you show that you can only bring down a S+B barbarian of the same level... by NOT using it?

Doesn't this highlight the point I've been making? That's it's situationally good for some classes and builds vs specific ACs and is not unbalanced particularly considering opportunity cost?

I'f we're wearing medium armor, I'll increase STR by 2 instead of con. Run your numbers again with my higher DPR from +1 hit and damage. My AC is still 19.

It's bloody evenly matched for numbers.

[numerek]

Hang on mate. So to prove -5/+10 is better... you show that you can only bring down a S+B barbarian of the same level... by NOT using it?

Doesn't this highlight the point I've been making? That's it's situationally good for some classes and builds vs specific ACs and is not unbalanced particularly considering opportunity cost?

I'f we're wearing medium armor, I'll increase STR by 2 instead of con. Run your numbers again with my higher DPR from +1 hit and damage. My AC is still 19.

It's bloody evenly matched for numbers.

I was simply showing that your scenario was not accurate, you were claiming the axe and board build would be doing 62% more damage than the GWM when in fact it does almost 20% less. The point is with only using half of the feat this thread is about the GWM still won in a scenario not devised by someone trying to prove that GWM is overpowered but by someone trying to show otherwise.
As for your changed scenario, I'm pretty sure the outcome would be the same the difference in to hit is nearly the same, your dpr will increase slightly but you lost your +5 hit points.
Anyways if you are changing the scenario then I would take polearm master and own you, I would have a bonus action attack every round and the first round I would get an attack of opportunity unless we start 5 feet from each other or 10 feet from each other and you win initiative. Which is still in keeping with the thread although in the OP spreadsheet GWM has a small advantage in dpr at level 4 vs polearm master I would still start with polearm master and pick up GWM later, both give a bonus action attack which is very good for barbarian dpr.

[Malifice]

I was simply showing that your scenario was not accurate, you were claiming the axe and board build would be doing 62% more damage than the GWM when in fact it does almost 20% less. The point is with only using half of the feat this thread is about the GWM still won in a scenario not devised by someone trying to prove that GWM is overpowered but by someone trying to show otherwise.
As for your changed scenario, I'm pretty sure the outcome would be the same the difference in to hit is nearly the same, your dpr will increase slightly but you lost your +5 hit points.
Anyways if you are changing the scenario then I would take polearm master and own you, I would have a bonus action attack every round and the first round I would get an attack of opportunity unless we start 5 feet from each other or 10 feet from each other and you win initiative. Which is still in keeping with the thread although in the OP spreadsheet GWM has a small advantage in dpr at level 4 vs polearm master I would still start with polearm master and pick up GWM later, both give a bonus action attack which is very good for barbarian dpr.

I didn't contrive the scenario brother. I was responding to the argument that a S+B barb is worse than a GWM barb.

Numerically they come out about even. The higher AC of the S+B barb forces the GWM barb to not use his GWM. The opportunity cost of taking GWM and using a heavy weapon (no shield, strength 2 less) evens them out against each other.

GWM is only good for a limited thing (DPR) for a limited number of classes vs a limited range of ACs usually taking advantage of a situational buff.

In the right set of curcumstances, with the right weapon, in Melee combat, vs a certain range of ACs, and in the hands of the right class it grants a good spike to DPR at an opportunity cost of that class not being better elsewhere. It's a circumstantially good feat to invest in when the stars align for certain classes who focus on DPR.

I see no need to change it.

[Tanarii]

Heya Kryx. Putting my disagreements with some of your assumptions in your calculations aside (per your request in the OP), im wondering: what's the DPR contribution difference between Polearm Master and Great Weapon Master? Put another way, is the bigger problem the +10 damage, the bonus attack, or is it BOTH together that are a big issue.

Personally the Polearm Master feat bothers me more for thematic reasons (I hate the idea that every character will be a Polearm weilder), but I'm assuming the DPR contribution comes primarily from two sources in combination: GWM and Advantage.

That suggests that the house rule to fix it is make the damage bonus less if you attack with advantage. Something like: -5 to hit, +10 damage if you attack without advantage, +5 if you attack with advantage. Obviously that's counterintuitive as all hell, but mathematically it may be more balanced.

[Mrglee]

Heya Kryx. Putting my disagreements with some of your assumptions in your calculations aside (per your request in the OP), im wondering: what's the DPR contribution difference between Polearm Master and Great Weapon Master? Put another way, is the bigger problem the +10 damage, the bonus attack, or is it BOTH together that are a big issue.

Personally the Polearm Master feat bothers me more for thematic reasons (I hate the idea that every character will be a Polearm weilder), but I'm assuming the DPR contribution comes primarily from two sources in combination: GWM and Advantage.

That suggests that the house rule to fix it is make the damage bonus less if you attack with advantage. Something like: -5 to hit, +10 damage if you attack without advantage, +5 if you attack with advantage. Obviously that's counterintuitive as all hell, but mathematically it may be more balanced.

From what I see with numbers, GWM gives +1 dpr over PM, but when mixed together it is a 20+ DPR gain, and a 12ish gain on just having one of the feats.

[Kryx]

As far as the rogue is concerned, I don't think Kryx's builds or strategies have been updated or optimized to the same degree as the heavy-weapon builds.
The heavy weapon builds aren't optimized at all. They simply have 1 or 2 feats (GWM, Polearm, or Polearm+GWM).


For example, the TWF rogue has a 78% sneak attack chance across the board. The swashbuckler will pretty much always be able to sneak attack, unless he's facing three or more opponents alone (facing two alone, his offhand attack may not be eligible for sneak attack unless he eats an AoO; he attacks #2 without sneak attack eligibility to freely move out of #2's reach). Change the sneak attack chance to 95%, he's up to 45 DPR at 20 compared to a GWM barbarian at 58.
The Barbarian doesn't have any archetype as well as none of them result in direct DPR.
If we gave the Barbarian an archetype then Battlerager would blow swashbuckler out of the water with his bonus action attack (no TWF battleragers? really WotC? Uggggh).


I'm not sure how you land on an assumption for reaction attacks with Sentinel, but give him that feat for double sneak-attack opportunities and you're in business.
That doesn't work. Sneak attack still must be made with a light weapons. No polearm is a light weapon.




If you're having fun playing a twf ranger, then who cares that it does less damage per round that an sharpshooter ranger
That's the point. There are many of us who couldn't have fun because that option is significantly worse. All the players know you'd be more of an asset as a Sharpshooter Ranger.




What i'd like to see is the new cantrips (Booming Blade/GFB) when used with a rogue/bard/warlock on that chart.
Many people overestimate these new cantrips.
I ran some quick numbers for EK:
2 weapon attacks + GFB would be decent. Likely better for a class that doesn't get 3rd and 4th attacks.

GWM fighter at 11: 44 DPR
GWM fighter at 11 w/o -5/+10: 39 DPR
Greatsword fighter at 11: 2x9.1 + (.65*7.5) = 18.2 + 4.875 + 4.875 = 27.95 DPR, though ~5 of it is on a secondary target which makes it less valuable.




Actually the DM has final say in what happens, even when dice are rolled DMs can fudge the results around or even blatantly ignore them or change goal posts after the roll so to speak.
Fudging rolls is a controversial topic. Even then.. using it to solve a math inequality is pretty ridiculous..




In a one on one fight at 5th level, the axe and board barbarian wins. +6 to hit v AC 16 = 55% of 19 = 10.45 DPR. +1 to hit v AC 19 = 15% of 43 = 6.45 DPR.
This is a false comparison that uses an AC far outside the range of the DMG's average of 13 for that level. Averages. The whole point is averages.




Theorization may be where the mechanics come from, but they were refined by actual playtesting.
So how much playtesting is required to balance? You'd have to run multiple builds hundreds of times through many different scenarios and record those numbers to see any kind of scientific data. The rest is just opinion and conjecture.




Heya Kryx. Putting my disagreements with some of your assumptions in your calculations aside (per your request in the OP), im wondering: what's the DPR contribution difference between Polearm Master and Great Weapon Master? Put another way, is the bigger problem the +10 damage, the bonus attack, or is it BOTH together that are a big issue.

Personally the Polearm Master feat bothers me more for thematic reasons (I hate the idea that every character will be a Polearm weilder), but I'm assuming the DPR contribution comes primarily from two sources in combination: GWM and Advantage.
It's the mix. But Polearm Master is just the readily available example. There are several other ways to get an extra attack: Sentinel, Mage Slayer, Haste, etc.

Having +10 to damage will obviously show up the most for the option with the most attacks.

We can discuss calculations - as long as it's constructive. I do want to improve the sheet so please do point out any errors.

[GoodbyeSoberDay]

To re-state my previous opinion: If everyone is super, no one is.

Give everyone (with proficiency in the weapon they're using) the ability to trade accuracy for damage, since it's such a basic concept it shouldn't be a feat anyway. GWM/SS become half feats because you already have that benefit. This actually benefits the low-DPR classes - who tend to have low base damage - the most, and they tend to need the love anyway. Yeah it bumps PM, but PM is almost its own different thread. Adding in the extra reaction attacks you get it can easily start trumping GWM as the first feat you should take; even for Barbarians!

And best yet, the casters don't care about the overall non-casting bump, because they're still doing awesome things like turning people into T-Rexes.

[DanyBallon]

So you guys are telling me that doing a few less damage at level 20 is so much worse that it's not worth playing the character concept you wanted? And you all seem to forget that different built have different role and most importantly, you are not soloing a monster, you are playing in a party with other characters. i.e. TWF ranger may do less damage per round, but while in melee with the BBEG, your party mage can be fireballing as many mooks as he can. Sharpshooter ranger, will surely do more damage, but you'll need a tank to hold up the BBEG. Here's an other example with two melee build: GWM barbarian and S+B barbarian, the later might not do as much damage against the BBEG (about 10-15 hp less if I remember well), but it will have higher AC, this mean he will get hit less often that his GWM counter part.



That's the point. There are many of us who couldn't have fun because that option is significantly worse. All the players know you'd be more of an asset as a Sharpshooter Ranger.

Sincerely, I feel very sorry for those who can't enjoy the game, unless they have the perfect built. Does it mean they are alway playing the same race/class combo, because anything else is suboptimal? IF so, they are missing a lot from what D&D have to offer... :smallfrown:


So how much playtesting is required to balance? You'd have to run multiple builds hundreds of times through many different scenarios and record those numbers to see any kind of scientific data. The rest is just opinion and conjecture.

There's no such thing as perfect balance, sorry, there was 4e, and it was all the same with a few different sparkle on top to give you the illusion classes were different . You playtest as long as you think you need so the game feels fun at the table, not until everything is perfect, as there will always be something you miss out or that someone will complain about.
Setting sarcasm on the side, 4e was the most solid mechanically and balanced edition of D&D, yet it failed because many thought it didn't feel like playing D&D anymore.
And I'll concede the point that math is what allow us to compare between builds, and that anything else is just conjuncture and opinion. But IMHO, that's exactly whats matter the most, as it is around a table that the game is played, and it all depends on the group you are playing with and how much fun you are having, not how balanced the classes theoretically are one against an other in a particular set of condition.

I'll give you an example, my actual character is a 3rd level EK with Int 13, I'm proficient in Arcana for a total of +3 on my Arcana ability check. So mathematically I should be succeeding my roll about 65% of the time, but for some obscure reason, I've been rolling 3 or lower on 90% of my Arcana ability checks, that it's now a running gag at the table, and I've made it a feature of my character, he think he is much smarter than he really is. There's no way a discussion about probability would have predict such a thing. I could rush and boost my Int to counter my badluck, but we are actually having fun with my lousy character, so why would I change that? And it don't affect to much as we also have a bard, that enjoy more than anything else, to show me off the he knows the answers... :smallwink:

[Kryx]

So you guys are telling me that doing a few less damage at level 20 is so much worse that it's not worth playing the character concept you wanted?

Sincerely, I feel very sorry for those who can't enjoy the game, unless they have the perfect built.
Such ridiculous condescending framing of the point... *sigh*



There's no such thing as perfect balance
There is competitive build options though. That's my goal - a world where TWF isn't awful, where -5/+10 builds don't reign supreme beyond the normal levels, a world where a Sorcerer is actually a viable alternative to a Wizard, etc, etc
If you're happy with the normal world then stick with it. Let us who disagree discuss it.

[Lollerabe]

The fact that fun is more important than numbers is a moot point.
Alot of us dislike playing something we know is inferior compared to other options that focuses on the same thing. That fx would be TWF - it's a dpr based fightstyle, but it just sucks compared to other dpr fightstyles. Again IMO that is TWF needing an overhaul or a buff, not a benchmark for what other styles should do.

But this was posted on page2:
Level 11
Fighter Polearm+GWM: 40 DPR
Fighter S&B: 36 DPR
Fighter Polearm+GWM w/o -5/+10: 34 DPR
Barb Polearm+GWM: 39 DPR
Barb Polearm+GWM w/o -5/+10: 31 DPR
Bladelock Polearm+GWM: 34 DPR
Bladelock Polearm+GWM w/o -5/+10: 34 DPR
Fighter TWF: 32 DPR
Fighter TWF w/ Rend Houserule: 35 DPR
Barb S&B: 23 DPR
Rogue TWF: 27 DPR
Monk: 27 DPR

Level 17
Fighter Polearm+GWM: 54 DPR
Fighter S&B: 50 DPR
Fighter Polearm+GWM w/o -5/+10: 44 DPR
Barb Polearm+GWM: 57 DPR
Barb Polearm+GWM w/o -5/+10: 48 DPR
Bladelock Polearm+GWM: 62 DPR
Bladelock Polearm+GWM w/o -5/+10: 52 DPR
Fighter TWF: 36 DPR
Fighter TWF w/ Rend Houserule: 41 DPR
Barb S&B: 36 DPR
Rogue TWF: 36 DPR
Monk: 35 DPR

Level 20
Fighter Polearm+GWM: 68 DPR
Fighter S&B: 65 DPR
Fighter Polearm+GWM w/o -5/+10: 55 DPR
Barb Polearm+GWM: 73 DPR
Barb Polearm+GWM w/o -5/+10: 57 DPR
Bladelock Polearm+GWM: 64 DPR
Bladelock Polearm+GWM w/o -5/+10: 55 DPR
Fighter TWF: 45 DPR
Fighter TWF w/ Rend Houserule: 50 DPR
Barb S&B: 42 DPR
Rogue TWF: 39 DPR
Monk: 35 DPR[/SPOILER]

So a SB figther deals MORE damage than a polearm + GWM figther if using the non -5/+10 GWM. That's.. Well just wrong isn't it? The PAM GWM even spend 2 feats instead of 1.

So from what i've gathered Kryx didn't have the numbers to support that GWM paladins 'break' anything, nor does the figther GWM seem to - compared to fx SB.

So the real culprit and cause for all this concern is the PAM+GWM recklessly attacking barb? So a single class based entirely on melee combat, who then invests 2 feats into more potent melee combat who constantly uses a suicidal feature blows other people out of the water - in the melee DPR spectrum. Im really trying to see the problem, and I just can't.

I mean he is not even the king of the combat pillar at all times, he's the king of melee single target combat - a form of combat that isn't always feasible/there.

Anyway why not just houserule that GWM and PAM don't stack then on the bonusaction from PAM then? If you have a problem with it, looking at the GWM flavor text that even makes sense. 'Putting weight behind your attacks' is not something you do with a quick jab from the buttend of a polearm.

[Kryx]

@Lollerabe: I made a mistake in my Barbarian S&B I posted earlier. I accidentally had him using a greatsword while raging. :P
I deleted that post. Here is the updated version:

To add some clarification to the debate of "How much damage is worth how much AC" I'm going to present 2 example data sets: GWF vs Defensive Fighting Style and Dueling vs Defensive Fighting Style.
These numbers vary by level (often scaling to be lower for S&B), but lets try to look at the overall.

A Fighter S&B Dueling build does ~16-22% more DPR than a Defensive build (52 vs 45 at level 20)
A Paladin S&B Dueling build does ~10-22% more DPR than a Defensive build (35 vs 31 at level 20)
A Fighter GWM GWF build does ~10% more DPR than a GWM Defensive build (35 vs 31 at level 20)
A Paladin GWM GWF build does ~9-11% more DPR than a GWM Defensive build (37 vs 33 at level 20)
A Fighter Polearm GWF build does ~7-9% more DPR than a Polearm Defensive build (54 vs 51 at level 20)
A Paladin Polearm GWF build does ~9-11% more DPR than a Polearm Defensive build (46 vs 41 at level 20)

In practice 5e's math places +1 AC around 10-15% more DPR. Therefore +2 AC should be somewhere around 20-30% more DPR.

Now looking at the complaint earlier, lets compare Barbarian's options:
With -5/+10:
http://i.imgur.com/EvH6Q4y.jpg

Without -5/+10:
http://i.imgur.com/PdlW7Ja.jpg

As you can see either way Polearm is doing ~65% more damage than a S&B build across all levels
GWM in the -5/+10 version does between ~60-90% more damage than a S&B build depending on the level.
GWM without -5/+10 does between ~35-65% more damage than a S&B build depending on the level.
Polearm+GWM in the -5/+10 version does between ~70-111% more damage than a S&B build depending on the level.
Polearm+GWM without -5/+10 does between ~25-64% more damage than a S&B build depending on the level.

Overall the graph without -5/+10 is more in line with other tradeoffs, but still offers huuuuge damage increases over S&B.

Though I think this also makes it evident that Polearm is too strong.





So a SB figther deals MORE damage than a polearm + GWM figther if using the non -5/+10 GWM. That's.. Well just wrong isn't it? The PAM GWM even spend 2 feats instead of 1.
Those numbers are incredibly wrong. No idea where he got those. S&B fighter is doing the following DPR:
19 DPR @5
36 DPR @11
41 DPR @17
52 DPR @20

[DanyBallon]

There is competitive build options though. That's my goal - a world where TWF isn't awful, where -5/+10 builds don't reign supreme beyond the normal levels, a world where a Sorcerer is actually a viable alternative to a Wizard, etc, etc
If you're happy with the normal world then stick with it. Let us who disagree discuss it.

Competitive where? Unless you're playing in a large enough group to have more than one character filling the exact same role, you won't see any difference. And even if there's more than one character filling a similar role in your party, the end goal is not being the best, it's overcoming challenges as a group. Comparing DPR in a bubble like you (not you as in yourself, but you as a group) are doing, is arguing for the sake of arguing.

But hey your right! I'll leave you arguing all you want if it's what makes you happy. Like I said, it's fun that matter the most, and you sure seems to be having fun with number crunching. Good for you :smallsmile:

Have a nice discussion

P.S. Don't get me wrong, me too I like a good discussion, and even enjoy arguing for the sake of arguing, but when my posts get snarky, I know that's time for me to leave the discussion, as fun is fading away for everyone the more I get snarky.

[Kryx]

That fx would be TWF - it's a dpr based fightstyle, but it just sucks compared to other dpr fightstyles. Again IMO that is TWF needing an overhaul or a buff, not a benchmark for what other styles should do.
Agreed. I have some options, but I need to do more math to find a great final solution


So from what i've gathered Kryx didn't have the numbers to support that GWM paladins 'break' anything, nor does the figther GWM seem to - compared to fx SB.
Because I don't take into account bless. Not because it isn't broken. Though I'm inclined to agree that normal Paladin is fine - it'd just be OoV Paladin where the problem really presents itself.


So the real culprit and cause for all this concern is the PAM+GWM recklessly attacking barb?
The real culprit is -5/+10 with advantage. That's true for all Barb and OoV variants.



Overall I think this thread has been very enlightening. It has narrowed the problem to a class and a specific archetype. That's good.

Though this whole time we we have forgotten Ranged. Lets swap over to that to see if we can get some insight:
http://i.imgur.com/bY9X53g.png
http://i.imgur.com/n9vQRNs.png
Summary: Hand Crossbow is really good. However even a Fighter Longbow/Heavy Crossbow are still doing good damage compared to their secondary damage role class counterparts.
For example a Rogue TWF wins 1-5, but 6+ his damage is ~75-85% of a Heavy Crossbow Fighter
Or a Monk wins 1-5, but 6+ his damage is ~75-100% of a Heavy Crossbow.
Without -5/+10 ranged does suffer some - I acknowledge that. I don't have a good solution for ranged.

Honestly I'd be fine giving GWM something like a flat damage increase. I just find -5/+10 inherently leads to problems in the case of advantage and bonuses to hit.

[smcmike]

The math is impressive, in general.

How did you pick the AC you are testing against? Obviously fiddling with this assumption changes the math pretty drastically, right?

I'm partially convinced, but come down on the "it's fine" side. I just don't see it as a very big problem that certain builds do significantly more damage than others under certain conditions.

I found it interesting that some people seem to think that the existence of this build hurts their ability to play a TWF ranger, even in a game where no one is using the -5/+10 feats, but that this would be fixed if the declaration was made that no one CAN use those feats. It's like optimization is so central that the only way you can allow yourself to play a sub-optimal build is by banning all better builds.

[Finieous]

The heavy weapon builds aren't optimized at all. They simply have 1 or 2 feats (GWM, Polearm, or Polearm+GWM).


They have max Strength, they have the heavy weapon feats, and they look for advantage. That's built in for barbarian; the fighter takes battlemaster and chooses the best maneuver for DPR most of the time. I'm not sure what else you'd do to optimize them.



The Barbarian doesn't have any archetype as well as none of them result in direct DPR.
If we gave the Barbarian an archetype then Battlerager would blow swashbuckler out of the water with his bonus action attack (no TWF battleragers? really WotC? Uggggh).


I agree that none of the archetypes offer direct DPR, but the swashbuckler does for the TWF rogue. I don't know anything about the battlerager.



That doesn't work. Sneak attack still must be made with a light weapons. No polearm is a light weapon.


I don't understand. The rogue doesn't need to use a polearm to benefit from Sentinel. The Sentinel rogue uses his reaction for Uncanny Dodge if he's attacked, and he uses it for an extra sneak attack if an adjacent enemy attacks someone else.



Those numbers are incredibly wrong. No idea where he got those. S&B fighter is doing the following DPR:
19 DPR @5
36 DPR @11
41 DPR @17
52 DPR @20

I accidentally pulled your houserule numbers for S&B! It looks like you're giving the S&B Fighter your mini Power Attack feat at 12. I actually like that feat, BTW, I just wouldn't replace GWM with it.

[Kryx]

How did you pick the AC you are testing against? Obviously fiddling with this assumption changes the math pretty drastically, right?
DMG 274.

I think there is a lot right with the default balance. There are a few cases that I think need fixing: TWF, Melee Ranger, -5/+10 with advantage, Polearm doing more damage than GWM even though it has reach.
I'll have to do some review and see if I can come up with something better.
Goals would be:
Polearm does ~90% of GWM.
TWF does ~90-95% of GWM.
S&B does ~75% of GWM.
Rogue can do about 80-90% of GWM (as is default)
Monk should be mostly even with S&B (as is default)
Ranged does about ~80% of GWM




They have max Strength, they have the heavy weapon feats, and they look for advantage. That's built in for barbarian; the fighter takes battlemaster and chooses the best maneuver for DPR most of the time. I'm not sure what else you'd do to optimize them.
I agree that none of the archetypes offer direct DPR, but the swashbuckler does for the TWF rogue. I don't know anything about the battlerager.

I don't understand. The rogue doesn't need to use a polearm to benefit from Sentinel. The Sentinel rogue uses his reaction for Uncanny Dodge if he's attacked, and he uses it for an extra sneak attack if an adjacent enemy attacks someone else.
I was talking about the Barbarian. He has max strength and smashes - nothing complex or optimized there. The Fighter does optimize w/ trip though.

Point being: The thread has narrowed in scope. Fighter w/ -5/+10 isn't of concern, even with trip. It's really the Barbarian. Barbarian isn't optimized, nor is the rogue. Rogue could optimize and take swashbuckler and a specific sentinel build as you said (sorry, mis-read before), but sentinel is very campaign dependent and would definitely put the rogue at risk in melee. Compare with "Sneak Opportunity Attack" which has a 7% chance of happening. You'd have to have sentinel happen fairly often. To counteract the Swashbuckler the Barb can take Battlerager which gives a bonus attack - that is worth more than swashbuckler.

Either way the goal is to balance around the norm, not optimization. Some ignore Barbarian and OoV as they consider them "outliers", but they are pretty core imo.



I accidentally pulled your houserule numbers for S&B! It looks like you're giving the S&B Fighter your mini Power Attack feat at 12. I actually like that feat, BTW, I just wouldn't replace GWM with it.
huh? None of my S&B guys do anywhere near 60+ DPR. Not sure what you're referring to.

[KorvinStarmast]

The math is impressive, in general.

How did you pick the AC you are testing against? Obviously fiddling with this assumption changes the math pretty drastically, right?

I'm partially convinced, but come down on the "it's fine" side. I just don't see it as a very big problem that certain builds do significantly more damage than others under certain conditions.

I found it interesting that some people seem to think that the existence of this build hurts their ability to play a TWF ranger, even in a game where no one is using the -5/+10 feats, but that this would be fixed if the declaration was made that no one CAN use those feats. It's like optimization is so central that the only way you can allow yourself to play a sub-optimal build is by banning all better builds.Maybe it's the internet that brings the inner munchkin out in people.

[Finieous]

Point being: The thread has narrowed in scope. Fighter w/ -5/+10 isn't of concern, even with trip. It's really the Barbarian.


Fair enough. Would you further say that it's really the Barbarian with both GWM and PM? Or do you still feel that the Barbarian's DPR with either one of these feats is too high?



Barbarian isn't optimized, nor is the rogue. Rogue could optimize and take swashbuckler and a specific sentinel build as you said (sorry, mis-read before), but sentinel is very campaign dependent and would definitely put the rogue at risk in melee. Compare with "Sneak Opportunity Attack" which has a 7% chance of happening. You'd have to have sentinel happen fairly often. To counteract the Swashbuckler the Barb can take Battlerager which gives a bonus attack - that is worth more than swashbuckler.


How does the battlerager's bonus action attack work? I really haven't paid any attention to the battlerager leaks. Is it higher DPR than PM?



huh? None of my S&B guys do anywhere near 60+ DPR. Not sure what you're referring to.

I downloaded your spreadsheet and I'm looking at the S&B tab. I'm pretty sure I didn't do anything to monkey with it. Maybe I have an old version?

http://i315.photobucket.com/albums/ll442/gbenage/SB%20DPR.jpg

[Kryx]

Fair enough. Would you further say that it's really the Barbarian with both GWM and PM? Or do you still feel that the Barbarian's DPR with either one of these feats is too high?
It's not just Barbarian and Polearm+GWM. It's any of the main damage builds (GWM, Polearm, Polearm+GWM) stacked with reliable advantage (any Barbarian, OoV Paladin)


How does the battlerager's bonus action attack work? I really haven't paid any attention to the battlerager leaks. Is it higher DPR than PM?
They have spiked armor and hit you for 1d4+str. Same as PAM, really.


I downloaded your spreadsheet and I'm looking at the S&B tab. I'm pretty sure I didn't do anything to monkey with it. Maybe I have an old version?
Looks like you have the version from a week or so ago when I was experimenting with giving "power attack" to all builds that wanted it. The end result was just a bloat of the math. :(

And my players voices their opinion that the power attack feature isn't fun. I'm inclined to agree.

[Finieous]

Looks like you have the version from a week or so ago when I was experimenting with giving "power attack" to all builds that wanted it. The end result was just a bloat of the math. :(


Heh. It has me thinking I might change "Great Weapon Master" to "Weapon Master" and including a -3/+6 option for one-handed weapons in the feat. I'm still happy with heavy-weapon builds doing more damage, but it seems like it creates options for the various one-handed builds to close the gap somewhat. All else being equal, adding options is preferable to removing them for me. Thanks!

ETA: -3/+6 may be too much, but I can play with the numbers.

[Fighting_Ferret]

I have an idea... would have to run numbers, but how about GWM and Sharpshooter only apply on the first successful attack on any round. The problem with them is that the +10 is cumulative on multiple strikes, far out damaging any normal weapons damage output, so... make it like a rogue's sneak attack, usable only once per round. They can still go for extra damage, and still get their better to hit to help minimize overall DPR.

Another option would be to limit the +10 only on main attacks made with the attack action. Bonus attacks and AoO wouldn't allow the extra damage to apply. This would most likely nerf Polearm down to just getting an extra 1-9 damage via their bonus action, comparable to dual wielding.

As for ranged, calling someone a sharpshooter means they take time to aim... I'd let them have the option of 1 shot per 2 attacks they could normally make with the bonus damage from sharpshooter and remove the -5 to hit, which is only a -3 to someone with Archery style.
Hand crossbows should be reduced in damage to at least a d6, if not a d4, with a reduced range.

Maybe I'm crazy... but anyway, there's my 2 cents worth... for what it's worth...

[Tanarii]

We can discuss calculations - as long as it's constructive. I do want to improve the sheet so please do point out any errors.It wouldn't be. But it's not particularly Germaine to this discussion. GWM, Polearm Master, Advantage, etc all have an calculable effect on DPR on a round by round basis as well as an adventuring day, and regardless of assumption of base hit chance, and the distortion appears to be significantly noticeable no matter which way you calculate it. But we're mostly comparing apples to apples regardless ... it's the distortion within builds of a given class that really shows the problem.

I'm starting to wonder if the 'real' culprit isn't just Advantage, or the lack of bonuses to damage. Classes that can get advantage reliably (Barbarian & Fighter) seem to show far more significant distortion. Meanwhile classes that add bonus damage (Paladin, Warlock, Ranger, Cleric) are affected less.

The big example is Warlock with +Cha damage to Pact weapon attacks. GWM becomes close to a worthless feat when you do 3d6+10 damage with your greatsword and two maxed stats, even with advantage.

[Mara]

Assuming that everything Kryx says is true, that all calculations are perfect, I'm not seeing the issue. It's a feat. It increases damage in most situations and not even to an amount that is all that game breaking. What it doesn't do is add a +1 to an ability check or save. It doesn't add utility. You advance your combat prowess at the opportunity cost of social or exploration ability.

The other styles increase AC and survivability if you take a combat specific feat for the style. If you don't, you gain more utility instead.

[Fighting_Ferret]

This assumes no magic items...
Lowest to hit bonus: +2 average is probably a +4 highest is +7
Highest to hit bonus: +11

Highest AC at level 1: 19 average is 14/15
Highest AC beyond level 1: 21 average is 17/18

Now the benefits of feats:
Shield Master: Pros:You can knock an opponent back/prone as a bonus action (most DMs allow you to do this before you attack, netting you advantage on all follow up attacks)
You can add the shields AC bonus to any dex save that target you (and only you)
If an effect would let you save vs half on a successful save, you take no damage, if you make the dex save. (See the above for conditions for bonus shield AC to save)
Cons:It only work on opponents that are large or smaller and they still get an opposed athletics/strength check. You still have to make the dex saves for he other 2 features.
Polearm Master: Pros:You get an extra attack with the butt of your polearm of choice at 1d4+ ability mod, reach with no penalties, and a free AoO when something enters/leaves your reach.
Cons:You lose out on the AC a shield provides.
Great Weapon Master: Pros:You get a free attack to an adjacent enemy if you land the killing blow on a enemy. You have the option of adding 10 damage to every attack made with a heavy weapon.
Cons:That +10 damage comes with a -5 to hit penalty and you lose out on a +2 to AC that a shield provides.
Sharpshooter: You get to use the full range of your weapon without disadvantage, ignore 1/2 and 3/4 cover, and add 10 to any attack you make with your ranged weapon of choice.
Cons:Not every location allows you to take advantage of your superior range, that +10 damage comes with a -5 to hit penalty, you need someone up front to take hits.
Crossbow Master: Pros:You can fire a bolt for every attack you get. You don't have disadvantage against targets within melee range of you, and you can get a bonus attack with a hand crossbow.
Cons:That hand crossbow has to be loaded in order to get the bonus attack (this is a matter of contention amongst DMs), and you lose out on the AC a shield provides.
Dual wielding: Pros:You can use a larger weapon die, you get a +1 to AC, you can draw 2 weapons instead of 1 as a free action as part of a move.
Cons:You lose half the AC a shield provides.
Sentinel: Pros:You get a free AoO if a creature within your reach if it attacks a target other than you, if your AoO hits that creature can't move that turn.
You need at least two creatures within range and an ally that hey can attack pretty close as well.

None of the feats are problematic on their own... it's the fact the polearm master plus great weapon master nets you up to 2 extra attacks (the extra butt attack from polearm and the AoO from polearm, but now both can have the +10 damage added to them. Sure your to hit bonus takes a hit, but you are getting 2x the hits with good regularly. Two times the chances to hit means better damage overall... especially when that damage is fixed damage (ie weapon die plus ability mod+ 10) Crossbow Expert experiences some of this problem with Sharpshooter and the character is wielding a hand crossbow, but at 2/3 attacks with regularity. The combat styles just add synergy to all the respective feats. It is the use of those synergies that Kryx is talking about. In a normal game, where players use characters... you would probably be fine... but there are a lot of people that use builds, especially designed to do a certain thing (mainly combat) extremely well. 5E was designed with bounded accuracy, so as to reduce the need to minimize the need for min/maxing, as all characters are viable as is (yes some are better at things that others, but overall they can all contribute, especially to combat).

[Nu]

An awfully large amount of people dismiss Kryx's argument because of incomparable, hypothetical factors, like using a GWM/Sharpshooter build might lower your utility or survivability.

I would argue that this is not a very good way to counter Kryx's math, because he's showing that a GWF can do 40~50% more damage than a rogue using sneak attack. There's going to be disparity in any system, but I agree that that much is too much.

Feats in 5E are great, but I think they cover much too wide a spectrum of power. On one end you have things like Great Weapon Master, Polearm Master, and Sharpshooter, which have the above effect on the game. Then you have interesting and nice feats that add a lot and are generally useful, like Inspiring Leader, Lucky, and Healer. Then you have things that are just so situationally useful that I'd feel awful if I ever gave up an ASI or other feat for them, like Keen Mind, Linguist (you can learn new languages during downtime!), and Observant.

My personal preference would be to see feats brought to the level of the middle group that includes Inspiring Leader, Lucky, and Healer. I feel as though these feats add a lot to the game without causing any particular build or choice to overshadow others. GWM/PM/Sharpshooter simply give too large of a benefit compared to many other choices you could make, and they are not as situational as many make them out to be. DnD's rules focus on combat, and if your game isn't, you're probably doing more freeform roleplaying than actually playing DnD. Which is fine, but it's irrelevant to the rules. The rules interact with non-combat situations in minimal ways that mostly involve a pass-fail d20 roll. Again, a good group can work with and around this, but that also has nothing to do with the rules and doesn't enter into the debate.

Likewise, if I consider all character concepts equal in my mind, then it does honestly bother me if some are mechanically inferior than others. Parallel imbalances don't bother me that much, but I do agree with Kryx here--the math shows that there is simply too big a difference in the damage-dealing capability of a character who uses one of the -5/+10 feats and one who does not. That's why there's math, Kryx isn't saying one character does more damage than another, it's rather that the difference is too great.

[KorvinStarmast]

An awfully large amount of people dismiss Kryx's argument because of incomparable, hypothetical factors, like using a GWM/Sharpshooter build might lower your utility or survivability.

I would argue that this is not a very good way to counter Kryx's math, because he's showing that a GWF can do 40~50% more damage than a rogue using sneak attack.
So what?

I do not understand, again, why it is that the rogue is being compared 1 for 1 to a specialist in melee fighting.

If all you are doing is fighting in your campaign, and if all you are doing is measuring your DPS against each other, then your problem isn't a feat.

[Nu]

So what?

I do not understand, again, why it is that the rogue is being compared 1 for 1 to a specialist in melee fighting.

If all you are doing is fighting in your campaign, and if all you are doing is measuring your DPS against each other, then your problem isn't a feat.

That's rather simplifying the issue, isn't it? My point would be that there aren't Feats that improve the rogue's ability to accomplish other tasks covered by the rules (and anything that is not covered by the rules is beyond the scope of the debate and irrelevant to it), not in the same way a feat like GWM/Sharpshooter helps a two-weapon wielder do damage. The feats are simply too good compared to the alternative. If the difference was smaller, there would be less issue.

[Saggo]

So what?

I do not understand, again, why it is that the rogue is being compared 1 for 1 to a specialist in melee fighting.

If all you are doing is fighting in your campaign, and if all you are doing is measuring your DPS against each other, then your problem isn't a feat.
The premise was pretty well outlined. Without feats there is a range of damage that was argued as decently balance. With feats, certain classes balloon well out of that range while other classes get little to no benefit.

If you don't think game balance is a worthwhile goal, then of course you won't ever care.

[Corran]

So what?

I do not understand, again, why it is that the rogue is being compared 1 for 1 to a specialist in melee fighting.

If all you are doing is fighting in your campaign, and if all you are doing is measuring your DPS against each other, then your problem isn't a feat.
A better balanced system leads to more fun and a more enjoyable game overall imo. Here is a very general example. Say you want to play an archer. Creating the character, you think of his personality and all the character traits that will make him unique and enjoyable for you to play that character. Obviously, one thing you definitely had in mind for that character, would be for him to be a real threat when using his bow (by which I mean that you most likely thought of that character being a damn exceptional archer). Afterall, a great portion of the game revolves around combat. So you begin in your new campaign playing that character. Everything is fine, until you enter an encounter. Then, you realise, that the game system favours heavily the melee characters over the ranged ones, making it almost impossible for your archer to be considered even close to as deadly as every other melee character is. Wouldnt that specific imbalance in the game system take something out of your character concept (ie his deadliness)? Wouldnt sth like that reduce the amount of fun you will have when playing this characrer? Afterall, besides all the nice little behaviour traits and rp elements you thought for this character to have, you clearly wanted him to be good at what he does (shooting things dead). My point is, balance helps everyone shine and have his moment in the spotlight, and prevents characters from dominating it, so everyone gets to be the hero of the day every now and then, so that equals to more fun. Simple as that.

[Lollerabe]

Kryx: Aren't you putting too much emphasis on the oov paladin?

Once per short rest as a bonus action he gains advantage on one target for 1min. Do you find that reliable enough to qualify as broken with GWM?

If no, then my previous statement was true, the 'problem' is the reckless attacking barb. In which case we are back to square one - it's a class based solely on melee single target damage, with next to no value outside of that.

With that said your chart also assumes that he uses reckless, while this makes sense math wise your chart can't calculate how often it's feasible to do so.

That BBEG doesn't need to have a lot of mooks close to him, before reckless will get you killed faster than you kill (at least in our games, rages and reckless attacking can be seen fairly easily and is often kited/countered).

I'm not trying to argue how you should rule at your table obviously - that's on you and your table.

I'm just trying to say that this scenario where the barb deals 90-100% more damg than the rogue doesn't happen as often as your chart can make it seem like.

So as Malifice said - when the stars align this one character built entirely around dropping high hp / low ac mobs, does this at a speed completely uncontested.

Is that really a problem?

Nobody complains when a wizard drops the perfect fireball and wins the encounter singlehandedly - we expect the wizard to do this from time to time, and everyone in the party feels awesome when he does it.

Anyway I'm repeating myself. I just can't shake the feeling that you are being a bit stubborn about this. Is it possible that you are a bit negatively biased towards gwm?

[Shaofoo]

I'm well aware of what the DM does. We can reasonably assume the DM, as I said, often resolves scenarios with dice rolls.

The point isn't that DMs use the dice results, the point is that DMs always override dice results if he wants to. If he wants something to happen then it will, he isn't shackled to the dice.


Linguist and Actor feats are mechanically defined by their ability score increases, their advantage rolls, and their skill checks. These are demonstrably measurable with numbers. You can correlate their successes against advantage and skill throws. I can't find a single feat that doesn't include a measurable mechanical improvement.

Can you give me the mathematical advantage of having three more languages learned or being able to mimic voices? Even the cipher and advantage over disguises is useless if you don't actually use them.

For some tables I assume having a language learned is a big thing since it opens up so much while in other tables it is next to useless because everyone talks the same language (the most you get is a different accent). If you are playing a door kicker campaign then I doubt you'll be wanting to disguise yourself.


Not that that matters. I never said "Math is the end all be all." I said it was a medium. Objective and necessary.

It can be used to gauge certain aspects but it should be tempered. You can't throw math and expect it to stick on the game as a whole, especially when you have a guy that can change the rules.




Fudging rolls is a controversial topic. Even then.. using it to solve a math inequality is pretty ridiculous...

This isn't about fixing problems this is getting what the DM wants and what the DM wants the DM gets. Personally since the DM angle is too broad for anyone to ever truly measure in any meaningful way basically your analysis is based on the DPR side of the game, I don't even say battle because there are ways that you can influence a battle without DPR at all and not all battles have the objective of beating up the other side to defeat. Sure as the game is presented battle and damage are big cornerstones to the game but the DM can easily change that around and throw full on skill challenges with traps instead of battles.


A better balanced system leads to more fun and a more enjoyable game overall imo.

Is it relevant if I mention 4e?

[JNAProductions]

Is it relevant if I mention 4e?

*Raises hand* I liked 4E!

That being said, I like 5E more, so.

[Safety Sword]

Is it relevant if I mention 4e?

A perfectly balanced turd is still a turd?

If -5/+10 mechanics are ruining your life, then you have two options. Fix it, or counter it.

When I DM, I counter it. Higher AC. Make it a real choice whether to use the feat or not.

In isolation this feat may increase damage by a large amount. In a real game where there are multiple other factors I have never noticed it break anything.

[Nu]

Anyway I'm repeating myself. I just can't shake the feeling that you are being a bit stubborn about this. Is it possible that you are a bit negatively biased towards gwm?

You would seem to be overstating the scenario required for a fighter/barbarian/etc with GWM to be effective. A group of tightly packed enemies will likely have a few dropped by the repeated attacks + bonus attack made by a buffed combatant with GWM if they're weak enough to fall to a fireball.

"Counters" and "kiting" would imply the party is using weak tactics and not working together efficiently. The battlefield controller should not allow enemies to simply kite their allies around, the support characters should not allow their party members to be locked down or countered for any significant period of time. But if they do, then it speaks to a different problem, one that's largely unrelated to the effectiveness of GWM. Anything can be ineffective if the party does not try to use it effectively. You speak of problems that plague every melee character, and there are generally better ways to address it than your valuable feat slots.


Is it relevant if I mention 4e?

Well, it'd be relevant only if you observed it from a neutral perspective. Personally, I do enjoy 4th edition. And there are many aspects of 4th edition that went into 5E's ruleset, and more it could learn from.

[Xetheral]

Goals would be:
Polearm does ~90% of GWM.
TWF does ~90-95% of GWM.
S&B does ~75% of GWM.
Rogue can do about 80-90% of GWM (as is default)
Monk should be mostly even with S&B (as is default)
Ranged does about ~80% of GWM

These seem to low to me. Can a player reliably notice a 10% difference over the course of a 4-hour play session? (I did a quick look for any research on that subject, but didn't find any. If anyone knows of any studies, I'd be curious.) Even if it is noticeable, the random factor introduced by the dice will sometimes overwhelm a 10% difference. To get an idea for how often, let's consider an example, using similar assumptions as your sheet:

Here's the probability distribution of total Attack Action damage over a 5-round combat for a Level 10 Barbarian with and without GWM, versus AC 17, using Reckless Attack: http://anydice.com/program/6e5e. The Barbarian without GWM does more total damage in 30.24% of combats, despite doing less damage on average. (Value calculated here (http://anydice.com/program/6e60) by subtracting one curve from the other, as the probability at which the "At Least" total becomes positive.) If we assume a 10% difference is the threshold for noticeability, the two damage totals will be indistinguishable in 20% of combats (using +/- 16 damage as the cutoff).

Here's the same comparison over a 25-round combat day: http://anydice.com/program/6e5f. The Barbarian without GWM does more total damage on 11.84% of days, despite doing less damage on average. (Value calculated here (http://anydice.com/program/6e61).) Again, if we assume a 10% difference is the threshold for noticeability, the two damage totals will be indistinguishable on 30% of days (using +/- 80 damage as the cutoff).

Note that those calculations show a 16% average difference in damage, larger than your spreadsheet's 10% difference. (This is at least partially because I can't factor in a fractional chance of the PM reaction attack in AnyDice without going over the 5 second calculation limit.) With your more comprehensive number, the damage totals will be closer together, and the frequency with which the two numbers are indistinguishable will shoot up. Thus, my numbers are underestimates.

So, as-is, if players can notice a 10% difference, then a player of a Level 10 Barbarian wouldn't notice a difference between taking +2 STR or GWM on at least a third of adventuring days. And those are two builds that are relatively close in damage. If you further compress the damage ranges as you suggest, the differences will become even harder to notice, and I don't think that's a good thing.

(This is my first time using AnyDice near its complexity limit, so I may have made a mistake. If so, please let me know!)

[Shaofoo]

*Raises hand* I liked 4E!

That being said, I like 5E more, so.




Well, it'd be relevant only if you observed it from a neutral perspective. Personally, I do enjoy 4th edition. And there are many aspects of 4th edition that went into 5E's ruleset, and more it could learn from.

I love 4th edition and I also love 5th edition. I am not saying one is better than the other.


A perfectly balanced turd is still a turd?

But it is balanced.

I think that 4e is a good cautionary tale for those who look for balance without looking at the bigger picture. Love or hate it 4e did solve a lot of "problems" that D&D used to have (and I use problems in quotations because those might not have been problems in the games as a whole but they were complained about, sometimes to length) of course a lot would say that 4e isn't true D&D.

I just think that at times we should take a step back.

[Saggo]

The point isn't that DMs use the dice results, the point is that DMs always override dice results if he wants to. If he wants something to happen then it will, he isn't shackled to the dice.



Can you give me the mathematical advantage of having three more languages learned or being able to mimic voices? Even the cipher and advantage over disguises is useless if you don't actually use them.

For some tables I assume having a language learned is a big thing since it opens up so much while in other tables it is next to useless because everyone talks the same language (the most you get is a different accent). If you are playing a door kicker campaign then I doubt you'll be wanting to disguise yourself.



It can be used to gauge certain aspects but it should be tempered. You can't throw math and expect it to stick on the game as a whole, especially when you have a guy that can change the rules.

What a DM decides to do is just as anecdotal as how much fun a certain class is, so I'll to operate under the assumption that scenarios are resolved with dice rolls, as the system was designed to do. You can do whatever you want as a DM and you can let your DM do whatever you want, but what we all have is the rulebook.

Incidentally, if you don't use something, of course it's useless to have.

"Tempered" math is a weird way to put things. It's all just applied statistics.

[Kryx]

I haven't had as much time in the later day today to respond as I ran a session tonight, though I think Saggo and Corran have argued my points for me well.




Kryx: Aren't you putting too much emphasis on the oov paladin?
Once per short rest as a bonus action he gains advantage on one target for 1min. Do you find that reliable enough to qualify as broken with GWM?
It is usable 3 times per day. Using the calculated rounds per encounter that's ~15 rounds out of the total 25.7. So ~58% of the time. That's the number I use. To answer your question: That's pretty often. You can see the results on the charts. It is the best DPR option for several builds.


With that said your chart also assumes that he uses reckless, while this makes sense math wise your chart can't calculate how often it's feasible to do so.
There are situations where a Barb wouldn't recklessly attack? With decent AC and resistance to damage in my experience it's pretty much always in the Barbarian's favor to do so. Doubly so with -5/+10.


rages and reckless attacking can be seen fairly easily and is often kited/countered
Uhh, there is a balance here. If you're specifically trying to disable the Barbarian then you'd also be specifically trying to disable the caster. Every class is a target for disabling.
If you argument is "Barbs do a lot of damage so the DM should limit them" then that's pretty lame for anyone who plays a useful class imo.


I'm just trying to say that this scenario where the barb deals 90-100% more damg than the rogue doesn't happen as often as your chart can make it seem like.
In my games it would surely happen. Though it sounds like you handle certain classes differently.


I just can't shake the feeling that you are being a bit stubborn about this. Is it possible that you are a bit negatively biased towards gwm?
I've flip flopped several times on -5/+10 and -3/+6. Just because I now disagree based on what the math says doesn't mean I'm being stubborn. It means I have outlined a problem which you, and others, think isn't a problem. Simple disagreement of where balance lies and that's fine.
If my math is wrong then I'd switch and say it's not a problem (though based on my player's preferences likely still replace it with more reliable damage of the same level as the normal case)




When I DM, I counter it. Higher AC. Make it a real choice whether to use the feat or not.
That punishes every build option and just makes encounters harder overall.. Instead of the default being 65% chance to hit you're looking at 50-60%. Overall I find succeeding (hitting) more often than not to be more fun. If you prefer closer to 50/50 though, then go for it.




These seem to low to me.
Please do suggest if you think the percentages I outlined should be different. It was a rough estimate. I refined it a bit more in my Balance sheet which I posted below.


Can a player reliably notice a 10% difference
This comes back to the point of balance. Some people prefer more balance. Some people don't. If I can do it well then I don't see a reason to not try to fix the math errors.


The Barbarian without GWM does more total damage in 30.24% of combats
I don't have time tonight to check all the number, maybe some else can or I can look closer tomorrow.
By my numbers a Barbarian at 10:
GWM: Picks up GWM at 4, stat at 8. 36 DPR
Polearm: Picks up Polearm at 4, stat at 8. 35 DPR (fixed an issue where I didn't have a stat increase at 8)
Polearm+GWM: Picks up GWM at 4, Polearm at 8. 39 DPR

I can look close tomorrow. Though I don't think assuming Vuman is a good assumption for balance discussions. Nor does it look like you have added GWM's cleave nor PAM's provoke to the damage.






I think there is a lot right with the default balance. There are a few cases that I think need fixing: TWF, Melee Ranger, -5/+10 with advantage, Polearm doing more damage than GWM even though it has reach.
I'll have to do some review and see if I can come up with something better.
Goals would be:
Polearm does ~90% of GWM.
TWF does ~90-95% of GWM.
S&B does ~75% of GWM.
Rogue can do about 80-90% of GWM (as is default)
Monk should be mostly even with S&B (as is default)
Ranged does about ~80% of GWM
Earlier today I worked up a sheet to compare the balance of both RAW and my houserules in comparison to these guidelines (https://docs.google.com/spreadsheets/d/1d-9xDdath8kX_v7Rpts9JFIJwIG3X0-dDUtfax14NT0/edit#gid=2141556455).
Overall RAW would do ok if Fighter and Barb weren't so far out (except TWF, hand crossbow, and Barb S&B). Earlier I had averaged the 4 GWM classes together and that was much better looking at the problems were hidden in the averages. Which tells me for the most part it's a matter of providing an alternative that doesn't have a problem when combined w/ advantage.

On my houserules: I did earlier take out Polearm Bonus attack. There is definitely some things that I'll need to refine. Ideally it looks like I'll want something of the same power as -5/+10 in the normal cases.

Thought I'd share my unfinished results. My next step would be to redo the Dighter's DPR so it doesn't spend all of it's superiority dice on tripping (mentioned earlier), but maybe about half while the rest would be on riposte or other dice. That should help make the Fighter more realistic to a normal use case. After that I can see what kind of buff GWM would need without -5/+10 that would keep it in a normal range (for the advantage case as well).

[Shaofoo]

What a DM decides to do is just as anecdotal as how much fun a certain class is, so I'll to operate under the assumption that scenarios are resolved with dice rolls, as the system was designed to do. You can do whatever you want as a DM and you can let your DM do whatever you want, but what we all have is the rulebook.

Incidentally, if you don't use something, of course it's useless to have.

"Tempered" math is a weird way to put things. It's all just applied statistics.

This isn't just me and my DM, this assumption applies to everyone because I assume that everyone plays with a DM. We all have a DM as well as the rule book, I have yet to see a D&D game operate without one. Of course unlike the rulebook DMs can vary wildly and that is my point. You can't just handwave the DM because it is as necessary as the rulebook itself (unless you know of a way for DM-less playing).

And I didn't mean to temper math, I meant to temper the expectations that math will resolve everything. A lot of people seem to throw out words that the game is imbalanced based on the DPR numbers but I want to say that DPR is but a part of the game and the game itself is the sum of its parts. You can't balance a game around just making the DPR numbers balanced.

And because the DM varies you can't be sure just how big a part is DPR in the first place. It can be the cornerstone of the game to it can be inconsequential in the grand aspect of things, either spectrum and everything in between can occur.

[Icewraith]

In terms of other combat-enhancing feats that don't just help out the melee bruisers, there's always Alert and Lucky to consider. Regularly getting to apply your damage first and not being surprised are huge combat buffs. Both are even better for Rogues, who need to dual wield or grab a feat to get two reliable attacks per round in order to land that Sneak Attack, and then try to set up those reaction attacks.

Other classes may not have options that boost DPR as dramatically, but they do have feats that significantly enhance their capabilities.

For instance, the Wizard takes Resilient:Con and suddenly can hang onto a Haste spell even when someone's plinking him with arrows. Ditto for Clerics with Bless. These characters don't care if the Barbarian is doing craploads more damage than them because part of it is THEIR damage.

This thing really only seems like it would really be a problem if there are six players at the table and they're all playing DPR classes and they're all trying to out damage each other every single round, yet only one of them actually optimizes their character.

Instead what happens is someone goes GWM+Polearm Barb or Fighter, someone else goes Shield Master maybe Paladin or Valor Bard, someone else goes spell sniper Warlock and Hexes STR or DEX for the Shield Master, someone else goes Fighter or Ranger and plinks with crossbow/sniper, someone else plays a Lucky Alert Halfling Rogue Assassin and starts off nearly every first round of combat with a bang and solves most of the party's skill problems, and someone... well let's be honest, in this sort of party someone plays a Resilient:Con cleric and Blesses the rest of the party, then spams cantrips or low-damage attacks while waiting for a prepared-spell shaped hole to fit one of their prepared spells into.

The GWM+Polearm player does insane damage but also misses a lot (sometimes he doesn't even do any damage in a round!), the Shield Master doesn't mind being behind on damage because he's handing out the advantage (and a Paladin smites on his crits for tremendous amounts of damage) and the next time a dragon shows up he won't be as worried about the breath weapon, the Warlock is Hexing enemies for the Shield Master on top of his own consistent damage, the Archer has even more reliable damage and impressive burst, the Rogue's damage spikes are impressive (especially if the rest of the party sets the Rogue up for off-turn attacks), and the Cleric knows they are keeping everyone else buffed and alive, and that Bless is accounting for a good chunk of GWM's damage.

The DM notes the composition of the party and takes a bit more care in his encounter planning.

[Pex]

I neglected to mention my paladin, who uses a shield, never a two-handed weapon, and will never take great weapon master feat, is an Oath of Vengeance paladin. Either I'm playing the game wrong or not everyone will take this feat just because it looks so good on paper. Bias as I am, I really don't think it's the former.

[Saggo]

This isn't just me and my DM, this assumption applies to everyone because I assume that everyone plays with a DM. We all have a DM as well as the rule book, I have yet to see a D&D game operate without one. Of course unlike the rulebook DMs can vary wildly and that is my point. You can't just handwave the DM because it is as necessary as the rulebook itself (unless you know of a way for DM-less playing).

And I didn't mean to temper math, I meant to temper the expectations that math will resolve everything. A lot of people seem to throw out words that the game is imbalanced based on the DPR numbers but I want to say that DPR is but a part of the game and the game itself is the sum of its parts. You can't balance a game around just making the DPR numbers balanced.

And because the DM varies you can't be sure just how big a part is DPR in the first place. It can be the cornerstone of the game to it can be inconsequential in the grand aspect of things, either spectrum and everything in between can occur.
What handwave? The recommended/expected/default way a DM arbitrates a game is by adjusting stats like DC and AC, thus I can assume Proficiency and Advantage matter, and I can use scary math to find the statistical chance of success. The chance of success vs a DC doesn't go away or change just because any given DM ignores the ruleset in favor of drama, it just means it wasn't applied. None of that presupposes DM-less playing, so you built a scenario that was never in question.

You're right, a game is a sum of its parts, that doesn't automatically preclude balancing each part. The premise hasn't strayed much from the start, DPR ratios without feats compared to DPR ratios with feats.

"Math will resolve everything" is not a thing I said. Or implied. To repeat, math is a medium, a way of discussing this game, that we all have access to and the game is predicated on. A "baseline" if you will. To go full circle, what I will always argue is simple, Math Matters.

[Mara]

The premise was pretty well outlined. Without feats there is a range of damage that was argued as decently balance. With feats, certain classes balloon well out of that range while other classes get little to no benefit.

If you don't think game balance is a worthwhile goal, then of course you won't ever care.
Your fighter grabs GWM and Polearm master my Rogue grabs Mobile and Skulker. Both have both feats and have maxed their main stat by level 12.

You can argue that more DPR is better, but I wouldn't discount greater hit-and-run tactics on top of the Rogues already staggering utility by this point.

[smcmike]

I'd like to see more discussion of xetheral's underlying point - kryx sort of bats it away by questioning the math, but if you make sure the math is right, I think this is a useful angle to take - how often will one build outshine another in a combat, or a combat day, rather than (or in addition to) the average DPR for each build. I don't know what the numbers will be, or what they should be, but I feel like it might be a bit more relatable to played experience. I have a very hard time saying what the appropriate gap in DPR should be.

[Kane0]

I like the disadvantage / +10 idea. My group uses the cancel adv/dis at 1:1 houserule, will definitely try it.

[Saggo]

Your fighter grabs GWM and Polearm master my Rogue grabs Mobile and Skulker. Both have both feats and have maxed their main stat by level 12.

You can argue that more DPR is better, but I wouldn't discount greater hit-and-run tactics on top of the Rogues already staggering utility by this point.

I mean, who doesn't like more DPR?

Really, the question for that would be how much of a difference in DPR should Mobile and Skulker be worth. I don't think anyone is seriously arguing Rogue should match.

[steppedonad4]

I like the disadvantage / +10 idea. My group uses the cancel adv/dis at 1:1 houserule, will definitely try it.

So if you have more than one source of advantage/disadvantage you can end up with advantage or disadvantage even though you have a source of the counterpoint?

[Mara]

I mean, who doesn't like more DPR?

Really, the question for that would be how much of a difference in DPR should Mobile and Skulker be worth. I don't think anyone is seriously arguing Rogue should match.

There are few situations after this point that the rogue would ever be hit with an attack. The rogue could solo most encounters at zero resource expenditure (aside from a few arrows possibly). In a team situation the feats make scouting very doable, granting the party a surprise attack. If scouting goes south (somehow a minimum of 23 stealth failed to go undetected) then you can still run away through rough terrain at 120ft per round. (double dash).

I'm not saying GWM or PM is trash just that higher DPR is not the only thing this game is about.

[steppedonad4]

I'm not saying GWM or PM is trash just that higher DPR is not the only thing this game is about.

I really wish more players would realise that.

[JoeJ]

That's rather simplifying the issue, isn't it? My point would be that there aren't Feats that improve the rogue's ability to accomplish other tasks covered by the rules (and anything that is not covered by the rules is beyond the scope of the debate and irrelevant to it), not in the same way a feat like GWM/Sharpshooter helps a two-weapon wielder do damage. The feats are simply too good compared to the alternative. If the difference was smaller, there would be less issue.


Feats that synergize well with rogue abilities:
Crossbow Expert
Defensive Duelist
Dual Wielder
Dungeon Delver
Healer
Lucky
Mage Slayer
Mobile
Sentinel
Sharpshooter
Skulker

Feats that allow the rogue use their reaction to make another attack - Defensive Duelist, Mage Slayer, and Sentinel - are especially good because they can combine with sneak attack.

[Saggo]

There are few situations after this point that the rogue would ever be hit with an attack. The rogue could solo most encounters at zero resource expenditure (aside from a few arrows possibly). In a team situation the feats make scouting very doable, granting the party a surprise attack. If scouting goes south (somehow a minimum of 23 stealth failed to go undetected) then you can still run away through rough terrain at 120ft per round. (double dash).

I'm not saying GWM or PM is trash just that higher DPR is not the only thing this game is about.
It was an honest question. What DPR ratio is all that worth?

I ask because if I said "Rogue can do all that but only 10% DPR" I'd guarantee most would disagree, vehemently, meaning most people have a rough ratio already in mind. Or at least ratios that don't work in mind.

[steppedonad4]

Feats that allow the rogue use their reaction to make another attack - Defensive Duelist, Mage Slayer, and Sentinel - are especially good because they can combine with sneak attack.

Defensive Duelist doesn't allow another attack. In fact, it's actually not that great for a rogue given Uncanny Dodge is arguably the better ability of the two.

[Kane0]

So if you have more than one source of advantage/disadvantage you can end up with advantage or disadvantage even though you have a source of the counterpoint?

For example a barbarian using reckless attack would have advantage to hit. Using this version of GWM he would not have disadvantage nor advantage (one source of each, they cancel out). If that same barbarian were attacking a prone target he would once again have advantage (2 sources of adv, 1 source of disadv).

[Shaofoo]

What handwave? The recommended/expected/default way a DM arbitrates a game is by adjusting stats like DC and AC, thus I can assume Proficiency and Advantage matter, and I can use scary math to find the statistical chance of success. The chance of success vs a DC doesn't go away or change just because any given DM ignores the ruleset in favor of drama, it just means it wasn't applied. None of that presupposes DM-less playing, so you built a scenario that was never in question.

You said that the only thing that we have in common are the rulebooks and that is not true. If you meant the only thing that we have in common that is equal among us then you are right but everyone has a DM and depending on the DM then the game changes.

The DM can be fully all about dice but not add more than a couple of battles, so DPR matters very little. If you made a character that is all about social interactions but you spend your days out in the wilderness away from most social encounters then such skills matter very little.

The chance of the sucess was 100% or 0% because the point is for an action to happen or not happen if the DM was to assign a result. Just because the task had a DC assigned doesn't mean that there are auto success or failures attached to them, a door can have a very high DC to lockpick but if you have the key then you automatically succeed. Sure the theoretical math stays but what matters is results in the game.


You're right, a game is a sum of its parts, that doesn't automatically preclude balancing each part. The premise hasn't strayed much from the start, DPR ratios without feats compared to DPR ratios with feats.

"Math will resolve everything" is not a thing I said. Or implied. To repeat, math is a medium, a way of discussing this game, that we all have access to and the game is predicated on. A "baseline" if you will. To go full circle, what I will always argue is simple, Math Matters.

I never said you did but the topic carries that implication that the game is lesser because of these feats existing, and it isn't just because of what you wrote but because of what others did. Maybe I just read it wrong and everyone is aware that the game isn't all about DPR and that there is much more to the game.

And sure you can discuss whatever but I would like to hear that everything is considered, even things outside of damage (which is has so far with the Rogue).

And sure Math Matters but use it wisely. It can be very easy to get lost in the calculations.


I mean, who doesn't like more DPR?

I recently played a Bard that spent most of his battles slinging Vicious Mockery for the disadvantage and for the satisfaction of killing people with insults and yo mama jokes. I didn't care that I was probably doing less than 2 damage per round since I was playing a concept that everyone was at least okay with.

If the Dm came up to me and offered to up my Vicious Mockery damage dice I wouldn't complain at all but I would not go out of my way to increase my DPR if it means sacrificing my character concept. I was fine with having the fighter and ranger have all the kills while I just stood in the background talking smack.
Of course I pulled out the big guns when it was needed.

[Malifice]

I'd like to see more discussion of xetheral's underlying point - kryx sort of bats it away by questioning the math, but if you make sure the math is right, I think this is a useful angle to take - how often will one build outshine another in a combat, or a combat day, rather than (or in addition to) the average DPR for each build. I don't know what the numbers will be, or what they should be, but I feel like it might be a bit more relatable to played experience. I have a very hard time saying what the appropriate gap in DPR should be.

The argument goes 'Spending a feat (or a whole level of class advancement) on GWM makes for a significant boost in raw DPR vs a narrow (but common) range of AC values for a narrow subset of classes (Barbarians, BM Fighters, Vengance Paladins)'.

The counter arguments are one or more of the following:



Those four classes represent a very narrow field of the 40+ odd classes available, and are designed to be excellent at dealing big amounts of damage, at the cost of next to zero options in the other pillars of the game.
Raw DPR isnt the be all and end all of a characters viability in combat. Monks (for example) get their combat oomph from Stunning strike. Moon Druids and Fiend 'Locks get it from tanking. Despite being good single target big hiters, those four classes have a total of zero AoE effects at their disposal.
The opportunity cost of taking the feat that could have been used on other feats or ASI's. While a decent boost in DPR is nice, failing a save can result in death or dominate effect (so reducing DPR to 0 or creating a -DPR situation on a dominate, where that damage is being used against the party). Resilient (Wisdom) or Lucky go a long way to ensuring you remain in the fight dealing damage.
A +2 to Strength is (in many cases) simply a better option than taking those feats. It buffs Str saves (the most common in the game) combat manouvers (grapple, shove) manouver DC's for BM Fighters, carrying and lifting capacity, and provides a further +1 to hit and damage with all melee and thrown weapons. In many cases your DPR is actually better from just boosting Str by +2 (and gaining +1/+1) instead of -5/+10.
The DPR only increases against a narrow (but common) range of expected AC's. Against some AC's (particularly high AC's) it actually becomes a poor choice to use the -5/+10, meaning you lose the advanatge of your feat in those battles.

[JoeJ]

Defensive Duelist doesn't allow another attack. In fact, it's actually not that great for a rogue given Uncanny Dodge is arguably the better ability of the two.

Oops! Good catch. The synergy with DD is that it makes use of the finesse weapon that the rogue is probably using. Mage Slayer and Sentinel are the ones that increase the opportunities to land a sneak attack.

[MaxWilson]

I'd like to see more discussion of xetheral's underlying point - kryx sort of bats it away by questioning the math, but if you make sure the math is right, I think this is a useful angle to take - how often will one build outshine another in a combat, or a combat day, rather than (or in addition to) the average DPR for each build. I don't know what the numbers will be, or what they should be, but I feel like it might be a bit more relatable to played experience. I have a very hard time saying what the appropriate gap in DPR should be.

Also, it's useful to know not just how often a build will shine by default, but also how easily the build can be made to shine by the player. An obvious example would be that a Skulker's Rogue's "hide when lightly obscured" ability looks fairly useless on paper until you realize that Darkvision only turns darkness into dim light, which means that all the rogue has to do is make sure the party has no torches and he can hide anywhere in the Underdark. Because players have agency, you don't just look at how strong the build is under average conditions--you need to account for the player(s) choosing not to play in average conditions.

Another example could be an entire party of Mobile adventurers, which plays out radically differently than a regular party.

[Xetheral]

Please do suggest if you think the percentages I outlined should be different. It was a rough estimate. I refined it a bit more in my Balance sheet which I posted below.

I don't think there is an abstract correct ratio. I'm just suggesting that your percentages may be too low to accomplish your goal of increasing balance, because the signal of the different average damages may be too low to show up in the noise created by the dice.


This comes back to the point of balance. Some people prefer more balance. Some people don't. If I can do it well then I don't see a reason to not try to fix the math errors.

No matter how much one prefers balance, does a X% difference matter if it's not perceivable by the players? (I can't claim conclusively that it isn't, I'm just raising the philosophical question.) That raises an interesting related question: if by chance the math was perfect but due to imperfect perception the players still felt it was imbalanced, would it be worthwhile to deliberately skew the math until its perceived as balanced? By favoring infrequent big hits, the perception that -5/+10 is imbalanced could be even stronger than the math suggests.


I don't have time tonight to check all the number, maybe some else can or I can look closer tomorrow.
By my numbers a Barbarian at 10:
GWM: Picks up GWM at 4, stat at 8. 36 DPR
Polearm: Picks up Polearm at 4, stat at 8. 35 DPR (fixed an issue where I didn't have a stat increase at 8)
Polearm+GWM: Picks up GWM at 4, Polearm at 8. 39 DPR

I can look close tomorrow. Though I don't think assuming Vuman is a good assumption for balance discussions. Nor does it look like you have added GWM's cleave nor PAM's provoke to the damage.

Cool, let me know if I'm making a mistake with the AnyDice syntax--it's a bit tricky. I mentioned in my post being unable to include the conditional extra attacks due to AnyDice's 5-second complexity limit, but excluding them just results in the builds being farther apart in average DPR than the numbers calculated by your sheets, which make my results underestimates: (all else equal) the closer the averages, the more likely it becomes that the underdog will come out on top in a given battle or adventuring day.


I'd like to see more discussion of xetheral's underlying point - kryx sort of bats it away by questioning the math, but if you make sure the math is right, I think this is a useful angle to take - how often will one build outshine another in a combat, or a combat day, rather than (or in addition to) the average DPR for each build. I don't know what the numbers will be, or what they should be, but I feel like it might be a bit more relatable to played experience. I have a very hard time saying what the appropriate gap in DPR should be.

As much as I like my method, looking at things probabilistically as I did in my example is computationally intensive, and hard to generalize to the impressive detail level of Kryx's DPR model. AnyDice is an amazingly powerful tool, and even it can't fully handle the limited example I chose. I intend for my example to illustrate my point that even a 15% difference in average DPR will frequently be unnoticeable amid the random factor introduced by the dice, rather than as a general approach to analyzing balance. For normal use, Kryx's approach is preferable because it's simpler to use while simultaneously being more comprehensive.

[djreynolds]

I posted a thread about beating a barbarian, in it we see that the barbarian doesn't need a shield because of his damage reduction and in the end offense is his best defense. That is his flavor. That's his purpose. And it is really cool.

The barbarian is really just a well made class and this may be the problem and not GWM feat. Now I'm not sure what everyone else is rolling, but with 15/14/13/12/10/8 the barbarian can have a decent armor class with or without armor, I would definitely select GWM at 4th and use it. The barbarian's lack armor makes him great for stealth, and unlike the paladin the barbarian has no need to multiclass for an adequate ranged attack, because they have a decent dexterity score of at least 14.

Now I'm not suggesting everyone dump their class and play a barbarian, but they are great and they can tank and turtle when they have to. They're really the best made class IMO, humbly said. They have a great skill set designed for battle and skills for adventuring. But they have their pitfalls.

Rage is limited, it last for 1 minute about 10 rounds, and he only has so many a day.
Reckless attack, is just that and it opens him up to big damage from the enemy in return.
And in social aspects they should bring a lot of attention to themselves, whether that's negative or positive is on the game.
And intimidation isn't the best social skill in most social circles, merchants tend to over charge, reckless attack and rage don't work well with the entire city guard.

I like shield master for them because of advantage on strength checks during rage, but GWM is the best feat to capitalize on their strengths. In fact, if wolf totem is selected, it makes shield master obsolete for advantage to allies because the 3rd level Wolf totem is giving that already. But if you rolled 4d6 for incredible stats and can fit it, take shield master because you will need a shield at some point, like when you run out of rage or have weigh not using rage now for later and you know have an option for you bonus action.

What needs to be designed is a combat expertise feat. Taking the dodge action means not attacking for disadvantage and defensive duelist forces you to have 13 in dexterity, not expensive, and to use a rapier and its only your reaction to one attack.
It really was one of my favorite feats for the defensive tank.

[Saggo]

You said that the only thing that we have in common are the rulebooks and that is not true. If you meant the only thing that we have in common that is equal among us then you are right but everyone has a DM and depending on the DM then the game changes.

Check again, I did, just to be safe, and I never said only. I felt it was safe to assume when one argues scenarios are often resolved with a dice roll that the DM is the one creating and arbitrating the scenario.

At any rate, not having a DM is the abnormal case and I never built an argument for it, so thats a rabbit trail we can quit pursuing.


The chance of the sucess was 100% or 0% because the point is for an action to happen or not happen if the DM was to assign a result. Just because the task had a DC assigned doesn't mean that there are auto success or failures attached to them, a door can have a very high DC to lockpick but if you have the key then you automatically succeed. Sure the theoretical math stays but what matters is results in the game.

Most players likely found the key with a Perception check, so you know, more math.

DM fiats are arbitrary by definition and anecdotal by nature, which is why an objective baseline for a public forum (any literal forum, not just the internet) is useful and dare I say necessary. We all have the rule set. The rule set is applied statistics. Statistics is just math. With a little logical leap, that means we all have the math, regardless of what a DM arbitrarily rules.


And sure you can discuss whatever but I would like to hear that everything is considered, even things outside of damage (which is has so far with the Rogue).
This is a thread on the math of the combat portion so you're not going to find much discussion on things not combat or math, but it was actually pretty explicitly called out in the opening argument, if you happened missed:


By choosing to allow -5/+10 Polearm+GWM becomes the best melee build choice by far. GWM and Polearm itself as now lesser options (at all tiers). Classes that previously played other roles (skills, CC, stunning) that used to put up fairly good damage (about 75-90% depending on level and class) now put up ~55-70%.

[Shaofoo]

Check again, I did, just to be safe, and I never said only. I felt it was safe to assume when one argues scenarios are often resolved with a dice roll that the DM is the one creating and arbitrating the scenario.

At any rate, not having a DM is the abnormal case and I never built an argument for it, so thats a rabbit trail we can quit pursuing.


My point was that even when dice are rolled that the DM chooses the results. Sure it is common that the DM would use the dice to determine but he isn't shackled.



Most players likely found the key with a Perception check, so you know, more math.

I said that the DC check to open a lock is Very Hard unless you have the key, I never said that the party had to find a key. For all you know I described the door of the PC home base. You can't just assume math is always at work.


DM fiats are arbitrary by definition and anecdotal by nature, which is why an objective baseline for a public forum (any literal forum, not just the internet) is useful and dare I say necessary. We all have the rule set. The rule set is applied statistics. Statistics is just math. With a little logical leap, that means we all have the math, regardless of what a DM arbitrarily rules.

Sure I never said that we have no math but what math there is and how much it matters does vary. Sure we all need a baseline but we must also be aware that such baselines only matter for discussion and not in actual play. For discussion there needs to be a common ground but we also must be aware that common ground stops as soon as we leave the forum.



This is a thread on the math of the combat portion so you're not going to find much discussion on things not combat or math, but it was actually pretty explicitly called out in the opening argument, if you happened missed:

I read this part instead




Here I'll present my argument for why -5/+10 is bad for the game.


The way he framed it is that this feat is bad for the whole game and that it is so powerful that the other parts of the game are rendered invalid due to the sheer DPR. So while I never expected there to be anything other than DPR number crunching it did display a picture that the game as a whole is lesser because of it.

[smcmike]

The argument goes 'Spending a feat (or a whole level of class advancement) on GWM makes for a significant boost in raw DPR vs a narrow (but common) range of AC values for a narrow subset of classes (Barbarians, BM Fighters, Vengance Paladins)'.

The counter arguments are one or more of the following:



Those four classes represent a very narrow field of the 40+ odd classes available, and are designed to be excellent at dealing big amounts of damage, at the cost of next to zero options in the other pillars of the game.
Raw DPR isnt the be all and end all of a characters viability in combat. Monks (for example) get their combat oomph from Stunning strike. Moon Druids and Fiend 'Locks get it from tanking. Despite being good single target big hiters, those four classes have a total of zero AoE effects at their disposal.
The opportunity cost of taking the feat that could have been used on other feats or ASI's. While a decent boost in DPR is nice, failing a save can result in death or dominate effect (so reducing DPR to 0 or creating a -DPR situation on a dominate, where that damage is being used against the party). Resilient (Wisdom) or Lucky go a long way to ensuring you remain in the fight dealing damage.
A +2 to Strength is (in many cases) simply a better option than taking those feats. It buffs Str saves (the most common in the game) combat manouvers (grapple, shove) manouver DC's for BM Fighters, carrying and lifting capacity, and provides a further +1 to hit and damage with all melee and thrown weapons. In many cases your DPR is actually better from just boosting Str by +2 (and gaining +1/+1) instead of -5/+10.
The DPR only increases against a narrow (but common) range of expected AC's. Against some AC's (particularly high AC's) it actually becomes a poor choice to use the -5/+10, meaning you lose the advanatge of your feat in those battles.


I understand this argument, and generally agree with the points you listed. I also feel that a large damage gap between Conan and his halfling acrobat friend is appropriate (but hope that the DM gives everyone a chance to shine).

I just thought Xetheral brought up an interesting point that more directly addressed the math problem within its narrow boundaries.



As much as I like my method, looking at things probabilistically as I did in my example is computationally intensive, and hard to generalize to the impressive detail level of Kryx's DPR model. AnyDice is an amazingly powerful tool, and even it can't fully handle the limited example I chose. I intend for my example to illustrate my point that even a 15% difference in average DPR will frequently be unnoticeable amid the random factor introduced by the dice, rather than as a general approach to analyzing balance. For normal use, Kryx's approach is preferable because it's simpler to use while simultaneously being more comprehensive.

I didn't realize how hard it was to compute. Still, I like your point, and would like to see Kryx address it squarely.

[Kryx]

how often will one build outshine another in a combat, or a combat day, rather than (or in addition to) the average DPR for each build.
If you want to see DPR over a whole day just multiply DPR by 25.7. For example:

Barbarian GWM at 11 does 42.52 DPR. So that's 1092.76 damage in a day.
Rogue TWF at 11 does 26.95 DPR. So that's 692.61 damage in a day.

Still the same percentages. Rogue does 63.38% of a Barb GWM at 11. Or viewed in the other perspective Barb does 57% more damage than a Rogue.




I like the disadvantage / +10 idea. My group uses the cancel adv/dis at 1:1 houserule, will definitely try it.
This wouldn't fix the issue. Barb now has 65% chance to hit as normal while other classes have 42.25%. Both have +10 to damage. It's basically the same situation.




I don't think there is an abstract correct ratio. I'm just suggesting that your percentages may be too low to accomplish your goal of increasing balance, because the signal of the different average damages may be too low to show up in the noise created by the dice.
Everyone has an expected level of balance, they just might not have thought it through.
2% is too small for humans to notice, but luckily we have math to do that for us.


No matter how much one prefers balance, does a X% difference matter if it's not perceivable by the players?
I don't agree that it's not perceptible (especially in this case) nor that perception is what should determine balance.


Cool, let me know if I'm making a mistake with the AnyDice syntax--it's a bit tricky.
Here is your second example w/o Vuman: http://anydice.com/program/6e79
Looking at your original math a Polearm does less than Polearm+GWM so I'm not sure what conclusions you're drawing. By your second original link the Polearm does 822.66 damage and the Polearm+GWM does 953.72

I don't understand your point. Some combat will have different results, yes, but overall one option does more damage than another. Maybe you're hinting back to "how much can players notice" and in this case I can guarantee a player will notice it when a Barb consistently does nearly double the damage of a Rogue (based on some of the numbers I put up a page or two ago).


Still, I like your point, and would like to see Kryx address it squarely.
Please let me know if I have not addressed it "squarely". As I said above his math actually shows my same conclusion that -5/+10 gives more damage over a day (though it falls short of capturing Cleave or Polearm Provoke as we've both acknowledged).




The way he framed it is that this feat is bad for the whole game and that it is so powerful that the other parts of the game are rendered invalid due to the sheer DPR. So while I never expected there to be anything other than DPR number crunching it did display a picture that the game as a whole is lesser because of it.
That original phrasing is a bit off. As the thread progressed the exact problem has been refined to specifically call out where -5/+10 is an issue and that is when combined with advantage. The most common cases being Barbarian, OoV Paladin, BM Fighter(to a lesser extent), and Bladelock 17+. The situation is still the same.



Malifice's comments are the same as the beginning as if this thread hasn't been discussing them the whole time. Most have been previously addressed. Summary:
It's not a narrow case. 3/12 classes can present the problem.
No one has ever said DPR was the "be all end all". Please stop biasing the argument by incorrectly stating the opposing opinion's argument.
It's a matter of valuation. Dealing way outside the normal bounds of damage is what is being discussed, not a 25% increase in DPR.
AC is based on the DMG. If you choose to alter monsters then of course it's different. But that's homebrew.

[Tanarii]

No matter how much one prefers balance, does a X% difference matter if it's not perceivable by the players? (I can't claim conclusively that it isn't, I'm just raising the philosophical question.)i think you've got it backwards. The players in general are likely to percieve it as *more* game changing than it is looking at just the DPR increase. We're talking about an effect that comes with a huge damage increase on hit, but more misses. The human mind will generall tend to see either of those outliers and amplify the effect of GWM as either useless (dude you always miss stop GWMing ffs!) or overpowered (dude +10 damage h4xx0r, DM ban nao!).

[Malifice]

If you want to see DPR over a whole day just multiply DPR by 25.7. For example:

Barbarian GWM at 11 does 42.52 DPR. So that's 1092.76 damage in a day.
Rogue TWF at 11 does 26.95 DPR. So that's 692.61 damage in a day.

Still the same percentages. Rogue does 63.38% of a Barb GWM at 11. Or viewed in the other perspective Barb does 57% more damage than a Rogue.




This wouldn't fix the issue. Barb now has 65% chance to hit as normal while other classes have 42.25%. Both have +10 to damage. It's basically the same situation.




Everyone has an expected level of balance, they just might not have thought it through.
2% is too small for humans to notice, but luckily we have math to do that for us.


I don't agree that it's not perceptible (especially in this case) nor that perception is what should determine balance.


Here is your second example w/o Vuman: http://anydice.com/program/6e79
Looking at your original math a Polearm does less than Polearm+GWM so I'm not sure what conclusions you're drawing. By your second original link the Polearm does 822.66 damage and the Polearm+GWM does 953.72

I don't understand your point. Some combat will have different results, yes, but overall one option does more damage than another. Maybe you're hinting back to "how much can players notice" and in this case I can guarantee a player will notice it when a Barb consistently does nearly double the damage of a Rogue (based on some of the numbers I put up a page or two ago).


Please let me know if I have not addressed it "squarely". As I said above his math actually shows my same conclusion that -5/+10 gives more damage over a day (though it falls short of capturing Cleave or Polearm Provoke as we've both acknowledged).




That original phrasing is a bit off. As the thread progressed the exact problem has been refined to specifically call out where -5/+10 is an issue and that is when combined with advantage. The most common cases being Barbarian, OoV Paladin, BM Fighter(to a lesser extent), and Bladelock 17+. The situation is still the same.



Malifice's comments are the same as the beginning as if this thread hasn't been discussing them the whole time. Most have been previously addressed. Summary:
It's not a narrow case. 3/12 classes can present the problem.
No one has ever said DPR was the "be all end all". Please stop biasing the argument by incorrectly stating the opposing opinion's argument.
It's a matter of valuation. Dealing way outside the normal bounds of damage is what is being discussed, not a 25% increase in DPR.
AC is based on the DMG. If you choose to alter monsters then of course it's different. But that's homebrew.

It's not 3/12 classes. It's one archetype of each of 2 classes and a Barbarian using reckless attack (which carries its own risk/ reward). So about 4/50 class archetypes from the PHB.

And it's not 'home brew' to suggest the existence of higher ACs than the DMG suggests as 'expected' at various levels. At 5th level they suggest AC14 which is easily surpassed by any NPC in armour or shield. Instead of hand waving the phenomenon away how about being honest about the mechanical effects of high AC against GWM. As in - simple STR increases become a mathmatically better choice at certain (common) AC ranges over GWM.

If you're going to be intellectually dishonest and hand wave legitimate criticisms as 'house rules' or misrepresent the number of arcetypes that benefit from the feat then I'm not going to discuss this with you any further man. I'm crirical of your theory. I don't think it stands up to thorough falsification. If you want to discuss those points in an open and frank manner you know where to find me.

You're not counting all the variables, you're discounting opportunity cost, it's a corner case phenomenon and it breaks nothing to leave it as is.

[smcmike]

Kryx, it seems to me you again completely miss the point I was interested in: the odds that one build would outdamage another in any given encounter, which is a very different question than the average damage per round or damage per encounter. I'm not saying that this angle would completely change the analysis - you can tell me why you think this doesn't add anything, if you like. But I admit I'm a little mystified by your current response of just repeating DPR and multiplying out to DPE and DPD.

Similarly, you can argue that the DMG is a good baseline for AC. I might even agree with this assumption. But variance from it is far less "homebrew" than cutting out or changing feats.

[Saggo]

My point was that even when dice are rolled that the DM chooses the results. Sure it is common that the DM would use the dice to determine but he isn't shackled.
It goes without saying that this is true, but you muddled it with arguments about DM-less play that were never brought into question and the implications that DM fiats are wildly frequent.

At any rate with regards specifically to DM fiats invalidating mechanics, good relationships establish agreements on what's expected and that includes tabletop DM-player relationships.


I said that the DC check to open a lock is Very Hard unless you have the key, I never said that the party had to find a key. For all you know I described the door of the PC home base. You can't just assume math is always at work.

Or the door to a boss' lair. That's not helping anyone's case.


Sure I never said that we have no math but what math there is and how much it matters does vary. Sure we all need a baseline but we must also be aware that such baselines only matter for discussion and not in actual play. For discussion there needs to be a common ground but we also must be aware that common ground stops as soon as we leave the forum. That's not really true for the simple fact that if enough of a ruleset is ignored by DMs or players, we're no longer playing the same game and discussing "5th Edition" is rendered moot. Thus anyone can assume that another table is using a non-trivial or significant amount of the ruleset, or more importantly a non-trivial or significant amount of the math.

[Theodoxus]

Seems to me, the best thing would be to either get rid of the -5/+10 as suggested, or add it as a native option to all weapons (though I still prefer -(Prof), +(Prof*2) for that).

The problem seems to stem from the fact that Polearms are considered Great Weapons, so there's synergy between GWM and PAM. Similarly, there's synergy between XBX and SS.

S&B and TWF don't have additive synergistic feats. Duelist and Shieldmaster provide side benefits that aren't cumulative. AFAIR, TWF only has one direct feat to support it.

I'd prefer to grant half feats that provide a unique benefit - they're already in game - though the question is whether they should be unique (can only affect one class of weapons) or not.

If the Cleave of GWM only worked with non-reach Heavy Weapons, that removes all the benefits of Polearm Master. If the Reaction provided by PAM required a reach weapon, that removes all the benefits of GWM (and quarterstaff, but opens other reach weapons, which are arguably polearms, unlike a staff, which is controversial and makes no sense with shields, etc.)

XBX and SS are harder to separate... though I could see combining them. Removes the loading feature, negates partial and half cover and removes disad when shooting within 5'. Of those, the cover is the strongest feature - but all three combine to make a decent feat and any bowman (x-bow or normal) would want.

That leaves S&B and TWF. TWF specifically has been argued round and round, with no real resolution ever agreed upon. Some see it more defensive, akin to S&B with slightly more damage potential; others think it should be on par with great weapon fighting. Where a Greatsword deals 2d6 on 1 attack, and paired short swords deal 2d6 on 2 attacks. Honestly, I don't know where the numbers should align, but I do like the idea of trading accuracy for damage - perhaps trade accuracy for protection, ala the 3.P defensive style feats?

[choryukami]

Ranger and Valor Bard can use Great Weapons too.

What I would do as a house rule is to improve Two Weapon Fighting by adding Two Weapon Rend to the Dual Wielder Feat.

*At level 6 or higher, if you make an attack with the attack action and a bonus action attack with an off-hand weapon and hit at least once with each one, you deal extra damage equal to your ability modifier when the second attack hits.

Something like that. Since DW is underpowered.

[Xetheral]

I didn't realize how hard it was to compute. Still, I like your point, and would like to see Kryx address it squarely.

Thanks! :) As an example of the computation problem, before I realized that I could have AnyDice calculate the likelihood that one build outdamaged the other (the second link in each example), I exported the data from the comparison (the first link in each example) and tried to have Excel due a brute-force comparison. It worked for the per-combat example (yielding the same number AnyDice did later), but choked on the per-day example: even after turning off the Undo feature, Excel quickly chewed through 8 gigs of RAM and returned an "insufficient resources" error. I'm sure I could streamline the calculation (using array functions to create a "SumProductIf" function would be the most likely candidate, or else turn to VBA), but it's still illustrative of the sheer scope of the problem.


Everyone has an expected level of balance, they just might not have thought it through.

Hmm, interesting point. While I agree with the literal statement, I'm not sure my expected level of mathematical balance for damage dealing can be defined by setting proportional differences between the classes for average DPR. For example, I'm not sure that my preferred balance point between sustained-DPR classes and Nova classes hinges on their average DPR ratio. For me, the rate of the Novas and absolute size of the Novas would also be important considerations.


I don't agree that it's not perceptible (especially in this case)

That's fine. :) What level of difference in the average will be perceptible will greatly vary from person to person anyway, and, as Tanarii notes below, the perception may be skewed more by outliers than by averages.


I don't agree that... perception is what should determine balance.

Fair enough. From my standpoint it seems that balance is fundamentally an issue of player perception, but it's fine that we disagree. Note that I still think the math is important because it gives us a way to potentially predict that perception without needing problematic quantities of survey data.


Looking at your original math a Polearm does less than Polearm+GWM so I'm not sure what conclusions you're drawing. By your second original link the Polearm does 822.66 damage and the Polearm+GWM does 953.72

I don't understand your point. Some combat will have different results, yes, but overall one option does more damage than another.

Sorry if I was unclear! I'll try to re-explain.

Smcmike has the right of it: I calculated two things:
The frequency with which, due to the variability of the dice, the damage dealt by the build with the lower average will exceed the damage dealt by the build with the higher average. The frequency with which, due to the variability of the dice, the damage dealt by the two builds will be close enough that an average difference may be imperceptible.
I calculated both figures for each a 5-round combat and a 25-round adventuring day so that the results would be compatible with your spreadsheet. What I found was that in 30.24% of combats, the variability introduced by the dice was sufficient that the character who does less average damage nevertheless will deal more damage. Over the course of a day, that figure drops to 11.84%. I also found that in 20% of combats, the damage totals were within an arbitrary threshold of 10% of each other. Over the course of a day, that figure rises to 30%. (The links can be used to calculate for any threshold you like if you'd prefer something other than 10%. If it's not clear how to do that, let me know and I'll explain.)

Because my calculations can't include the extra attacks granted by the two feats, the average DPR shown in the calculations are understatements. This results in my showing a 15% difference in average DPR, which is more than your 10% difference for the same builds. The closer the average DPR, the higher all the figures will be in the preceding paragraph. Accordingly, all those numbers are themselves understatements.

The conclusion I draw is that if you change the rules to bring (e.g.) TWF to within 5 to 10 percent of GWM, you'll get a situation where "who does more damage?" will be answered primarily by the dice. Even with a 15% difference, my calculations show that the random factor is already very large, and that will get bigger if you introduce smaller average differences.

(Note: a 15% difference in average DPR between TWF and GWM isn't going to have the exact same distribution as the 15% difference I explored in my calculations. (That's why I labeled by calculations as illustrative.) I believe the difference is likely to be negligible, but to the extent that it isn't that means that not all 15% differences are created equal, which complicates everyone's math.)


i think you've got it backwards. The players in general are likely to percieve it as *more* game changing than it is looking at just the DPR increase. We're talking about an effect that comes with a huge damage increase on hit, but more misses. The human mind will generall tend to see either of those outliers and amplify the effect of GWM as either useless (dude you always miss stop GWMing ffs!) or overpowered (dude +10 damage h4xx0r, DM ban nao!).

I completely agree! My argument was that if Kryx changed the game to bring the overall DPR figures much closer than they currently are, the random chance might start outweighing the difference in the average. As for whether or not -5/+10 is a problem at the table:


By favoring infrequent big hits, the perception that -5/+10 is imbalanced could be even stronger than the math suggests.

[Tanarii]

Personally, I've decided not to worry as much about GWM as much I was previously, and not worry about Sharpshooter at all. The feats seems to mostly affect classes that both: 1) can't get added damage; 2) regularly get advantage. That means BM Fighters (with Trip) and Barbarians.

I have no problem with Barbarians 'requiring' GWM as a build option. As far as I'm concerned they already required a two-handed weapon build, and if one of their five ASIs is' 'comitted' to GWM instead of Str/Con, got no issues there. So *within* the Barbarian class, I don't see a build vs build issue. In terms of Barbarian vs other classes, I might consider house-ruling the damage bonus down a bit when used with advantage when running a game.

Unfortunately for Fighters it is an internal build issue, giving a fighting style, Archetype, and Maneuver a significant damage advantage over the others. As far as I'm concerned, that's the biggest issue ... not the disparity between damage other classes, but the disparity within the class, making Fighter damage builds other than BM/GWM/Trip builds relatively superfluous. The fix is probably the same as with Barbarians ... reduce the bonus damage when the feat is used with advantage.

I still don't like Polearm Mastery at all. Personal bias. Anything that makes Polearms even close to as common as other weapons in personal combat is that stuff I don't like. Mass combat, against stupid charging animals (ie Boar Spear), or niche monk/martial artist builds is my preference.


I completely agree! My argument was that if Kryx changed the game to bring the overall DPR figures much closer than they currently are, the random chance might start outweighing the difference in the average. As for whether or not -5/+10 is a problem at the table:Yup I totally missed your other statement.

[Kryx]

@Malifice: I summarized the above for others as I've literally addressed your complaints several times now and yet you just repeat them as if there was no discussion happening. Without a dialogue it's just two people angrily spouting opinions at one another. No thank you.




Kryx, it seems to me you again completely miss the point I was interested in: the odds that one build would outdamage another in any given encounter, which is a very different question than the average damage per round or damage per encounter. I'm not saying that this angle would completely change the analysis - you can tell me why you think this doesn't add anything, if you like. But I admit I'm a little mystified by your current response of just repeating DPR and multiplying out to DPE and DPD.
You did not specify what was interesting. As apparent by my exchange with Xetheral I did not grasp his full argument. So please if you're after something then ask for it directly.

On to the topic at hand:
Damage in a given encounter is very different from DPR. It is relevant to ensure that there isn't too much nova, but that's an entirely different topic than damage over an adventuring day.
If you only looked at damage in a single encounter then classes with resources would be able to burst an encounter. Paladin for example would double or triple a Fighter in a single encounter, but then would be a limp noodle the rest of the adventuring day.
More on Nova in my reply to Xetheral below.


Similarly, you can argue that the DMG is a good baseline for AC. I might even agree with this assumption. But variance from it is far less "homebrew" than cutting out or changing feats.
Altering the base game is homebrew. Choosing a different scale for AC for your monsters changes the balance of the whole game. It would be something to mention, but it isn't a factor for a discussion based on the game's rules.




The problem seems to stem from the fact that Polearms are considered Great Weapons, so there's synergy between GWM and PAM.
The problem is -5/+10 and advantage. That is true for GWM, Polearm+GWM, and SS+advantage.




What I would do as a house rule is to improve Two Weapon Fighting by adding Two Weapon Rend to the Dual Wielder Feat.
TWF does indeed suck, but that's true no matter if -5/+10 is in the game or not. It's just exaggerated more if -5/+10 is in the game. See https://docs.google.com/spreadsheets/d/1d-9xDdath8kX_v7Rpts9JFIJwIG3X0-dDUtfax14NT0/edit#gid=2141556455 which compares many options to GWM.




Hmm, interesting point. While I agree with the literal statement, I'm not sure my expected level of mathematical balance for damage dealing can be defined by setting proportional differences between the classes for average DPR. For example, I'm not sure that my preferred balance point between sustained-DPR classes and Nova classes hinges on their average DPR ratio. For me, the rate of the Novas and absolute size of the Novas would also be important considerations.
Nova vs sustained is a different topic than most of the others though.
For example it seems most people have a fairly clear idea how much stronger melee classes should be compared to ranged as they may not always be able to target opponents and they have to put themselves in vulnerable positions to do that damage. Most people seem to have an idea how much stronger a Fighter should be compared to a Rogue or a non spellcasting class like Ranger compares to a pure fighting like like Fighter.

I do agree that Nova wouldn't fall in that same category. Though if we make the assumption that groups play by the DMG recommended adventuring day (which seems the only logical starting point for discussing balance as that is what 5e is balanced on) then I don't think Nova is a huge concern. Even if the game designers did put in some ridiculous nova capabilities that is balanced out by that character being awful in every other encounter. Paladin for example would do really mediocre damage without resources.
Given the Adventuring Day assumption I'm not too concerned about nova, but maybe there is more to it. I'd be curious to hear what you think.


That's fine. :) What level of difference in the average will be perceptible will greatly vary from person to person anyway, and, as Tanarii notes below, the perception may be skewed more by outliers than by averages.
Ya, what Tanarii writes is true for the most part I think.


Fair enough. From my standpoint it seems that balance is fundamentally an issue of player perception, but it's fine that we disagree. Note that I still think the math is important because it gives us a way to potentially predict that perception without needing problematic quantities of survey data.
I'm glad we can disagree here. It is honestly refreshing to just discuss it directly. I do favor the math as I believe, as Saggo has pointed out, that the game is built on math. Based on the numbers I've run the developers recognize that as well as their game is fairly balanced in most regards.
I've honestly never understood the playtest or "feeling" oriented balance approach. To me that's secondary after the math to make the balanced mechanics feel fun. Though not everyone agrees.


The frequency with which, due to the variability of the dice, the damage dealt by the build with the lower average will exceed the damage dealt by the build with the higher average.
The frequency with which, due to the variability of the dice, the damage dealt by the two builds will be close enough that an average difference may be imperceptible.

The conclusion I draw is that if you change the rules to bring (e.g.) TWF to within 5 to 10 percent of GWM, you'll get a situation where "who does more damage?" will be answered primarily by the dice. Even with a 15% difference, my calculations show that the random factor is already very large, and that will get bigger if you introduce smaller average differences.
I got the imperceptible argument before. The argument that the dice play a big factor was missed.
Dice factor
Dice do indeed play a big factor in a bounded accuracy system. I don't think it's necessarily bad that two builds wouldn't seem that different. Two builds that focus pure DPR shouldn't be very perceptively different no matter how they achieve that DPR (via Polearm, GWM, or TWF barring other (some unmeasurable) considerations like reach, nova, spells, etc). By RAW you could say that many builds of similar nature are within 10% of eachother (with some larger variances).

In the end I'm left wondering why is the frequency relevant? In my eyes total damage dealt in a round/day/whatever is the metric to be concerned about as the dice will have their way. The game is balanced on the dice average, so why is average not a fair representative?

Perceptibility
By RAW you could say that about a lot of builds. Most builds of similar nature are within 10% of eachother (with some larger variances). I guess I don't see the problem of lack of perceptibility. Though there are probably many who do.




Personally, I've decided not to worry as much about GWM as much I was previously, and not worry about Sharpshooter at all.
For ranged it's quite important as both Longbow and Heavy Crossbow are beat by Hand Crossbow in nearly every case.

[Corran]

I don't agree that it's not perceptible (especially in this case) nor that perception is what should determine balance.

Exactly. Being happy in our ignorance isn't really an option now, is it? The game should be balanced whether random player Bob can understand it or not.

[Tanarii]

For ranged it's quite important as both Longbow and Heavy Crossbow are beat by Hand Crossbow in nearly every case.Yeah, like PAM, I really don't like the bonus attack granted by CE either ... especially since my initial assumption was it was there to allow off-hand Hand Crossbow attacks (ie for Drow). I've focused on the Advantage + bonus damage issue, which is an artifact of penalty to hit being overly mitigated by advantage. But bonus attack + bonus damage has it's own issues, as it always does. If you add too much bonus damage to bonus attack (no matter how it is obtained), it's an issue. That's always been true for TWF in every edition, which is why it tends to be nerfed all to hell and back.

[Coffee_Dragon]

Do the Hand Crossbow numbers assume self-loading crossbows?

[Kryx]

Do the Hand Crossbow numbers assume self-loading crossbows?
Nope. By RAW Crawford chose the worst option and said the bonus attack can be made with the same crossbow.

I, personally, don't allow one hand crossbow to make the bonus attack. I treat Hand Crossbows like any other TWF weapon like shortswords, daggers, thrown handaxes, etc. I allow the purchase of wrist crossbows which strap on to the arm.

[Coffee_Dragon]

Nope. By RAW Crawford chose the worst option and said the bonus attack can be made with the same crossbow.

I, personally, don't allow one hand crossbow to make the bonus attack. I treat Hand Crossbows like any other TWF weapon like shortswords, daggers, thrown handaxes, etc. I allow the purchase of wrist crossbows which strap on to the arm.

But you do allow firing two bolts every turn? If so, why is it more reasonable to load two crossbows once each instead of the same one twice?

[Kryx]

But you do allow firing two bolts every turn? If so, why is it more reasonable to load two crossbows once each instead of the same one twice?
TWF requires a sacrifice of having the offhand holding a second weapon. Second weapon also matters if you use magic items.
Not to mention TWFing with a wrist crossbow prevents switching of weapons as it's strapped on.

Plus my biased opinion that demon hunters are cool and a single one feels munchkin-y.

[smcmike]

You did not specify what was interesting. As apparent by my exchange with Xetheral I did not grasp his full argument. So please if tyou're after something then ask for it directly.

On to the topic at hand:
Damage in a given encounter is very different from DPR. It is relevant to ensure that there isn't too much nova, but that's an entirely different topic than damage over an adventuring day.
If you only looked at damage in a single encounter then classes with resources would be able to burst an encounter. Paladin for example would double or triple a Fighter in a single encounter, but then would be a limp noodle the rest of the adventuring day.
More on Nova in my reply to Xetheral below.


Altering the base game is homebrew. Choosing a different scale for AC for your monsters changes the balance of the whole game. It would be something to mention, but it isn't a factor for a discussion based on the game's rules.


Nova vs sustained is a different topic than most of the others though.
For example it seems most people have a fairly clear idea how much stronger melee classes should be compared to ranged as they may not always be able to target opponents and they have to put themselves in vulnerable positions to do that damage. Most people seem to have an idea how much stronger a Fighter should be compared to a Rogue or a non spellcasting class like Ranger compares to a pure fighting like like Fighter.

I do agree that Nova wouldn't fall in that same category. Though if we make the assumption that groups play by the DMG recommended adventuring day (which seems the only logical starting point for discussing balance as that is what 5e is balanced on) then I don't think Nova is a huge concern. Even if the game designers did put in some ridiculous nova capabilities that is balanced out by that character being awful in every other encounter. Paladin for example would do really mediocre damage without resources.
Given the Adventuring Day assumption I'm not too concerned about nova, but maybe there is more to it. I'd be curious to hear what you think.

Dice factor
Dice do indeed play a big factor in a bounded accuracy system. I don't think it's necessarily bad that two builds wouldn't seem that different. Two builds that focus pure DPR shouldn't be very perceptively different no matter how they achieve that DPR (via Polearm, GWM, or TWF barring other (some unmeasurable) considerations like reach, nova, spells, etc). By RAW you could say that many builds of similar nature are within 10% of eachother (with some larger variances).

In the end I'm left wondering why is the frequency relevant? In my eyes total damage dealt in a round/day/whatever is the metric to be concerned about as the dice will have their way. The game is balanced on the dice average, so why is average not a fair representative?

Perceptibility
By RAW you could say that about a lot of builds. Most builds of similar nature are within 10% of eachother (with some larger variances). I guess I don't see the problem of lack of perceptibility. Though there are probably many who do.


I'm sorry if I have seemed argumentative or indirect. That was not my intention.

My question isn't about a nova build v. a non-nova build. I agree that this would complicate the discussion, and would prefer to stick to the parameters you laid out, of GWM v non-GWM melee builds - the fighter or barbarian or rogue cases you are concentrating on (though the existence of Nova builds suggests that balancing internally in this subset may raise larger concerns).

In a given random adventuring day- the sort of day you are basing your DPR numbers on - what are the odds that the GWM build does more damage than the S&B build? On any given encounter during that day?

I do agree that DPR is a valuable tool in determining balance, I just thought that this sort of consideration could help explain what it really means.

[Coffee_Dragon]

TWF requires a sacrifice of having the offhand holding a second weapon.

That's a meta-consideration. In the game world, what's the reason you can load your left-hand crossbow with your right hand (that also holds a crossbow) and your right-hand crossbow with your left hand (that also holds a crossbow) but not loading one crossbow with the other hand (that is free) twice?


Not to mention TWFing with a wrist crossbow prevents switching of weapons as it's strapped on.

Well, apparently not, if you can use the hand for other stuff already?

[Kryx]

In a given random adventuring day- the sort of day you are basing your DPR numbers on - what are the odds that the GWM build does more damage than the S&B build? On any given encounter during that day?
I must be missing something as I would answer these questions the same as before by multiplying the DPR by 5 for an encounter or 25.7 for the adventuring day. The ratios stay the same as DPR. But that wasn't what you wanted last time.




Hand crossbow
If you want to discuss hand crossbows and the fluff/mechanics of how they work then please start a thread. It is not relevant to this discussion.
To fully explain how I use hand crossbows on my sheet:
RAW follows RAW which is 1 handed hand crossbow per the feat and Crawford.
Houserules treats hand crossbow like TWF weapons.

[Tanarii]

I must be missing something as I would answer these questions the same as before by multiplying the DPR by 5 for an encounter or 25.7 for the adventuring day. The ratios stay the same as DPR. But that wasn't what you wanted last time.He's probably asking for the variance. I don't know the exact statistical term for that. As in ... what is the likelyhood that Character A without GWM will actually outperform Character B with GWM on a given day. Average is average ... it doesn't always account show variance or deviation from the average. As the average between the two increases, that likelyhood will eventually approach, but potentially not be, zero.

Edit: Of course, probability of variance across a larger number of rolls goes down, provided you can come up with reasonable assumptions. Which is the entire reason people focus on DPR anyway. And others object to it so much. ;)

[Shaofoo]

It goes without saying that this is true, but you muddled it with arguments about DM-less play that were never brought into question and the implications that DM fiats are wildly frequent.

The main point was never about DM-less play, you just held onto it for some reason. The point is that we all have a DM and the rules, there was nothing about playing with no DM and I believe you took my off comment at full value because I said the only way it can all be equal is that there would be no DM.


At any rate with regards specifically to DM fiats invalidating mechanics, good relationships establish agreements on what's expected and that includes tabletop DM-player relationships.

That has nothing to do with the topic at all. I am only adding DM fiat as a way of saying that even when the dice are cast the DM can disregard it.


Or the door to a boss' lair. That's not helping anyone's case.

Actually it proves that you are arguing with bad faith. You say that there has to be some way to discuss the game and I place a situation and you add variables to it just to prove a point that somehow math is always present even though you twisted my original premise into your own. The point is that just as there can be dice rolls they can also be circumvented by the DM and it isn't necessarily fiat. And I am not interested in playing Six Degrees of Mystery Keys so we can drop this little simulation.


That's not really true for the simple fact that if enough of a ruleset is ignored by DMs or players, we're no longer playing the same game and discussing "5th Edition" is rendered moot. Thus anyone can assume that another table is using a non-trivial or significant amount of the ruleset, or more importantly a non-trivial or significant amount of the math.

Does it matter?

And you are proving my point.

If you want to argue in a controlled environment (and not change the environment on me) then do so but you can't somehow pretend that the math will hold up in all the tables, just from the mere reason that tables have different focuses and some tables might actually change the rules, sometimes drastically.

[Kryx]

He's probably asking for the variance.
Got it. Then he'd have to do math simulations much more complex than mine to reach the answers for that.

Though I'd refer back to the quoted discussion with Xetheral:

Dice factor
Dice do indeed play a big factor in a bounded accuracy system. I don't think it's necessarily bad that two builds wouldn't seem that different. Two builds that focus pure DPR shouldn't be very perceptively different no matter how they achieve that DPR (via Polearm, GWM, or TWF barring other (some unmeasurable) considerations like reach, nova, spells, etc). By RAW you could say that many builds of similar nature are within 10% of eachother (with some larger variances).

In the end I'm left wondering why is the frequency relevant? In my eyes total damage dealt in a round/day/whatever is the metric to be concerned about as the dice will have their way. The game is balanced on the dice average, so why is average not a fair representative?

[Icewraith]

Perceptibility is actually extremely important. The whole idea behind optimization, aside from generic "being better", it that your choices in character build have a perceptible, ideally positive, effect on gameplay.

The idea behind balance is that if there are multiple ways to specialize in performing a particular game-related task, none of the ways should be so overwhelmingly superior to the others that a player regrets choosing differently.

What is missing from this discussion is that the player using -5/+10 , on top of having spent character resources to be able to do it in the first place, is actively choosing to take an additional risk in not hitting the target, in exchange for an increased benefit when they do connect. If using -5/+10 didn't significantly and perceptibly improve average damage dealt over time in the long run, nobody would use it, since the cost of missing an attack is quite high. Whiffing an additional time or two on the wrong monster can result in its ability (especially breath weapons) recharging and killing one or more party members. Missing all your attacks in a round also sucks from the perspective of a player. That's a perceptible disadvantage that the feat's increased damage when it hits needs to offset in order to mechanically reflect the player's choice.

The idea that -5/+10 with advantage is too strong also runs into the issue that, aside from the character usually having to expend a resource (such as maneuver dice, attacks, ki points, bonus actions and/or feats) or take a risk (opposed strength check, inflicted save, grant opponent advantage on attacking you) to gain advantage, advantage is often merely a way of countering disadvantage. A number of monsters have webs, fears, or the like that inflict disadvantage on one or more players. This is especially true of the big, set-piece bosses with legendary and lair actions. So, just because the Barbarian has a reliable means of generating advantage doesn't mean they will always get to roll 2d20 and pick the best two, and mindlessly use their -5/+10.

So yes, -5/+10 feats are "stronger" than other feats. Other feats only grant bonuses and do not involve risks. If taking a -5 penalty to hit resulted in a slight, imperceptible increase in average damage over time, it wouldn't be worth using.

The numbers look ridiculous at first glance, but when you consider the GWM+PM Barbarian is arguably taking the highest risk (-5 to hit penalty, in melee, opponents have advantage to attack you) on top of devoting a large number of resources (two feats), to do at best double (? This may be a rogue value, and rogues shouldn't necessarily compete with Barbarians when it comes to DPR) the average damage over time of a barbarian using the worst possible fighting style (from an offense perspective) and taking no risks who has not spent any character resources against an arbitrary average ac that isn't necessarily representative of the AC seen in the tougher fights, and that the cost of the increased miss rate isn't immediately made up for by the average damage but can result in 2-3 round miss streaks where you basically give your opponents an extra turn to kill you or your allies (who you can't just Bear totem)...

It's definitely strong, but I'm pretty sure nerfing -5/+10 removes the whole dynamic of high-risk, high-reward that is kind of the point of the feat.

[smcmike]

Right, I guess I should have said "variance." Seemed pretty clear to me, so if I had an attitude about it it's because your responses sounded purposefully obtuse to me. But I guess it was just miscommunication.

And it seems obvious to me that variance can matter in principle, at least. The obvious example - a class feature that required a natural 20 to hit, but provided +1,000 damage would give an outstanding DPR, but would also be mostly worthless outside of fights against very high-hp enemies.

Now, this is an absurd case, and may not have much bearing on the discussion - it's meant only to show that variance can matter. It's more about perceptibility though. If one build outdamages another about a third of the time and gets outdamages about half of the time, that's not a huge gap, really.

[Xetheral]

Altering the base game is homebrew. Choosing a different scale for AC for your monsters changes the balance of the whole game. It would be something to mention, but it isn't a factor for a discussion based on the game's rules.

I could be mistaken, but I believe scmmike's point was that DM's choosing RAW enemies from the MM will not produce the same average AC at all tables as that listed in the DMG. If a given campaign happens to run into more offensively-oriented monsters at a given CR, the average AC at that table will be lower. If a given campaign happens to run into more defensively-oriented monsters at a given CR, the average AC at that table will be higher. Thus, average AC will vary from table to table even among DMs who play strictly by the book.


Nova vs sustained is a different topic than most of the others though.
For example it seems most people have a fairly clear idea how much stronger melee classes should be compared to ranged as they may not always be able to target opponents and they have to put themselves in vulnerable positions to do that damage. Most people seem to have an idea how much stronger a Fighter should be compared to a Rogue or a non spellcasting class like Ranger compares to a pure fighting like like Fighter.

I do agree that Nova wouldn't fall in that same category. Though if we make the assumption that groups play by the DMG recommended adventuring day (which seems the only logical starting point for discussing balance as that is what 5e is balanced on) then I don't think Nova is a huge concern. Even if the game designers did put in some ridiculous nova capabilities that is balanced out by that character being awful in every other encounter. Paladin for example would do really mediocre damage without resources.
Given the Adventuring Day assumption I'm not too concerned about nova, but maybe there is more to it. I'd be curious to hear what you think.

I probably should have said "spike" rather than "nova". I didn't mean to specifically refer to classes that expend resources, but rather to draw the distinction between classes whose damage comes from lots of low-damage attacks vs classes whose damage comes from fewer high-damage attacks. If there were two such classes whose damage output I'd expect to be similar, simply ensuring that the average damage they each do is identical wouldn't be enough for me to consider them balanced. I'd also want to make sure that the frequency and size of the spikes didn't interfere with the perception of balance among the players.


I'm glad we can disagree here. It is honestly refreshing to just discuss it directly. I do favor the math as I believe, as Saggo has pointed out, that the game is built on math. Based on the numbers I've run the developers recognize that as well as their game is fairly balanced in most regards.
I've honestly never understood the playtest or "feeling" oriented balance approach. To me that's secondary after the math to make the balanced mechanics feel fun. Though not everyone agrees.

I think the "table experience" aspect is important to the math because its the only way to verify if one's mathematical results hold true in practice. In other words, it provides an ability to check your work to make sure you didn't make a mathematical error or misinterpret your results. Calculations may be unambiguous, but interpreting the meaning of their results is not, particularly in the field of statistics with its dueling interpretive philosophies. If a game were to be mathematically perfect and yet still widely perceived as imbalanced by the players, who is mistaken? The designers? Or the Players? I'd argue that the players can't be wrong, as a matter of definition, but am I correct in thinking you disagree with me?


I got the imperceptible argument before. The argument that the dice play a big factor was missed.
Dice factor
Dice do indeed play a big factor in a bounded accuracy system. I don't think it's necessarily bad that two builds wouldn't seem that different. Two builds that focus pure DPR shouldn't be very perceptively different no matter how they achieve that DPR (via Polearm, GWM, or TWF barring other (some unmeasurable) considerations like reach, nova, spells, etc). By RAW you could say that many builds of similar nature are within 10% of eachother (with some larger variances).

In the end I'm left wondering why is the frequency relevant? In my eyes total damage dealt in a round/day/whatever is the metric to be concerned about as the dice will have their way. The game is balanced on the dice average, so why is average not a fair representative?

Answering your second question first, an average is only a fair representation to the extent that it reflects the underlying data. Depending on the distribution of that data, an average can be misleading. For an extreme example, consider lotteries: occasionally a lottery will go without a winner for long enough that the expected value (an average) of buying a ticket goes above 1. Following this average suggests that the optimal course of action is to buy as many tickets as possible, because each one is worth more than it costs to buy. In reality, the odds of winning are so vanishingly small that virtually everyone who buys tickets will lose money. In other words, the frequency of those for whom the average turned out to be misleading (the many losers) is vastly higher than the frequency of those for whom the average turned out to be a useful guide (the few winners).

For a more-relevant example, consider the same level 10 barbarians from before, but this time over only a single round. Here's the AnyDice representation: http://anydice.com/program/6e82. Click on the "graph" tab, and you'll see that for the barbarian without GWM, the damage is vaguely-normally distributed. Here the average of 32.91 is a mostly fair representation of what that barbarian can expect to do on a single round. However, for the barbarian with GWM, the distribution is crazy. I'd call it bimodal, but even that's a stretch. Here, the average of 38.15 is misleading, because such a barbarian will very rarely do about that much damage: he's far more likely to do either much more or much less.

Over time, of course, the average will start to matter more, which is why the per-combat and per-day charts I linked previously both feature more-normal distributions. (Yes, this counter-intuitively means that the per-round and per-day averages are of differing utility despite the one simply being a multiple of the other, because the underlying distributions are entirely different.) But what that means is that the average only reflects the underlying data over longer timescales. For an example of why this is problematic, if an adventuring day happens to span 2 or 3 sessions, that introduces the possibility that all of the sessions are imbalanced even if the adventuring day is balanced as a whole. Will the game still be fun if the individual sessions are imbalanced but the long-run average is balanced? That's hard to say, and will undoubtedly vary from table to table.

All of the above is doubly true when modeling rather than just looking at distributions of dice. Dice are nice because (assuming they're fair or at least differently-weighted) their results are uncorrelated. This means that if you roll enough dice you'll eventually get a normal-ish distribution, no matter how convoluted the rolling scheme is (for more on this, see here (https://en.wikipedia.org/wiki/Central_limit_theorem)). When modeling you run into the problem that your variables are far from random and might indeed be correlated. For an illustrative example, let's consider the variable in your DPR model for number of rounds of combat per day (useful because we've discussed its parameters before--I'm not trying to reopen the old debate). At any given table the number of rounds of combat per day will depend both on the number of combats and how long they are. So long as those two sub-variables are uncorrelated (or inversely correlated) the distribution underlying the average number of rounds of combat per day might be approximately-normal. But what if DMs who have fewer combats also tend to run shorter combats? In that case, the distribution underlying the average number of rounds of combat per day could easily be bimodal, with two clusters relatively far apart. If there are equal numbers of both types of DMs, the average could be located between the two modes, and thus describe almost none of them. That average would thus poorly describe the underlying distribution, creating a situation where your model might incorrectly describe the DPR behavior of certain classes at almost every table, despite accurately describing the average across all tables.

Note that I don't think this is likely! I personally suspect that few of the variables underlying your model will be correlated strongly enough across the entire playerbase to create a widespread problem. But just due to normal variation it's virtually guaranteed that at certain tables, even if they play entirely by RAW, that certain variables will coincidentally be correlated strongly enough that your model may over- or under-state class DPR for enough classes that it ultimately poorly describes class balance at those tables. It’s impossible to know how many of those tables there are, and it’s impossible to identify them in advance. That’s one of the chief reasons I’ve been such a proponent of your model’s application to abstract discussions, but persistently critical of the assumption that its numerical results can be said to apply at the individual table level.
Going back to your first question regarding frequency, it’s relevant because, as Tanarii pointed out, human perception is likely weighted heavily by the outliers. I’d go so far as to suggest that humans are pretty terrible at intuiting which of two semi-random series of numbers has a larger average. (I have no data, I’m just guessing based on personal experience.) The more-random the series, the worst I suspect humans are. The frequency with which the calculated average accurately describes the events at the table is thus highly relevant to the magnitude with which that average is perceived.

[Saggo]

The main point was never about DM-less play, you just held onto it for some reason. The point is that we all have a DM and the rules, there was nothing about playing with no DM and I believe you took my off comment at full value because I said the only way it can all be equal is that there would be no DM.
You spend an full paragraph (http://www.giantitp.com/forums/showsinglepost.php?p=20003957&postcount=137) establishing DMs exist at every table and make an incorrect accusation (http://www.giantitp.com/forums/showsinglepost.php?p=20004877&postcount=152) against me, both as opening remarks. So, that reason.

You and I seem to have different opinions on what makes an off comment.


That has nothing to do with the topic at all. I am only adding DM fiat as a way of saying that even when the dice are cast the DM can disregard it.
Well, that was an off comment, but "nothing, at all" is a bold statement that just isn't true.


Actually it proves that you are arguing with bad faith. You say that there has to be some way to discuss the game and I place a situation and you add variables to it just to prove a point that somehow math is always present even though you twisted my original premise into your own. The point is that just as there can be dice rolls they can also be circumvented by the DM and it isn't necessarily fiat. And I am not interested in playing Six Degrees of Mystery Keys so we can drop this little simulation.
Insulting my character wasn't necessary. That's also an incorrect accusation. "Most" and "likely" definitely don't imply always, that's 2 levels of uncertainty. Your situation was also nondescript, vague enough that simply adding a Perception check rendered it moot. I'm sorry if you think that means your premise was twisted. Plus, If the someone counters with "Surprise, it was their base all along.", probably shouldn't be surprised if someone else responds in kind.


Does it matter?

And you are proving my point.

If you want to argue in a controlled environment (and not change the environment on me) then do so but you can't somehow pretend that the math will hold up in all the tables, just from the mere reason that tables have different focuses and some tables might actually change the rules, sometimes drastically.
Yes it does and no I didn't, but honestly that's not up for you or me to decide. I don't appreciate the attacks, either.

I'm going to repeat what I said, since there wasn't any supporting argument provided against it. "If enough of a ruleset is ignored by DMs or players, we're no longer playing the same game and discussing "5th Edition" is rendered moot. Thus anyone can assume that another table is using a non-trivial or significant amount of the ruleset, or more importantly a non-trivial or significant amount of the math."

I'm going to add that it is the onus is on any DM or player to provide the context of their table since they chose to deviate from the established ruleset, even if deviation is encouraged. That means we can continue to assume only the baseline math of the ruleset is present until the context is provided. At which point the respectful thing to usually do is include the context in the discussion.

[JoeJ]

I could be mistaken, but I believe scmmike's point was that DM's choosing RAW enemies from the MM will not produce the same average AC at all tables as that listed in the DMG. If a given campaign happens to run into more offensively-oriented monsters at a given CR, the average AC at that table will be lower. If a given campaign happens to run into more defensively-oriented monsters at a given CR, the average AC at that table will be higher. Thus, average AC will vary from table to table even among DMs who play strictly by the book.

It's a bigger problem than that even. The table in the DMG lists AC as only one factor among several used to determine monster CR. It's not intended to work in reverse. IOW, we can't use the AC on the table to determine the average AC for monsters of that CR because we don't know how the balance between different factors works out for the creatures in the MM.

What can be done, if somebody has the time, is go through the MM and calculate the average AC at each CR.

[Coffee_Dragon]

Depending on the distribution of that data, an average can be misleading. For an extreme example, consider lotteries: occasionally a lottery will go without a winner for long enough that the expected value (an average) of buying a ticket goes above 1. Following this average suggests that the optimal course of action is to buy as many tickets as possible, because each one is worth more than it costs to buy. In reality, the odds of winning are so vanishingly small that virtually everyone who buys tickets will lose money. In other words, the frequency of those for whom the average turned out to be misleading (the many losers) is vastly higher than the frequency of those for whom the average turned out to be a useful guide (the few winners).

What. Let's say three people are given the option of buying a ticket for $1, and one of the three tickets will win $5. Are you saying each person should not buy the ticket offered, since they have a 2 in 3 chance of not being the winner?

[smcmike]

What. Let's say three people are given the option of buying a ticket for $1, and one of the three tickets will win $5. Are you saying each person should not buy the ticket offered, since they have a 2 in 3 chance of not being the winner?

That depends how much they need that dollar. For example, change that example to their life savings of $100,000 for a chance to receive half a million. Sounds a bit different, doesn't it?

Or do you go out and spend all your money on lotto tickets as soon as the jackpot exceeds the odds?

[Xetheral]

What. Let's say three people are given the option of buying a ticket for $1, and one of the three tickets will win $5. Are you saying each person should not buy the ticket offered, since they have a 2 in 3 chance of not being the winner?

I'm referring to situations such as Powerball, which has a positive expected value when the jackpot rises above $276,505,966. (This is a trivial calculation ignoring taxes, multiple winners, etc.) But the odds of winning the jackpot are still only 1 in 175,223,510. (The two numbers don't match up as expected due to the existence of intermediate prizes, but the positive expected value is due almost entirely to the chance of winning the jackpot.)

[Shaofoo]

You spend an full paragraph (http://www.giantitp.com/forums/showsinglepost.php?p=20003957&postcount=137) establishing DMs exist at every table and make an incorrect accusation (http://www.giantitp.com/forums/showsinglepost.php?p=20004877&postcount=152) against me, both as opening remarks. So, that reason.

You and I seem to have different opinions on what makes an off comment.


So then point to me where does it say that I am talking about DM-less gaming? You say I talk about DMs always being present and I made some comment about you, point to me where did I open the topic about DM-less gaming.



Insulting my character wasn't necessary. That's also an incorrect accusation. "Most" and "likely" definitely don't imply always, that's 2 levels of uncertainty. Your situation was also nondescript, vague enough that simply adding a Perception check rendered it moot. I'm sorry if you think that means your premise was twisted. Plus, If the someone counters with "Surprise, it was their base all along.", probably shouldn't be surprised if someone else responds in kind.

First of all it isn't an insult of your character, I mean maybe you can take it as an insult but I didn't mean any.

I wanted to prove that you can get around certain checks without having to roll the dice because the DM says so. And again you demonstrate your bad faith again by accusing me of "surprise" when you yourself has said "surprise" beforehand. I find it funny that you say that you shouldn't be surprised considering that you were the one to initiate it, and not only that but disregard my comment as well as an off hand remark.

The point was just the door, you wanted to add dice by making the key be found even though the key being found or not found is never in question. You wanted to prove that somewhere along the line a die roll had to be made.


Yes it does and no I didn't, but honestly that's not up for you or me to decide. I don't appreciate the attacks, either.

No attacks dude, none at all

I am making observations and you seem to be guilty of things you are accusing me of, that is no attack I am just saying what I'm seeing.


I'm going to repeat what I said, since there wasn't any supporting argument provided against it. "If enough of a ruleset is ignored by DMs or players, we're no longer playing the same game and discussing "5th Edition" is rendered moot. Thus anyone can assume that another table is using a non-trivial or significant amount of the ruleset, or more importantly a non-trivial or significant amount of the math."

You seem to be discussing about the actual discussion while I am discussing what you can take from the discussion. And like I said you proved my point. You can't expect conversations to hold in all tables because things can change.


I'm going to add that it is the onus is on any DM or player to provide the context of their table since they chose to deviate from the established ruleset, even if deviation is encouraged. That means we can continue to assume only the baseline math of the ruleset is present until the context is provided. At which point the respectful thing to usually do is include the context in the discussion.

Since there are as many potential DMs as there are people on earth it is impossible to establish any sort of consensus to a common DM unless we limit the discussion to just one table or just speak RAW.

Like I say again, you seem to want to talk about the actual discussion, I am talking about applying the discussion in real life. A lot of things that people used to think that looked bad on paper later were found out to not be so bad when applied in real life.

Besides I did try to provide context but you dismissed it.

[Kryx]

And it seems obvious to me that variance can matter in principle
Xetheral does a pretty good job below outlining what I believe would be your argument, but based on what he says and I present below that variance is gone when considered on an adventuring day.

So in the context of an adventuring day where it is evened out, does variance truly matter?




I could be mistaken, but I believe scmmike's point was that DM's choosing RAW enemies from the MM will not produce the same average AC at all tables as that listed in the DMG. If a given campaign happens to run into more offensively-oriented monsters at a given CR, the average AC at that table will be lower. If a given campaign happens to run into more defensively-oriented monsters at a given CR, the average AC at that table will be higher. Thus, average AC will vary from table to table even among DMs who play strictly by the book.
The actual AC faced can vary. Sometimes it'll vary low and sometimes high depending on the encounter. But in the end if the DM is picking without the purpose of higher AC it'll even out to the expected average value (assuming the devs followed their own guidelines - I'd love to see actual math on that).


If there were two such classes whose damage output I'd expect to be similar, simply ensuring that the average damage they each do is identical wouldn't be enough for me to consider them balanced. I'd also want to make sure that the frequency and size of the spikes didn't interfere with the perception of balance among the players.
That feels like a bit more than needed, but maybe it comes back to my mindset of "in the end the average is all that really matters" (more on this below).


I think the "table experience" aspect is important to the math because its the only way to verify if one's mathematical results hold true in practice.
If done in a scientific way, I would agree. But "table experience" is largely based on feeling. Often inaccurate feelings based on the scenario or perception imo. If playtesting was hundreds of scenarios for different characters and the numbers were recorded (ie: scientific process) then I'd be all for it, but as it has stood in the D&D community since it beginning it is simply "does it feel right? eeeeh close enough" which has often results in broken things - see monk, rogue, sorcerer of 3.X all being quite underpowered.

As you say, you can do some sample tests to see if your math was right, but that's verification, not balance.


If a game were to be mathematically perfect and yet still widely perceived as imbalanced by the players, who is mistaken? The designers? Or the Players? I'd argue that the players can't be wrong, as a matter of definition, but am I correct in thinking you disagree with me?
The designers would be wrong. "Perception is everything" is a famous quote and it does hold true that perception plays a huge part in all of life. But perception isn't what determines balance on a system level. The designer's job is to make the player's perception match the system's (balanced) math as closely as possible.


extreme example
Imo it's best to ignore extreme examples as they often spawn other discussions (as witnessed above). Focusing on how it impact's 5e's math is the important part.


For a more-relevant example, consider the same level 10 barbarians from before, but this time over only a single round. Here's the AnyDice representation: http://anydice.com/program/6e82. Click on the "graph" tab, and you'll see that for the barbarian without GWM, the damage is vaguely-normally distributed. Here the average of 32.91 is a mostly fair representation of what that barbarian can expect to do on a single round. However, for the barbarian with GWM, the distribution is crazy. I'd call it bimodal, but even that's a stretch. Here, the average of 38.15 is misleading, because such a barbarian will very rarely do about that much damage: he's far more likely to do either much more or much less.

Over time, of course, the average will start to matter more, which is why the per-combat and per-day charts I linked previously both feature more-normal distributions. (Yes, this counter-intuitively means that the per-round and per-day averages are of differing utility despite the one simply being a multiple of the other, because the underlying distributions are entirely different.) But what that means is that the average only reflects the underlying data over longer timescales
Ah, ha! I understand your argument now! Graphs, they speak to everyone! On a per round average you're right that a -5/+10 barbarian is unlikely to do near his DPR (he'll do more or less, but rarely the average) whereas without -5/+10 they are likely to do near his average DPR.
But as you point out the average reflects the underlying data over longer timescales. Since the adventuring day is all that we care that should be the metric we look at. Looking at the non-Vuman version I created before http://anydice.com/program/6e7 the average is indeed turned into a normal curve. So please excuse my ignorance, but what would be the danger of using the data now that it is a nice curve (besides the edge cases just below here)?


For an example of why this is problematic, if an adventuring day happens to span 2 or 3 sessions, that introduces the possibility that all of the sessions are imbalanced even if the adventuring day is balanced as a whole. Will the game still be fun if the individual sessions are imbalanced but the long-run average is balanced? That's hard to say, and will undoubtedly vary from table to table.
There will always be edge cases. I can tell from the mechanics, math, and the words in the DMG that the adventuring day is what the designers used to balance the game. That's my concern, not individual rounds or perception cases.


application of the above to your DPR model as a whole
Honestly, you should be the one creating the math and balance charts. You understand it far more than I do.

Quick note on the number of rounds: As discussed before, it will vary. I think the best we can do is estimate the average and run with it. If DMs play with different expectations they should prepare for the results (which is likely going to be casters and resource based classes being more powerful based on the polls I've seen). However there was an argument written on this topic that the party doesn't always have to follow the normal average rule, but just enough that the players will always hold a bit back to ensure they don't get stomped by the unseen encounter. So there is a bit of psychology involved as well.




It's a bigger problem than that even. The table in the DMG lists AC as only one factor among several used to determine monster CR. It's not intended to work in reverse. IOW, we can't use the AC on the table to determine the average AC for monsters of that CR because we don't know how the balance between different factors works out for the creatures in the MM.

What can be done, if somebody has the time, is go through the MM and calculate the average AC at each CR.
What is presented in the DMG is what the system expects at those CRs. Why can't it be used in reverse? Again it comes down to averages - some will be higher, some will be lower, some will be spot on.
The MM's actual data would be very useful to see how closely the designers matched their own system though.



@Saggo/Shaofoo: Your discussion seems to be on an unproductive path. Maybe let it go.

[Saggo]

So then point to me where does it say that I am talking about DM-less gaming?
I did. Twice, with evidence.


No attacks dude, none at all
"Bad faiths" and "twisting premises" and "pretends" are all that, if benign. Google ad hominem.


Like I say again, you seem to want to talk about the actual discussion,
I was very clear on that, and it should go without saying it's discussion with the intent of application.


I am talking about applying the discussion in real life.
If you don't see how "another table is using a non-trivial or significant amount of the ruleset, or more importantly a non-trivial or significant amount of the math." is applicable to real life, then this will just keep circling around. At this point, it should be self-evident why math is important to discussion and actual table play.


@Saggo/Shaofoo: Your discussion seems to be on an unproductive path. Maybe let it go.
Yeah, it got there. Sorry for muckin' up the thread!

[Strill]

Perceptibility is actually extremely important. The whole idea behind optimization, aside from generic "being better", it that your choices in character build have a perceptible, ideally positive, effect on gameplay.No, that's not the point of optimization at all. The point of optimization is to provide an ACTUAL effect on gameplay, regardless of whether it's perceptible.

I've seen tons of threads and posts complaining about how rogues are OP because they get to roll lots of damage dice, even though that doesn't actually make them any more powerful than other classes. That is a perceptual advantage effect that has no basis in reality.

If we were attempting to optimize perceptual efficacy, we'd just look for whichever class lets you roll the most dice, regardless of how much damage they actually do.

[JoeJ]

What is presented in the DMG is what the system expects at those CRs. Why can't it be used in reverse? Again it comes down to averages - some will be higher, some will be lower, some will be spot on.
The MM's actual data would be very useful to see how closely the designers matched their own system though.

What the DMG presents is a way of computing the CR of combinations of traits. There's no reason to believe that any individual trait will maintain the given value as an average, only that all of them together will. If, for example, a large number of creatures at a given CR have higher damage or special abilities, then the average AC for that CR would be lower than given on the table (assuming for the sake of argument that the CR in the MM matches the table at all).

[Coffee_Dragon]

That depends how much they need that dollar. For example, change that example to their life savings of $100,000 for a chance to receive half a million. Sounds a bit different, doesn't it?

Well, destitution would weigh rather heavily in the loss case!


I'm referring to situations such as Powerball, which has a positive expected value when the jackpot rises above $276,505,966. (This is a trivial calculation ignoring taxes, multiple winners, etc.) But the odds of winning the jackpot are still only 1 in 175,223,510. (The two numbers don't match up as expected due to the existence of intermediate prizes, but the positive expected value is due almost entirely to the chance of winning the jackpot.)

I have no personal experience with these types of lotteries, but a quick spot of googling indicates that if taxes and the prospect of shared jackpots are taken into account, there is no positive expected value, as you'd expect.

More on track, while there are probably good arguments against this whole "balance the DPRs" approach, I don't think questioning the use of statistical analysis is one of them. Granularity of everyday gaming just isn't relevant to an analysis of class-based maths.

[Tanarii]

BTW I looked it up and technically Variance (and Standard Deviation) are factors *within* a set of datapoints, not comparison between two sets. So they'd give you a value for how swingy a GWM Barbarian might be (given a set of potential hit/miss and damage rolls), but not the difference between, or probability of difference between, a GWM and a +2 ASI Barbarian. What we're interested in is the probability of variance within each set, and the difference/overlap between the two sets.

The way to compare is to give the expected range of DPR along with average. The way you determine that, IIRC, is Noise. That's SQRT(X)/X.


Xetheral does a pretty good job below outlining what I believe would be your argument, but based on what he says and I present below that variance is gone when considered on an adventuring day.

So in the context of an adventuring day where it is evened out, does variance truly matter?Potentially. But we're going to get into whether it is even possible to get a meaningful DPR-across-adventuring-day discussion or not again if we go into some of it. So I'll stay away from those aspects. ;)

But regardless, if there are 25 combat rounds per day (approx), then YES it potentially matters quite a lot.
The noise is:
1 attack: +/-20%
2 attacks: +/-14.1%
3 attacks: +/- 11.5%

I *think* that directly affects the hit chance, since the possibilities are evenly distributed. So a barbarian with two attacks (and a bonus attack) using GSM might expect to hit as little as 30% in an adventuring day, average 40%, and get lucky with 50% (all before advantage). Meanwhile the ASI could be anywhere from 55% - 75% with an average of 65%. (Pls note I'm using your assumptions here). That's a fair amount of variability within each set, which can allow for quite a large amount of overlap. (Note that modern polls, for example, are generally considered unreliable with noise outside of about +/-5%. Thats why they have thousands of data points, not less than 100.)

[Baptor]

Your other tactics are ways of mitigating the issue. I would prefer that the issue not exist in the first place instead of me designing encounters specifically the mitigate the issue. That seems silly to me.

Fudging rolls is a controversial topic. Even then.. using it to solve a math inequality is pretty ridiculous..

Okay, so we are a gadzillion pages further in than these comments but I wanted to bring them up. Bravo Kryx, bravo.

Having to create encounters tailored to dance around the OP character and fudging dice rolls should always be the last ditch exception, not the rule. If you have to throw in custom monsters into every combat to offset the super character or constantly lie about the OP character's attack rolls, then you have made a mistake and need to address it. I am tired of people saying this is a viable long term fix: it isn't.

Kryx your beef is -5/+10, mine are magic items with +hit and +AC. I think we are kindred spirits of a sort.

That said, I've run a game with both GWFs and Sharpshooters and its been OK. I am not saying Kryx is wrong, and he is probably right, but I pick my battles carefully when it comes to taking away things. For now the feat stays.

[smcmike]

Okay, so we are a gadzillion pages further in than these comments but I wanted to bring them up. Bravo Kryx, bravo.

Having to create encounters tailored to dance around the OP character and fudging dice rolls should always be the last ditch exception, not the rule. If you have to throw in custom monsters into every combat to offset the super character or constantly lie about the OP character's attack rolls, then you have made a mistake and need to address it. I am tired of people saying this is a viable long term fix: it isn't.

Kryx your beef is -5/+10, mine are magic items with +hit and +AC. I think we are kindred spirits of a sort.

That said, I've run a game with both GWFs and Sharpshooters and its been OK. I am not saying Kryx is wrong, and he is probably right, but I pick my battles carefully when it comes to taking away things. For now the feat stays.

Only one person is arguing that fudging die rolls is the solution. I think the rest of us can agree that that is a silly non-response.

The "custom monsters" is a bit of a straw man too. The argument is that the DM can choose standard monsters that happen to have higher ACs than what Kryx is testing against. This seems like a perfectly reasonable (and obvious) step to take if GWM is disruptive at your table.

[Mara]

I haven't seen math that puts GWM more than twice rogue DPR.

Yes that is effectively two characters worth of DPR because of both the strength of martials and the strength of the feat. I think it is a good way to compensate for lower caster DPR who fallback to cantrips after a few GOD wizard like spells. (I feel like good party DPR is rogue DPR * number of members)

Another big factor is not everyone wants to play a polearm wielder. I also think this problem only really applies to fighters. Barbarians can be very MAD if they want to toss their armor at some point to get 20 AC or more. Paladins are also very MAD with limited ASI. This feat combo is really for fighters. Battlemasters are very SAD and can afford to get 5-6 feats, the other kinds of fighters are looking at 2-3 feats. Low-int EKs are losing a lot. Not max con low dex champions are giving up lots of combat potential and utility.

Really, it's only battlemasters that can afford to grab both utility/BFC feats and DPR boosting feats. You run into the "problem" of battlemasters living up to their name.

[Baptor]

Only one person is arguing that fudging die rolls is the solution. I think the rest of us can agree that that is a silly non-response.

The "custom monsters" is a bit of a straw man too. The argument is that the DM can choose standard monsters that happen to have higher ACs than what Kryx is testing against. This seems like a perfectly reasonable (and obvious) step to take if GWM is disruptive at your table.

That's so funny. You think I was arguing with someone. I was just agreeing with Kryx, which for whatever reason, appears to have become really unpopular.

Also, misuse of straw man, just in case you ever are really arguing with someone. I didn't misrepresent anyone's argument, I just agreed with Kryx that I don't think tweaking monsters is a long term solution that I agree with. You might find that enjoyable for all I know. YMMV, and all that.

[Saggo]

What can be done, if somebody has the time, is go through the MM and calculate the average AC at each CR.


The MM's actual data would be very useful to see how closely the designers matched their own system though.

So this made me curious, and I really want to know what the average AC is. When I got home, made a pretty simple spreadsheet for MM's Avg AC by CR, using SUMIFs.

https://docs.google.com/spreadsheets/d/1-eul_N14RmAYuUjvJW4CsYvPmoCeks3LhbgivSM_n2o/edit?usp=sharing

I've gotten through a 3rd of the book, about 100 creatures. Even then you can see something of a pattern. I plan to finish it as I have time, limiting myself to just the MM. Let me know if the link's not working.

[Kane0]

Wow, makes the AC 26 boss fights we have with our level 12 party look really impressive.

Though to be fair, at the end of our last session the party legitimately decided to fight the ancient blue dragon starting next week because it was the least challenging thing to take on first.

[smcmike]

That's so funny. You think I was arguing with someone. I was just agreeing with Kryx, which for whatever reason, appears to have become really unpopular.

Also, misuse of straw man, just in case you ever are really arguing with someone. I didn't misrepresent anyone's argument, I just agreed with Kryx that I don't think tweaking monsters is a long term solution that I agree with. You might find that enjoyable for all I know. YMMV, and all that.

Agreeing with one side of an argument is arguing, though I prefer to think of what we are doing as a discussion - I happen to think Kryx has a lot of good points, and I certainly appreciate the work he's put in.

He also misinterpreted a number of points, though, including the suggestion that DMs just use higher AC adversaries, which he took to mean rewriting the MM. This was a straw man before you showed up, and it does become not a straw man simply because you happen to agree with it.

I don't mean to continue this meta discussion though - I agree with your points, such as they are. Fudging dice and rewriting the monster manual aren't good solutions.

[MaxWilson]

Battlemasters are very SAD and can afford to get 5-6 feats, the other kinds of fighters are looking at 2-3 feats. Low-int EKs are losing a lot.

IME, EKs use magic primarily for utility and defensive purposes like Find Familiar/Shield/Mage Armor/Expeditious Retreat/Blur. Int has zero effect on any of those. What do you see low-Int EKs as losing out on? Arcane Strike guarantees that they'll be effective against single targets even with mediocre Int (Int 12 or so).

To put it another way, what does an EK have to gain from pumping Int that is better than what he could get from a feat like Mobile, Lucky, Tough, or Alert? Why is it worth the opportunity cost?

[Saggo]

So this made me curious, and I really want to know what the average AC is. When I got home, made a pretty simple spreadsheet for MM's Avg AC by CR, using SUMIFs.

https://docs.google.com/spreadsheets/d/1-eul_N14RmAYuUjvJW4CsYvPmoCeks3LhbgivSM_n2o/edit?usp=sharing

I've gotten through a 3rd of the book, about 100 creatures. Even then you can see something of a pattern. I plan to finish it as I have time, limiting myself to just the MM. Let me know if the link's not working.

Finished the data entry, it was surprisingly quick since the book only has 1.3 mobs a page. I kept it to just the numbers for the public sheet, but I have all the names in a separate sheet if anyone cares. I can add some fancier stuff like min, max, median, outliers, etc, when I get more free time. I don't have the DMG directly in front of me, but the average AC in the MM is about equal or only 1 higher than what the DPR sheet uses.

[Mara]

IME, EKs use magic primarily for utility and defensive purposes like Find Familiar/Shield/Mage Armor/Expeditious Retreat/Blur. Int has zero effect on any of those. What do you see low-Int EKs as losing out on? Arcane Strike guarantees that they'll be effective against single targets even with mediocre Int (Int 12 or so).

To put it another way, what does an EK have to gain from pumping Int that is better than what he could get from a feat like Mobile, Lucky, Tough, or Alert? Why is it worth the opportunity cost?

Boom spells. Ranged attack cantrips. I also tend to run EKs with the sage background. Vhuman with ritual caster(wizard). Athletics and perception from fighter. Arcane and history from sage. Then nature from Vhuman. Warcaster too if I'm going sword and board duelist.

I see my EKs being the party brain and point person.

[Malifice]

Agreeing with one side of an argument is arguing, though I prefer to think of what we are doing as a discussion - I happen to think Kryx has a lot of good points, and I certainly appreciate the work he's put in.

He also misinterpreted a number of points, though, including the suggestion that DMs just use higher AC adversaries, which he took to mean rewriting the MM. This was a straw man before you showed up, and it does become not a straw man simply because you happen to agree with it.

I don't mean to continue this meta discussion though - I agree with your points, such as they are. Fudging dice and rewriting the monster manual aren't good solutions.

It's not a queston of fudging or rewriting the MM. The AC ranges used in Kryxs formula are those from the DMG and they're quite low.

For example CR5 encounters use AC14. High AC beasties are reasonably common at this level (Knights, Orogs, Animated Armor, Mages with shield spell, Drow, Hobgoblins etc). Against those critters using 5/10 is counter productive (and a wasted feat).

[MaxWilson]

Boom spells. Ranged attack cantrips. I also tend to run EKs with the sage background. Vhuman with ritual caster(wizard). Athletics and perception from fighter. Arcane and history from sage. Then nature from Vhuman. Warcaster too if I'm going sword and board duelist.

I see my EKs being the party brain and point person.

Hmmm. Boom spells = Thunderclap and Sword Burst? (Greenflame Blade keys off of your melee weapon attack instead of Int, so Dex/Str, which I guess is one of its benefits.) Yeah, that's one downside to skipping Int ASIs. I don't think it's a serious one since you're primarily a fighter, with ranged weapons attacks that are far better than a cantrip thanks to Sharpshooter/Archery style/Lucky feat/possible Crossbow Expert. But I see now why you'd enjoy pumping Int. Thanks for explaining.

[Mara]

Hmmm. Boom spells = Thunderclap and Sword Burst? (Greenflame Blade keys off of your melee weapon attack instead of Int, so Dex/Str, which I guess is one of its benefits.) Yeah, that's one downside to skipping Int ASIs. I don't think it's a serious one since you're primarily a fighter, with ranged weapons attacks that are far better than a cantrip thanks to Sharpshooter/Archery style/Lucky feat/possible Crossbow Expert. But I see now why you'd enjoy pumping Int. Thanks for explaining.I meant fireballs, thunderwave, firebolt, and dumping dex.

You don't have a lot of spell slots or spells but one damage spell per tier shouldn't be too bad.

For the most part it's the skills that make the int worth it. The sage background feature is also pretty important.

[Kryx]

The "custom monsters" is a bit of a straw man too. The argument is that the DM can choose standard monsters that happen to have higher ACs than what Kryx is testing against. This seems like a perfectly reasonable (and obvious) step to take if GWM is disruptive at your table.
If you're going to use a debate term then please use it correctly. A straw man is not what you're using it as.

A straw man is a common form of argument and is an informal fallacy based on giving the impression of refuting an opponent's argument, while actually refuting an argument which was not advanced by that opponent.
Custom monsters isn't a straw man - you and several others are arguing for that right now. I'm portraying your argument.

Increasing AC has the same effect as making all PCs bonus to hit 1 less. The only difference is you're doing it on the monster side instead of the PC side.
Specifically adjusting monsters to have higher AC or seeking out monsters with a higher AC is houserules. You're purposefully changing the game. If you don't seek it out then it'll come out toward the average AC.

Now lets take a look at how important that AC adjustment would be. We'll assume the variance in AC is 2 and it's equally as likely to be average-2, average, and average+2.

Level 11:
GWM Barb -5/+10 vs average-2: 43.5 DPR
GWM Barb -5/+10 vs average: 36.3 DPR
GWM Barb -5/+10 vs average+2: 27.9 DPR
Total: 107.7/3 = 35.9 DPR

GWM Barb vs average-2: 32.8 DPR
GWM Barb vs average: 30.5 DPR
GWM Barb vs average+2: 27.4 DPR
Total: 90.7/3 = 30.2 DPR

TWF Rogue vs average-2: 29.06 DPR
TWF Rogue vs average: 26.95 DPR
TWF Rogue vs average+2: 24.42 DPR
Total: 80.43/3 = 26.81 DPR

So when adding in variance to the calcs they are very close to the normal average. It only would matter if the GM specifically tries to raise the AC of monsters.


He also misinterpreted a number of points, though, including the suggestion that DMs just use higher AC adversaries, which he took to mean rewriting the MM. This was a straw man before you showed up, and it does become not a straw man simply because you happen to agree with it.
Still not a straw man. You even say yourself that rewriting the MM isn't a good solution, but you're suggesting doing effectively that.




I haven't seen math that puts GWM more than twice rogue DPR.
It's not double. It was nearly double against S&B, but that's because Barb S&B sucks. See http://www.giantitp.com/forums/showsinglepost.php?p=20001174&postcount=109 for the numbers on that.

http://i.imgur.com/StejwGS.jpg
Is how a Barbarian compares to a Rogue. Weighted average lowers the emphasis of level 1-2, and 19-20 (see the balance page to see the weight).
You'll notice the two options with -5/+10 are higher than Polearm. The crux of the argument here is that 25-30% is a good range. 45-50 is too much, but that's opinion based on the default game's DPR which some would disagree with.




When I got home, made a pretty simple spreadsheet for MM's Avg AC by CR, using SUMIFs. https://docs.google.com/spreadsheets/d/1-eul_N14RmAYuUjvJW4CsYvPmoCeks3LhbgivSM_n2o/edit?usp=sharing
...
Finished the data entry
Thanks for doing this! It'll help a fair amount. We can look at things like variance in AC as well.
Would you mind if I take this data and put it in my sheet? I'll want to add some color formatting to show where the DMG differs from actual MM monsters.




It's not a queston of fudging or rewriting the MM. The AC ranges used in Kryxs formula are those from the DMG and they're quite low.
Look at Saggo's sheet and compare with the DMG numbers. Or better yet here is a version where I quickly added DMG numbers to compare: https://docs.google.com/spreadsheets/d/15KB3fY-UrknHDIEqXDHbtdBg731V6iYi0U_K3PrpsTM/edit?usp=sharing
They are nearly all accurate with a handful within 1 AC of the DMG's numbers.

Of course there are certain monsters with higher AC, but there are also many with lower AC. It averages out. In practice it actually massively benefits the PC with -5/+10 as they wouldn't use it on high AC creature, only medium and low so the numbers I present would actually be higher than they are via the math.

[smcmike]

It's not a queston of fudging or rewriting the MM. The AC ranges used in Kryxs formula are those from the DMG and they're quite low.

For example CR5 encounters use AC14. High AC beasties are reasonably common at this level (Knights, Orogs, Animated Armor, Mages with shield spell, Drow, Hobgoblins etc). Against those critters using 5/10 is counter productive (and a wasted feat).

That was my point.

[smcmike]

If you're going to use a debate term then please use it correctly. A straw man is not what you're using it as.

Custom monsters isn't a straw man - you and several others are arguing for that right now. I'm portraying your argument.

Increasing AC has the same effect as making all PCs bonus to hit 1 less. The only difference is you're doing it on the monster side instead of the PC side.
Specifically adjusting monsters to have higher AC or seeking out monsters with a higher AC is houserules. You're purposefully changing the game. If you don't seek it out then it'll come out toward the average AC.

Now lets take a look at how important that AC adjustment would be. We'll assume the variance in AC is 2 and it's equally as likely to be average-2, average, and average+2.

Level 11:
GWM Barb -5/+10 vs average-2: 43.5 DPR
GWM Barb -5/+10 vs average: 36.3 DPR
GWM Barb -5/+10 vs average+2: 27.9 DPR
Total: 107.7/3 = 35.9 DPR

GWM Barb vs average-2: 32.8 DPR
GWM Barb vs average: 30.5 DPR
GWM Barb vs average+2: 27.4 DPR
Total: 90.7/3 = 30.2 DPR

TWF Rogue vs average-2: 29.06 DPR
TWF Rogue vs average: 26.95 DPR
TWF Rogue vs average+2: 24.42 DPR
Total: 80.43/3 = 26.81 DPR

So when adding in variance to the calcs they are very close to the normal average. It only would matter if the GM specifically tries to raise the AC of monsters.

Still not a straw man. You even say yourself that rewriting the MM isn't a good solution, but you're suggesting doing effectively that.

Look at Saggo's sheet and compare with the DMG numbers. Or better yet here is a version where I quickly added DMG numbers to compare: https://docs.google.com/spreadsheets/d/15KB3fY-UrknHDIEqXDHbtdBg731V6iYi0U_K3PrpsTM/edit?usp=sharing
They are nearly all accurate with a handful within 1 AC of the DMG's numbers.

Of course there are certain monsters with higher AC, but there are also many with lower AC. It averages out. In practice it actually massively benefits the PC with -5/+10 as they wouldn't use it on high AC creature, only medium and low so the numbers I present would actually be higher than they are via the math.

"Doing effectively that."

It's that middle word that gets you in trouble. I know what the term means, thanks. We disagree over whether whether you are fairly characterizing the argument, not over the definition of the straw man. You state that I suggest using custom monsters. I have not done so - I have suggested choosing standard monsters that will challenge your particular party. This sounds like standard DMing to me.

But this is a silly side argument. Saggo's list is a valuable addition.

[DanyBallon]

If you're going to use a debate term then please use it correctly. A straw man is not what you're using it as.

Custom monsters isn't a straw man - you and several others are arguing for that right now. I'm portraying your argument.

Increasing AC has the same effect as making all PCs bonus to hit 1 less. The only difference is you're doing it on the monster side instead of the PC side.
Specifically adjusting monsters to have higher AC or seeking out monsters with a higher AC is houserules. You're purposefully changing the game. If you don't seek it out then it'll come out toward the average AC.


I believe that what they meant, is that DM is free to use monster from the higher AC side of the curve for a given CR.
i.e. at 10th level the average challenge AC is 15, then there's monsters with higher AC, and the DM may be tempted to use these against a GWM character, and doing so may not affect other character that don't suffer the -5 penalty to hit.

Doing so is fully within RAW, but may end up with less monster diversity, and weird monster appearance...

You've done a great job calculating average DPR for different archetype, as well as proving that the variance over long term end up close to a bell curve. But in order to do so you must assume some parameter, as number of round per encounter, number of encounter per day and that every table will play a day worth of encounter each session.
But in reality, such perfect condition never or rarely applies. Also, in order to see the disparencies you need at least two character to compare, and as it have been said before, unless there are two or more charater build around doing max DPR, nobody will care if build A is doing 10% damage than B, because the other comparative (rogue, archer, mage, cleric) have other role in the party than being the best damage dealer. No one will blame a player for playing a S+B barbarian instead of a GWM+PAM barbarian, as long as he fill his role of doing more damage than anyone else.
Those are the main reason many that they don't have any problems with GWM. It's not that your maths are wrong, it's just that unless you specifically look them, the problem just don't exist at most table. And if so, the DM by raw as plenty of opportunity to work around. And even if he pump up moster AC, or HP, it's not much different that houseruling how GWM works, don't you think?

[Kryx]

You state that I suggest using custom monsters. I have not done so - I have suggested choosing standard monsters that will challenge your particular party. This sounds like standard DMing to me.
Again, this comes back to the 65% hit chance. If you prefer a lower normal hit chance to fix the issue then that's your choice, but you're still changing the base game which changes the balance.

I'd have to test the math to see what effect that choice would have on the game, but those kind of house specific rules won't effect the model for all games.




I believe that what they meant, is that DM is free to use monster from the higher AC side of the curve for a given CR.
..
Doing so is fully within RAW
Using monsters from both sides of the curve is totally understandable - I'm not arguing against doing so as that strategy will equate to average AC.

Purposefully adjusting monsters or monster choice is a choice that the GM makes. That's not the base that is expected. The point I'm making above is that the effect on the math is the same as adjusting player's to hit (which I would argue is bad for the game). I'm also making the point that specifically choosing higher AC creatures is not expected, but a specific choice of a DM (ie: houserules).


But in order to do so you must assume some parameter, as number of round per encounter, number of encounter per day and that every table will play a day worth of encounter each session.
This has been discussed to death in the past. This is mostly not true. The only thing I assume is round duration of combats. That's it. Please see my sheet where I take the DMG numbers from 82-85 to calculate the rest.

And I actually assume on the high end for my one assumption. In practice (based on the polls I've read) classes with resources would actually do better.


But in reality, such perfect condition never or rarely applies.
This has also been discussed to death. It doesn't matter for system balance. System balance is based on expectations. If groups do not play by those expectations then the system balanced around those expectations will no longer by balanced. That is WotC's design, not mine.
Groups can adjust the expectations on my sheet to see how their playstyle affects the balance. Though the game suffers a lot when that expectation isn't met. For instance classes like Warlock & monk become far less valuable when the group doesn't take short rests.


And even if he pump up moster AC, or HP, it's not much different that houseruling how GWM works, don't you think?
It is a houserule, I agree. Though as I pointed out above I do not believe it is good for the game or applies nearly as evenly as that GM would hope for.

[Markoff Chainey]

While I do totally agree with Kryx that -5/+10 is bad for the game, I also do not want to simply ban things... our own personal rule is that -5/+10 can be taken once per attack action only and must be called before any dice are rolled. This way it can be used 1 per attack action or 1 per attack and 1 per bonus action when it is an attack... (not when it is a reaction because it must be called for...) so the maximum of uses is 2 opposed to 7 times per turn.

I thought about banning it altogether but stumpled about this!
http://www.giantitp.com/forums/showthread.php?447801-The-Balance-implications-of-Bonus-Action-Feats-and-5-10-Feats&highlight=heatmap

If you scroll down, you find a heatmap by Bladeyeoman and it basically says that most of the time, -5/+10 is really cool if you are fighting stuff that you would win against anyways.. there are exceptions, of course and it still has implications to the game. It is just our approach to the problem to tone it down a little (see above).

[Finieous]

I think everyone is right. Kryx's numbers accurately model an average by-the-book adventuring day. The math is what it is. At the same time, the weighted averages in the balance comparisons are an important addition: We shouldn't assume that all levels are equal, because they aren't in actual play.

By the same totem token, in actual play not all fights are equal five-round bouts against average opponents. At most tables, there will be some fights that are more important and more difficult, against tougher opponents, than others. These opponents will often have higher than average AC, sometimes considerably higher. The result in actual play is that GWM/Sharpshooter will often be comparatively less effective in the most important battles against the toughest opponents, and comparatively stronger in less important battles against average and below-average opponents.

This doesn't affect the model of the average adventuring day one whit, but it does affect the play experience and the DM's task of challenging the players ("The barbarian will mow through the 'trash mobs' and really shine, but his GWM isn't going to help against the frost giant jarl with magic plate armor and shield. That's when the rogue will really shine, if he can set up a surprise round, or the wizard can change the game by forcing saving throws.")

[Kryx]

-5/+10 is really cool if you are fighting stuff that you would win against anyways.. there are exceptions, of course and it still has implications to the game.
I think the conclusion of that is a bit off. It assumes lower AC = lower challenge. That isn't necessarily true as a monster can have more hp, more damage, or spells that mitigate it. CR varies.

Though I think your limit to once a turn is fairly good. 1/turn seems more standard than 1/action (so twice w/ bonus)

[Kryx]

By the same totem token, in actual play not all fights are equal five-round bouts against average opponents. At most tables, there will be some fights that are more important and more difficult, against tougher opponents, than others. These opponents will often have higher than average AC, sometimes considerably higher. The result in actual play is that GWM/Sharpshooter will often be comparatively less effective in the most important battles against the toughest opponents, and comparatively stronger in less important battles against average and below-average opponents.
This exact arguement would the other side of the coin and a reason for why GWM is not a good feat - it sucks in important battles. Based on the level 11 I posted above a Barbarian is doing 4% more damage than a rogue with an AC of average+2 enemy. That is with and without -5/+10.

Based on everyone's argument here 4% isn't near enough.

That is the other reason I'd prefer a different system to -5/+10: It's insane when it works and it's god awful when it doesn't.

[Finieous]

This exact arguement would the other side of the coin and a reason for why GWM is not a good feat - it sucks in important battles. Based on the level 11 I posted above a Barbarian is doing 4% more damage than a rogue with an AC of average+2 enemy. That is with and without -5/+10.

Based on everyone's argument here 4% isn't near enough.

That is the other reason I'd prefer a different system to -5/+10: It's insane when it works and it's god awful when it doesn't.

I don't know, so is a fireball. Less provocatively, there's value in being able to quickly and efficiently dispatch the mooks, thereby preserving finite party resources for the more important battles. These kinds of abilities have a long tradition in (A)D&D.

[Z3ro]

but you're still changing the base game which changes the balance.


You keep saying this with regards to monster AC, but I'm very confused where this is coming from.

Now custom monsters, sure, that's a houserule (though even that's a stretch as rules for creating monsters are part of RAW). But near as I can tell, there's no RAW saying what AC a specific challenge should have.

[DanyBallon]

This exact arguement would the other side of the coin and a reason for why GWM is not a good feat - it sucks in important battles. Based on the level 11 I posted above a Barbarian is doing 4% more damage than a rogue with an AC of average+2 enemy. That is with and without -5/+10.

Based on everyone's argument here 4% isn't near enough.

That is the other reason I'd prefer a different system to -5/+10: It's insane when it works and it's god awful when it doesn't.

GWM's and Sharpshooter's -5/+10 are meant to be high risk=high reward, so it's okay that it won't be as effective against a above average AC opponent. In such case, you have the choice to either Munch trough the low AC mook that support him, and doing so letteing you allies to focus on him without any concern as you have them covered, or just don't use the -5/+10 for this fight. Either feat still allow you other usefull tricks to pull out.



This has also been discussed to death. It doesn't matter for system balance. System balance is based on expectations. If groups do not play by those expectations then the system balanced around those expectations will no longer by balanced. That is WotC's design, not mine.
Groups can adjust the expectations on my sheet to see how their playstyle affects the balance. Though the game suffers a lot when that expectation isn't met. For instance classes like Warlock & monk become far less valuable when the group doesn't take short rests.


Game balance is based on two things, mathematical balacance between races/classes/archetypes, and fun to play thoses races/classes/archetypes. Considering only the mathematical aspect because it's the only one quantifiable creates a distorsion in the expectations.
i.e. mechanically, Beastmaster ranger holds it's own in DPR, it's not the best, nor the worst, as expectated, yet, people agrees, that it need a fix. Why, because, while mechanically sound, it doesn't play very well, people don't like that the animal companion act as a robot.

You can do all the math you want (and Xetheral and you are sure really good at it), unless we find a way to quantify fun, we won't be able to rely only on math to achieve game balance. What your math does, it shows us some possible problematic aspect of the game, and gear us into paying caution when playing, to see if it really happens in most games.

[Saggo]

Thanks for doing this! It'll help a fair amount. We can look at things like variance in AC as well.
Would you mind if I take this data and put it in my sheet? I'll want to add some color formatting to show where the DMG differs from actual MM monsters.
You're welcome! I have to be honest, it wasn't hard.


Of course there are certain monsters with higher AC, but there are also many with lower AC. It averages out. In practice it actually massively benefits the PC with -5/+10 as they wouldn't use it on high AC creature, only medium and low so the numbers I present would actually be higher than they are via the math.
The biggest thing I want to do, besides find the Avg, is figure out how many mobs are statistical outliers. Basically, what is the chance of fighting something outside of Avg. Something to that effect will let you know how much weight to give the Avg that isn't just conjecture.

The MM is pretty bottom heavy, with the number of mobs in each CR dwindling fast after 11 or so. Which makes since official campaigns are geared for levels 1-15.

I also stole your DMG numbers.

[Tanarii]

Okay I crunched through some numbers. In a 25 round day, the variance in number of hits is pretty steady for reasonable hit chances. For a hit rate of 40% - 65%, the variance ranges from 5.6 to 6.25, or around 6. The standard distribution is therefore just under 2.5.

What this means: in a 25 round day, the swing in number of successes is plus or minus 2.5 successes (hits) away from the expected number based on hit chance in 68% of all cases. In other words, it's fairly solidly plus or minus 10%.

So kryx, you should be able to solidly address "random variation from the average" arguments by showing a low floor for DPR based on 10% less hit rate (before advantage), and a high ceiling based on 10% more hit rate. 68% of all adventuring days should fall in that range. That's given your parameters for a standard adventuring day of approximately 25 rounds.

Edit: that's for 1 attack per round. Plus or minus 10% nets you 68% accuracy for 1/round, and 99% accuracy for 3+ attacks per round.

[Saggo]

It's not a queston of fudging or rewriting the MM. The AC ranges used in Kryxs formula are those from the DMG and they're quite low.

For example CR5 encounters use AC14. High AC beasties are reasonably common at this level (Knights, Orogs, Animated Armor, Mages with shield spell, Drow, Hobgoblins etc). Against those critters using 5/10 is counter productive (and a wasted feat).

The median, average, and DMG value* for CR5 are all AC15. That median shows that abnormally low or abnormally high ACs aren't skewing the average. It also shows that while ACs above the average are indeed common like you said, ACs below the average are equally just as common. Accepting that specific campaigns usually target a subset of creatures in a CR5, an AC 15 is an accurate representation for all campaigns across all tables.

The median for each CR is within 1 AC of the average and DMG values*. That means that the DMG values are most likely accurate enough representations for all campaigns across all tables.

*As far as I've been told. I'm ordering my own copy now.

[smcmike]

Yeah, Saggo and Kryx have a good point there, particularly with the demonstration that the DMG data was followed in the MM, more or less. The number Kryx used are right, and his math will work for any DM who is randomly sampling from the MM by CR, which is the most objective way to test for balance (though obviously there is a deep divide in the level of balance people expect in DPR).

On the other hand, experiences at the table will vary considerably. In a way, it's like testing the power of poison or fire damage - (potentially) great when they work, terrible when they don't. You can count up the numbers for each type of resistance in the MM, but that doesnt necessarily give you any real clue about what you'll actually face at the table. With power attack, the question you'll want to answer at low levels is this - does my DM love goblins or orcs? Because power attack won't do much good in raiding that hobgoblin encampment.

[Icewraith]

The median, average, and DMG value* for CR5 are all AC15. That median shows that abnormally low or abnormally high ACs aren't skewing the average. It also shows that while ACs above the average are indeed common like you said, ACs below the average are equally just as common. Accepting that specific campaigns usually target a subset of creatures in a CR5, an AC 15 is an accurate representation for all campaigns across all tables.

The median for each CR is within 1 AC of the average and DMG values*. That means that the DMG values are most likely accurate enough representations for all campaigns across all tables.

*As far as I've been told. I'm ordering my own copy now.

The median equaling the mean doesn't mean what you think it means. The median value just means the monster in the middle of the set has the mean AC, but doesn't mean that the set isn't full of abnormally high or low ACs as long as they're evenly distributed. The set can look like this:

5,5,5,5,5,15,25,25,25,25,25 and still have the median, mean, and DMG values for AC be 15, and still have 15 not be a good number to base your model on. That's why you also worry about terms like the mode (most common value in the set), outlier values (min and max), and standard deviation. 15 can be the mean and median number in your dataset but still be the least likely number in the set you'll encounter.

Kryx, if you're looking for bells and whistles to add onto your spreadsheet, repeat the DPR calculations for additional ACs such as those - most common (if it differs from the mean), high, and low.

Also, add a "boss monster" section. See if the average AC of monsters with lair or legendary actions at each level (some levels may have 0 "boss" monsters) differs significantly from the DMG expected value at that level, and see what the expected DPR is.

I think what we're showing here is that there are regularly occurring extremely lethal situations where -5/+10 is deadweight. I think lowering the penalty/damage would actually make the feat worse from a power perspective, not better. Of course any feat like this you're going to want to use when you've got uncountered advantage. Lowering the penalty to something like -3/+6 increases the reliability of the feat in all other situations and is more likely to be treated as "always on". -5/+10 is big enough that when you're facing a serious threat, or need to use advantage just to break even, suddenly it becomes a much more interesting, risky choice. If you roll big you'll win big, but if you roll mediocre you'll probably miss.

[Saggo]

The median equaling the mean doesn't mean what you think it means. The median value just means the monster in the middle of the set has the mean AC, but doesn't mean that the set isn't full of abnormally high or low ACs as long as they're evenly distributed. The set can look like this:

5,5,5,5,5,15,25,25,25,25,25 and still have the median, mean, and DMG values for AC be 15, and still have 15 not be a good number to base your model on. That's why you also worry about terms like the mode (most common value in the set), outlier values (min and max), and standard deviation. 15 can be the mean and median number in your dataset but still be the least likely number in the set you'll encounter.

The median value is just AC, not a mean AC. It does accurately tell us how common above average or below average ACs are, if they're equal.

I did make an assumption, where we can reasonably assume there isn't a significant amount of abnormally high or low numbers by just looking at the data sets for each CR, which tells us we can infer a relationship between the median and average. I'm still adding analysis to the sheet, so we'll see how accurate that assumption is.

[Safety Sword]

Whilst all of this statistical analysis is fascinating, it doesn't change the fact that when -5/+10 matters it isn't broken. Please don't take this post as criticism of people who analyse the game mathematically. There is merit in that. I'm a scientist, so data is life. That being said however:

It doesn't really matter if you do 10 extra points of damage to things that have low AC and hit points.

It's still a choice to make if the AC is high. I would argue that low AC targets are dead meat in any case, so accelerating their trip to HP = 0 doesn't make any practical difference to the game. Those guys are fodder to eat actions.

Low AC, high HP targets are just that, a bag of HP to burn through, so the feat helps in that limited case. Great. Working as intended.

In the case where AC is high you are going to lose significant amounts of DPR to misses. If you need advantage to make the feat "work", then more power to you. You need to play a certain way to make that happen. A predictable, easily countered way.

Conclusion (for me): It's fine. I can DM it in and out as I need to so that it doesn't break my combats.

Edit: What I'm saying here is that (as with all aspects of this game) the DM has ultimate control on how much influence this feat has. System checks complete. Roll for initiative, monkey boy.

[Tanarii]

In the case where AC is high you are going to lose significant amounts of DPR to misses. If you need advantage to make the feat "work", then more power to you. You need to play a certain way to make that happen. A predictable, easily countered way.you seemed to have missed the part where GWM becomes a required feat for all Barbarians. And makes any Fighter that is a damager a Battle Master Great Weapon Tripper, as well well as requiring the feat.

Saying 'you can counter Advantage' is not a good counter to making a feat required for one class, and locking one into a specific build and maneuver.

Personally I don't care that Barbarians have a feat tax. But I *do* care that Fighters that want to do one of their two class roles (damage) are head and heels above better at it if they choose one build.

[broodax]

Here is the thing. The average AC for a given challenge rating is actually on the high side of all the ACs an adventurer should face. It is not just some trash monsters that have low AC bringing down the average. If you're fighting a monster with challenge rating equal to your level, that means it's an effective challenge all by itself. It is a "boss" monster. The vast majority of enemies should, individually, be lower CR than a character's level, because they are encountered in groups.

If your campaign has a high volume of huge CR creatures encountered alone, that may be just grand for you, but it is not the type of campaign the designers had in mind, and I'd guess it is not similar to most tables. The game was just made with enemies that are quite easy to hit.

[JoeJ]

Finished the data entry, it was surprisingly quick since the book only has 1.3 mobs a page. I kept it to just the numbers for the public sheet, but I have all the names in a separate sheet if anyone cares. I can add some fancier stuff like min, max, median, outliers, etc, when I get more free time. I don't have the DMG directly in front of me, but the average AC in the MM is about equal or only 1 higher than what the DPR sheet uses.

Excellent! Thank you. It's always better to have actual date to put into a model.

[smcmike]

Here is the thing. The average AC for a given challenge rating is actually on the high side of all the ACs an adventurer should face. It is not just some trash monsters that have low AC bringing down the average. If you're fighting a monster with challenge rating equal to your level, that means it's an effective challenge all by itself. It is a "boss" monster. The vast majority of enemies should, individually, be lower CR than a character's level, because they are encountered in groups.

If your campaign has a high volume of huge CR creatures encountered alone, that may be just grand for you, but it is not the type of campaign the designers had in mind, and I'd guess it is not similar to most tables. The game was just made with enemies that are quite easy to hit.

It's true that most enemies will be below your CR, but there are also a fair number of high-AC low-CR monsters. Goblins have AC 15. Hobgoblins have AC 18. In general, humanoids have AC on the higher end of the scale, because they often wear armor. In a lot of campaigns humanoids and NPCs make up the bread and butter, so, especially at low
levels, I don't think your assumption is really warranted.

Power attack is great at cutting up things that don't wear armor and aren't otherworldly fast. It's definitely what you want if you are fight ogres or zombies.

[Zman]

I didn't get a chance to read every page, but I can see how the -5/+10 skews the damage curve.

I think we are forgetting each Feat gives additional bonuses and the problem lies more with with the scale of the bonus.

How about changing -5/+10 to -Prof/+Prof. Now it scales, is a real trade off and forces tough decisions. Even if it isn't used often, there still are other usable and worthy benefits for the Feats.

If the math doesn't quite work -Prof/+1.5xProf could as well. Or a fixed -4/+6.

[Kryx]

@zman: all the options you listed are nearly the same result as the current situation.

[djreynolds]

Perhaps restrict it to certain levels, like it must be 11th to select it. Actually it seems that its at 11 level when things go bad isn't. Or toss is that you need a minimum in an additional stat, like con because its so tiring to perform. So introduce something that forces it to be used con modifier x2 per short or you get exhaustion. Whatever is appropriate.

Probably won't help, and its not a solution, just a work around like the magical armor and bad bards and clerics.

But if you tug too hard on this knot it might unravel the world economy

[Malifice]

Of course there are certain monsters with higher AC, but there are also many with lower AC.

Higher AC monsters are more common than lower AC monsters. Any Orc/ monstrous humanoid/ mook/ NPC in chain mail and a shield is spamming AC 18, and thats common from a very low level.


It averages out. In practice it actually massively benefits the PC with -5/+10 as they wouldn't use it on high AC creature

Not using a feat against half the monsters you encounter doesnt benefit the person with the feat. It benefits the guys that just bumped Strength by +2 or that took a different feat.

Have you played 5E? How often do the Barbarians or anyone else for that matter swing with GWM 'on'? Around 50 percent of the time from my experience.

Against high AC foes the -5/+10 gets put away, while the +2 to Strength is always on.

[Kryx]

Perhaps restrict it to certain levels, like it must be 11th to select it. Actually it seems that its at 11 level when things go bad isn't. Or toss is that you need a minimum in an additional stat, like con because its so tiring to perform. So introduce something that forces it to be used con modifier x2 per short or you get exhaustion.
I, personally, will just replace the mechanic with something like flat damage. It'll maintain the expected ratio to other builds without being OP against equivalent and lower AC and awful against higher AC.

[djreynolds]

Up the ante. Incorporate mercenary bands, and enemy types with similar PC abilities and higher armor classes, not for every fight mind you. But here and there.

GWM and sharpshooter are powerful, but so are spells. No reason not to have enemy sorcerers with the elemental adept feat and empowered spells. Twinned stoneskin for him and his meat shield.

Work arounds are better. And the players will love the challenges.

[Kryx]

Increasing AC will just make the rogue do as much damage as outlined above so I don't understand this argument when many of you above said that -5/+10 was crucial as doing big damage was critical to these classes (which I agree with).

[DanyBallon]

Increasing AC will just make the rogue do as much damage as outlined above so I don't understand this argument when many of you above said that -5/+10 was crucial as doing big damage was critical to these classes (which I agree with).

Doing big damage is not crucial but an iconic option for those classes. Varying your monsters so GWM shine from time to time and other style get their chance to shine as well is what we are talking about doing instead of trying to fix GWM. GWM is balanced in that when you take the feat you get a feature that gives you a high reward but also have a greater risk of failure. Still the feat offer other features available even if you don't use the -5/+10 in a round. The feat would be problematic if most of the monsters you face would be of low AC, but, while the average AC for a given CR is following the DMG, many iconic monsters can have high AC even in the low end of CR, i.e. Goblins.

Are we bad DM to throw goblins against our PC? I don't think so. But we would if we have a character with GWM and never let him face orcs or any other monster with a lower AC once in a while

[Zman]

@zman: all the options you listed are nearly the same result as the current situation.

I disagree. -Prof/+Prof will significantly lower damage output. At high level it is -6/+6 which not only reduces the damage dealt by 4 per attack, but it also reduces the likelihood of landing a hit.

Without a 2:1 ratio damage gets curbed hard. It become more situational for the player, used when they have advantage or the enemy AC is lowish. Your argument has been that -5/+10 is an extreme damage outlier, I happen to agree. This would limit the damage impact.

Sure, using PA become situational, but the feat also comes with Cleave which is effective in its own right.

I'd like to see the math, there is no way this results in the current situation, it inherently will lessen it. Even -4/+6 will lessen it, though it may be tricky with with accuracy coming into play.

[Kryx]

Ah, I misunderstood your suggestions. -6/+6 is just an awful feat as it will not change DPR for the advantage case (20 barb is 47 DPR with no -5/+10 and with -6/+6) and lower it for the normal case.

I did the math on my homebrew a while ago with -3/+6. The result was nearly the same. I've since removed it.

You're welcome to copy my sheet and test any math you want.

[smcmike]

Increasing AC will just make the rogue do as much damage as outlined above so I don't understand this argument when many of you above said that -5/+10 was crucial as doing big damage was critical to these classes (which I agree with).

If we want to talk about the aesthetics of the feat, it works perfectly. An iconic barbarian with a giant sword should be an absolute monster against slow unarmored monsters. When you want to hack an ogre to death or chop off the head of a giant snake-god, you call the big guy with the big sword, and he swings it as hard as he can. The elegant swordsman is less useful in this situation, since there is just so much flesh to poke.

On the other hand, hacking away sucks against serious armor, and is too clumsy to hit that elegant swordsman consistently.

As a mechanic, it is also way more interesting than a flat bonus to damage.

[Kryx]

If we want to talk about the aesthetics of the feat, it works perfectly.
In your opinion, it does. I would disagree using the arguments of "barbarians should do more damage than other options" presented earlier.

But that's my personal opinion as well. Some people obviously like the high risk, high reward. I don't like the math results and my players don't like the concept. So therefore I'll change it.

[smcmike]

In your opinion, it does. I would disagree using the arguments of "barbarians should do more damage than other options" presented earlier.

But that's my personal opinion as well. Some people obviously like the high risk, high reward. I don't like the math results and my players don't like the concept. So therefore I'll change it.

Sure, of course. When it comes to matters of tastes there isn't much reason to argue. I was just laying out my vision.

[Zman]

Ah, I misunderstood your suggestions. -6/+6 is just an awful feat as it will not change DPR for the advantage case (20 barb is 47 DPR with no -5/+10 and with -6/+6) and lower it for the normal case.

I did the math on my homebrew a while ago with -3/+6. The result was nearly the same. I've since removed it.

You're welcome to copy my sheet and test any math you want.

I figured that the -Prof/+Prof would be too sever, which is why I also offered the alternative of -4/+6. I have to borrow your spreadsheet when I get to a PC, but breaking the 2:1 ratio is what is important. -3/+6 is just as problematic and likely better than -5/+10 vs some ACs.

[Saggo]

Higher AC monsters are more common than lower AC monsters. Any Orc/ monstrous humanoid/ mook/ NPC in chain mail and a shield is spamming AC 18, and thats common from a very low level.


By Level and CR 4 (first ASI/Feat) and Avg AC 14, 25.2% of MM mobs are above average. At Level/CR 8, avg AC 16, it's 20.1%. Level/CR 12 is an odd case, since the AVG is still 16 (Avg AC 17 at CR 11), but it's still just 25.4%. All that ignores CR0.

We can assume most DMs stopped using low CRs by level 12. Somewhat arbitrarily (CR4 at level 12 is basically as challenging as CR1/8 at Level 4), let's limit to CR4-CR12, avg AC 16, where we do get 39.8%. Not a majority, but enough that most people would call it common. However, by level 12, most Martials are able to cap Strength and all can pick GWM, so the "always on" vs "not always on" is less of a concern.

Obviously, % of available above average mobs is not the same as % of encountered above average mobs. Given that the actual encountered rate is virtually unknowable and no one (here at least) has a significant sample size, what is "common" will be subject to intuition and cognitive/confirmation bias. I suspect most confirmation bias can be attributed to humanoids coming in repeating batches vs other creatures coming in random batches. Parsing official campaigns would be cool, though.

[Zman]

Been playing with your spreadsheet and based upon my initial impressions my -4/+6 suggestion seems to be a significant improvement. Thoughts??

[Malifice]

By Level and CR 4 (first ASI/Feat) and Avg AC 14, 25.2% of MM mobs are above average. At Level/CR 8, avg AC 16, it's 20.1%. Level/CR 12 is an odd case, since the AVG is still 16 (Avg AC 17 at CR 11), but it's still just 25.4%. All that ignores CR0.

We can assume most DMs stopped using low CRs by level 12. Somewhat arbitrarily (CR4 at level 12 is basically as challenging as CR1/8 at Level 4), let's limit to CR4-CR12, avg AC 16, where we do get 39.8%. Not a majority, but enough that most people would call it common. However, by level 12, most Martials are able to cap Strength and all can pick GWM, so the "always on" vs "not always on" is less of a concern.

Obviously, % of available above average mobs is not the same as % of encountered above average mobs. Given that the actual encountered rate is virtually unknowable and no one (here at least) has a significant sample size, what is "common" will be subject to intuition and cognitive/confirmation bias. I suspect most confirmation bias can be attributed to humanoids coming in repeating batches vs other creatures coming in random batches. Parsing official campaigns would be cool, though.

I disagree with your statement that low CR critters no longer turn up at 12th level.

12 x CR 2 Orogs are a 'hard' challenge for 5 12th level PCs. Mooks and low level critters are a threat at all levels in 5e. Seeing a BBEG with evil humanoid minions (veterans, Knights, orogs, hobgoblins, thugs etc) is quite common.

Which is why GWM is not going to be useful in every encounter against every critter. It's situationaly great (against a narrow range of ACs, in melee, for certain classes). Considering the opportunity cost for taking the feat (you could have just bumped your Attack stat by 2) it's perfectly balanced IMO.