GWM vs DW - Not as bad as you might suspect

[mgshamster]

I've been running some numbers for comprison between the Great Weapon Master feat and the Dual Wielder Feat. A lot of analysis I've read shows that GWM vastly outperforms DW and that the two are unbalanced. But there's a minor quibble I've had from all the analysis I've read comparing the two - they don't take the chance to hit into account. Back in 3.X, every DPR calculation I used took into account the chance to hit, but I haven't seen that used in 5e yet (maybe I just haven't see the right posts!).

To determine the DPR, I'm using the following equation:

h(d+b) + td

Where h is the chance to hit, represented as a percent

d is the average damage from the dice

b is the bonus damage from modifiers and other non-critting sources

t is the chance to crit

===================================

Looking at the Fighter, without an archetype but with fighting style, at 20th level, he should have

GWM: +6 to hit (+5 ability score, +6 proficiency, -5 feat) 2d6 damage (avg 8.3), +15 damage per hit (10 from GWM, 5 from ability modifier). This gives us 23.3 damage per attack on avg, plus 0.4 damage from crits. The GWM also gets a bonus attack on the crit - He has four chances to roll a crit, making the probability of rolling at least one 0.185, which puts his average damage for the bonus attack at 4.4 before adjusting for chance to hit. Total damage for four attacks plus bonus is 99.2

DW: +11 to hit, 1d8 damage (avg 4.5), +5 damage per hit. This gives us 9.5 damage per attack, plus 0.2 damage from crits. Total damage for five attacks is 48.5

That's a huge difference. And this is where I see most analyses stop. But what happens when we take the chance to hit into account?

If the DW needs a 10+ On the d20 to hit (55% chance to hit), then the GWM needs a 15+ to hit (30% chance). The average DPR is now 30.9 GWM and 27.1 DW.

If the DW needs an 11+ to hit (50% chance), then the GWM needs a 16+ to hit (25% chance). The average DPR is now 26.0 GWM and 24.8 DW.

But what about that +1 to AC for the DW? How much is that worth? To test it, I pitted them against each other. Now, if the DW needs an 11+ to hit (50% chance), the GWM needs a 17+ to hit (20% chance). The average DPR is 21.3 GWM and 24.9 DW. That +1 is worth a bit over 4.7 points of damage, which is also 5% of the GWM's damage (obviously) when DW and GWM are pitted against each other.

Somewhere around the 50% mark is where the two are balanced, where the GWM pulls ahead when it's easier to hit and the DW pulls ahead when it's harder to hit. Even though the GWM is 1.2 points ahead at the 50% mark, the +1 AC on the DW makes up for the difference.

When they're working together against some bad guy, it means the DW is taking 5% less damage than the GWM. All other stats equal, if they are up against a bad guy which gives the DW a chance to hit of 50% or worse, the DW will do more damage per round; if it's 55% for the DW or better, the GWM will do more damage per round. When they're matched evenly for damage, then the GWM will hit fewer times and hit harder, the DW will hit more often but not as strong.

For the purposes of this analysis, I've assumed all stats are the same, except that the DW has +1 AC due to the feat.

Percent chance to hit is based on the d20. Each side is 5%, so if you need to roll a 10 on the d20, then count up each roll that's a success. 10-20 is 11 sides, which is 55%.

This analysis is really about the TWF Style + DW vs GWF Style + GWM. I used a high level Fighter because that's where Two Weapon Fighting is least advantageous. The fewer attacks from Extra Attack, the better DW becomes.

Separate analysis has shown that not using the -5/+10 puts GWM consistently ahead of DW, but not by much (approximately 8 points per round) and the +1 AC makes up for about a quarter to a half that difference. The primary difference here is simply the greater damage of the weapon - it accounts for 6 of those points, the feat accounts for the other 2. Again, this is for a level 20 fighter - as the ratio between the number of attacks increase, dual wielding closer to great weapon fighting and eventually surpasses it. For example, if you only have one attack, GWM is at 9.1 while DW is at 9.4, before ability score modifiers and probabilities to hit are accounted for (remember, DW will have a better probability to hit than GWM by 25%).

[Rhaegar14]

What you're failing to account for is that even without the -5/+10, the Great Weapon Fighter does somewhere around 52 damage if all four of his attacks hit, without accounting for the bonus hit on a crit. So even WITHOUT the bonus damage his DPR is better, and he has the flexibility of sacrificing accuracy for damage when fighting something with low AC, or when he's been set up with Advantage. This is why there exist tables that show what to-hit numbers you should use GWM against for your damage-per-attack (though I'd be hard pressed to locate one right this second).

And that's also assuming he didn't take Polearm Master to also have 90% of the benefit of two-weapon fighting style as well. This is a not insignificant factor in what makes two-weapon fighting so bad; great weapon fighters can get its primary advantage without really sacrificing much of their own.

For any other class, you might have a point. But a max-level great weapon Fighter with 4 attacks deals more damage than Dual Wielder simply from getting 4/5 attacks with bigger weapon dice.

[EvilAnagram]

What you're failing to account for is that even without the -5/+10, the Great Weapon Fighter does somewhere around 52 damage if all four of his attacks hit, without accounting for the bonus hit on a crit. So even WITHOUT the bonus damage his DPR is better, and he has the flexibility of sacrificing accuracy for damage when fighting something with low AC, or when he's been set up with Advantage. This is why there exist tables that show what to-hit numbers you should use GWM against for your damage-per-attack (though I'd be hard pressed to locate one right this second).

And that's also assuming he didn't take Polearm Master to also have 90% of the benefit of two-weapon fighting style as well. This is a not insignificant factor in what makes two-weapon fighting so bad; great weapon fighters can get its primary advantage without really sacrificing much of their own.

For any other class, you might have a point. But a max-level great weapon Fighter with 4 attacks deals more damage than Dual Wielder simply from getting 4/5 attacks with bigger weapon dice.

I believe he specifically took the difference in weapon dice into account.

[mgshamster]

What you're failing to account for is that even without the -5/+10, the Great Weapon Fighter does somewhere around 52 damage if all four of his attacks hit, without accounting for the bonus hit on a crit. So even WITHOUT the bonus damage his DPR is better, and he has the flexibility of sacrificing accuracy for damage when fighting something with low AC, or when he's been set up with Advantage. This is why there exist tables that show what to-hit numbers you should use GWM against for your damage-per-attack (though I'd be hard pressed to locate one right this second).

For any other class, you might have a point. But a max-level Fighter with 4 attacks deals more damage than Dual Wielder simply from getting 4/5 attacks with bigger weapon dice.

I did not fail to account for that - it's in my spoiler notes. DW is dealing about 47.5 damage for its five attacks, which is within a handful of points of GWM.

This means that they're relatively comparable - at the end game they're only a few points part from each other. Which is not nearly as bad as you're making it out to be.

In early game, DW is ahead of GWM. Somewhere in mid game they cross, and they're never that far apart. And that's if you don't use the -5/+10. If you do, then they're balanced right around the 50% chance to hit mark.

[Ruslan]

Any factors that help the to-hit chance skew the math significantly in favor of GWM.

Precision Attack (Battlemaster maneuver)? It helps the GWF fighter a lot. Sure, it also helps the DW fighter, but not as much.
Reckless Attack (Barbarian dip)? Again, helps GWF more than DW.
Got a Bardic Inspiration die up your sleeve? Same.
Got a Wolf Totem Ally or a Faerie Fire caster? Same.
+1 weapon? Again, the enhanced to-hit chance helps GWF more than it helps DW (and DW will need two weapons anyway)

And the higher you advance in levels, the more common sources of advantage become. Which is why GWF will pull ahead with levels.
Also, as was mentioned, the GWF fighter doesn't have to take -5/+10. Against the occasional high AC opponent, he can fight normally and be at least on par with DW.

[Rysto]

I did a similar analysis for a Paladin recently. When Improved Divine Strike kicks in at Paladin 11, GWM isn't very good anymore. A high-level paladin with GWM is almost never better off taking the accuracy hit. It's only worth positive DPR if the paladin has advantage on the attack roll and hits on 5 or better, or if the Paladin needs to crit to hit anyway.

Meanwhile, Improve Divine Smite synergizes nicely with TWF. The trick is that a Paladin can't get that fighting style natively, so they would have to do a 1 level Fighter dip to pick it up (I suspect that the synergy is why TWF isn't directly available to Paladins). A high-level Paladin with dual wielder but not TWF only slightly outperforms a GWF/GWM Paladin.

[Iguanodon]

Any factors that help the to-hit chance skew the math significantly in favor of GWM.

And any factors that increase damage tip the balance towards DW. Notably, these are most significant at low levels, rather than high: Hex, Hunter's Mark, that one Cleric spell I can never remember. Sneak Attack and the later Hunter abilities are absent from my list since they're just once per round.

My thinking is GW is best for pure martial classes, still, but if you have spells like Hex, it's probably better to go DW if you can (holding a focus might be a relevant factor, though).

[Finieous]

Meanwhile, Improve Divine Smite synergizes nicely with TWF. The trick is that a Paladin can't get that fighting style natively, so they would have to do a 1 level Fighter dip to pick it up (I suspect that the synergy is why TWF isn't directly available to Paladins).

IDS synergizes nicely with PAM, too, for the same reasons, and your paladin doesn't have to take a fighter dip for TWFing. ;)

[mgshamster]

Any factors that help the to-hit chance skew the math significantly in favor of GWM.

It does help. GWM does better when it's easier to hit. DW does better when it's harder to hit. Once you get to 95% chance to hit, GWM is dealing about twice as much as DW.

My point, however, is that it's not always dealing that much more, and they are fairly balanced (right around the 50% chance to hit mark). Once you pass that 50% mark, DW is the better performer (especially once you account for the extra AC).



Precision Attack (Battlemaster maneuver)? It helps the GWF fighter a lot. Sure, it also helps the DW fighter, but not as much.
Reckless Attack (Barbarian dip)? Again, helps GWF more than DW.
Got a Bardic Inspiration die up your sleeve? Same.
Got a Wolf Totem Ally or a Faerie Fire caster? Same.
+1 weapon? Again, the enhanced to-hit chance helps GWF more than it helps DW (and DW will need two weapons anyway)

And the higher you advance in levels, the more common sources of advantage become. Which is why GWF will pull ahead with levels.
Also, as was mentioned, the GWF fighter doesn't have to take -5/+10. Against the occasional high AC opponent, he can fight normally and be at least on par with DW.

Pretty much every time I see someone talk about wanting to play a DW character, the general recommendation is to avoid it because it's that much worse than GWM. My point is that it's not that much worse than GWM. Without the -5/+10, they're about on par, with DW slightly better at lower levels and GWM slightly better at higher levels. With it, GWM is better when it's easier to hit and DW is better when it's harder to hit.

Which means that when someone wants to play a DW character, we shouldn't be telling them it's going to suck. They'll be about on par. There will be times when GWM will out perform DW, and there will be times when DW out performs GWM. For the most part, GWM is slightly better - but not so much better that it's a blanket recommendation to avoid DW like we've been doing here on the playground.

Also, it should be noted that my analysis do not include the Attack Bonus, but rather the number required to roll on the d20 - this bypasses whatever attack bonuses may or may not be present - the attack bonus is irrelevant when you're looking at the number on the d20 that's needed to roll.

[Rysto]

IDS synergizes nicely with PAM, too, for the same reasons, and your paladin doesn't have to take a fighter dip for TWFing. ;)

Oh, I'm well aware (IMO it's too good on a Paladin), but I didn't want to derail the thread.

[GlenSmash!]

Nice analysis. Now I just need a Monkey grip feat and I'll take both DW and GWM. :smallbiggrin:

[Ruslan]

And any factors that increase damage tip the balance towards DW. Notably, these are most significant at low levels, rather than high: Hex, Hunter's Mark, that one Cleric spell I can never remember. Sneak Attack and the later Hunter abilities are absent from my list since they're just once per round.

My thinking is GW is best for pure martial classes, still, but if you have spells like Hex, it's probably better to go DW if you can (holding a focus might be a relevant factor, though).Alas, both Hex and Hunter's Mark take up a Bonus Action themselves, and thus not a perfect synergy with DW. You have to skip the offhand attack on the round you cast the spell, and every time you down an enemy you have to skip the offhand attack again to move the Hex or Hunter's Mark to a new target.

[mgshamster]

So let's look at PAM in combination with GWM. Even though we're now comparing two feats to one feat.

All PAM weapons are dealing 1d10, not 2d6, which reduced the average weapon damage from 8.3 to 6.3. It also removes the bonus attack on the crit, removing the 4.4 damage. But it adds a 1d4+5 or +15. So that's 17.63 total.

With 2d6, we're at 99.2 damage before probabilities. With PAM, it's 104.1. An extra 5 damage more. Without the -5/+10, it's 53.2 vs 54.1. A little better. PAM/GWM combo is now just more than a handful above DW when not using the -5/+10. The difference between DW and PAM/GWM is now 6.6 points per round.

When using -5/+10, the balance point is when you need to roll a 12 for DW, so PAM adjusts the balance point by 5%.

[krugaan]

that's some quality work there, mghamster. Now get in the wish/simulacrum and minor illusion threads and do your magic!

edit: I always figured they were roughly balanced, although I do think the PAM benefiting from GWM is slightly overdone.

[Finieous]

It also removes the bonus attack on the crit, removing the 4.4 damage.

Not a huge difference, but if you crit or drop an opponent with your Attack action, you can still use the bonus action for another 1d10 attack, rather than using it for the 1d4 attack, right?

[mgshamster]

Not a huge difference, but if you crit or drop an opponent with your Attack action, you can still use the bonus action for another 1d10 attack, rather than using it for the 1d4 attack, right?

Yes. Let's do a weighted average:

(0.18*21.8)=3.83
(0.82*17.63)=14.46

Total: 18.28 instead of 17.63.

Which makes GWM/PAM's total 104.7 instead of 104.1. A bit better. Not enough to push it over the "12+ on a d20" balance point, but it does change the difference with DW (no -5/+10) from 6.6 to 7.2.

[Tanarii]

Now do all the math with Advantage.

Really GWM comes down to two things:
1) Can your character get advantage regularly (Barbarian, Battle Master)? If yes, GWM is good.
2) Do you have a high base damage (Paladin, Bladelock), or extra damage from attacks (Ranger, Warlock)? If yes, GWM is bad.

[Roughishguy86]

I would just like to say thank you for this post.

[bid]

When they're working together against some bad guy, it means the DW is taking 5% less damage than the GWM.
Actually, it's closer to 10%. Another bonus for DW.
{9/20 vs 10/20 == 9/10 == 90%, assuming you get hit half the time}


PAM by itself is 12% better though:
- 53.2 per hit (+28.2 per crit) {4x (1d10+4/5+5) + (1d4+1/2+5)}
- 47.5 per hit (+22.5 per crit) {5x (1d8+5)}

GWM is made to clear through low-AC minions, call it a weak form of AoE.

[Foxhound438]

Any factors that help the to-hit chance skew the math significantly in favor of GWM.

Precision Attack (Battlemaster maneuver)? It helps the GWF fighter a lot. Sure, it also helps the DW fighter, but not as much.
Reckless Attack (Barbarian dip)? Again, helps GWF more than DW.
Got a Bardic Inspiration die up your sleeve? Same.
Got a Wolf Totem Ally or a Faerie Fire caster? Same.
+1 weapon? Again, the enhanced to-hit chance helps GWF more than it helps DW (and DW will need two weapons anyway)

And the higher you advance in levels, the more common sources of advantage become. Which is why GWF will pull ahead with levels.
Also, as was mentioned, the GWF fighter doesn't have to take -5/+10. Against the occasional high AC opponent, he can fight normally and be at least on par with DW.

Pretty much this. Dual wielding, especially for a pure fighter, simply can't get as much out of your resources. Granted, things like hex and such will help the dual wielder, but not by as much.

[Foxhound438]

Now do all the math with Advantage.

Really GWM comes down to two things:
1) Can your character get advantage regularly (Barbarian, Battle Master)? If yes, GWM is good.
2) Do you have a high base damage (Paladin, Bladelock), or extra damage from attacks (Ranger, Warlock)? If yes, GWM is bad.

exactly this.

[Gtdead]

Using the basic solo fighter playstyle to make these kind of comparisons will always give funny results.
Like I said in another thread about fighter dpr, don't count 4 attacks, trade one for shove and count 3 attacks with advantage. And then it all makes sense.

[Saggo]

In an isolated round, Dual Wield holds up, like you've shown. But in the course of a full adventuring day, it will continue to do less and less DPR vs GWM simply because (and my key complaint with TWF and DW by extension) it doesn't play nicely with other class features like GWM.

Any advantage, regardless of consistency, favors GWM. Action Surge favors it, and does nothing for TWF/DW. Hunter's Mark and Hex look like they favor TWF/DW, until you realize that you need to make 3 additional bonus action attacks to compensate for casting or moving them (3d6 to beat 1d8+4-5). Hunter's Whirlwind favors Duelist, Beastmaster is right out, and most of Rangers AoE spells are ranged specific. The list goes on.

Minor factors like these are why a Battlemaster can end up with a daily DPR of 62 with GWM but only 48 with TWF/DW.like at level 20 as shown on Kryx's DPR spreadsheet.

[mgshamster]

Now do all the math with Advantage.

I figured someone would ask, but I still didn't want to do it. I figured anyone should be able to use the formula I provided to do it themselves if they wanted to prove that it worked.

Anyways, let's look at a hit on a 11+ for DW (which is 16+ for GWM):

GWM: A hit on a 16+ is 25% chance to hit, or 44% with advantage, which is 43.4 DPR. With 17+, it's 35.4. With 18+, it's 27.7

DW: A hit on a 11+ is 50% chance to hit, or 75% chance with advantage. Which is 35.9 DPR. With 12+, it's 32.9. With 13+, it's 30.6.

So the balance point shifts from 11+ (50% mark) to 13+ (40% mark) with advantage.

[Gtdead]

I figured someone would ask, but I still didn't want to do it. I figured anyone should be able to use the formula I provided to do it themselves if they wanted to prove that it worked.

Anyways, let's look at a hit on a 11+ for DW (which is 16+ for GWM):

GWM: A hit on a 16+ is 25% chance to hit, or 44% with advantage, which is 43.4 DPR. With 17+, it's 35.4. With 18+, it's 27.7

DW: A hit on a 11+ is 50% chance to hit, or 75% chance with advantage. Which is 35.9 DPR. With 12+, it's 32.9. With 13+, it's 30.6.

So the balance point shifts from 11+ (50% mark) to 13+ (40% mark) with advantage.

Which shows that a +1 to hit for DW is 2.3 dpr increase, and for GWM is 7.7.
Want to add a bless too? I mean we are already in the CR24+ AC range here. That's the endgame of endgame.

Also consider what happens when a GWM fighter blows his precision strikes on action surge, and how DW compares.

[mgshamster]

Which shows that a +1 to hit for DW is 2.3 dpr increase, and for GWM is 7.7.
Want to add a bless too? I mean we are already in the CR24+ AC range here. That's the endgame of endgame.

Also consider what happens when a GWM fighter blows his precision strikes on action surge, and how DW compares.

It's not quite linear (damage added from crits causes the linearity to skew); but yeah, the change can be approximated as such for advantage.

Bless is irrelevant. All it does is give an attack modifier bonus, but we're bypassing attack modifiers by looking at the roll required to hit. It doesn't matter what your bonus is if you need to roll an 18 to hit - that's the target number on the d20, which determines the percent chance to hit, which determines how the damage is modified for a DPR calculation.

Interestingly, the advantage analysis also shows that if you need to debuff an enemy, target the GWM enemy with anything that causes disadvantage or a penalty to hit rather than something that does a flat reduces damage - it's more effective in the damage reduction. Not sure if there are debuff spells which actually reduce damage, but still.

[Gtdead]

If the enemy AC is ever increasing, yes it's irrelevant. But it's not. There is a hard cap of 25. We can't optimize for infinity cause then the numbers stop making sense. For example against monsters with infinite AC, dw deals 3.5 dpr and rogue deals 4.2. Does this mean that rogue compares favorably to fighter? Hardly.

[mgshamster]

In an isolated round, Dual Wield holds up, like you've shown. But in the course of a full adventuring day, it will continue to do less and less DPR vs GWM simply because (and my key complaint with TWF and DW by extension) it doesn't play nicely with other class features like GWM.

Any advantage, regardless of consistency, favors GWM. Action Surge favors it, and does nothing for TWF/DW. Hunter's Mark and Hex look like they favor TWF/DW, until you realize that you need to make 3 additional bonus action attacks to compensate for casting or moving them (3d6 to beat 1d8+4-5). Hunter's Whirlwind favors Duelist, Beastmaster is right out, and most of Rangers AoE spells are ranged specific. The list goes on.

Minor factors like these are why a Battlemaster can end up with a daily DPR of 62 with GWM but only 48 with TWF/DW.like at level 20 as shown on Kryx's DPR spreadsheet.

It depends on the chance to hit - that's been my entire point all along. If you're facing a chance to hit of 40% or less (13+ on the d20), with advantage, the DW will do more DPR than the GWM (-5/+10). That's an AC of 24+ . Not entirely unreasonable for a high level game. High AC opponents favor the DW. The GWM does better with lower AC opponents. At that point, the GWM has to turn off their -5/+10, which makes their DPR approximately the same, only differing by a handfull of points. Which means that the discrepancy is not nearly as bad as everyone seems to believe.

And remember, high level games are where you see the greatest discrepancy between DW and GWM. As you go down in level, that discrepancy diminishes and eventually disappears somewhere in the mid level range. At low level, DW is favored.

Where is the majority of the game played? Between 1-15th level; so the discrepancy between them is much much lower than most people claim uses Kryx's and other's analysis (which don't take into account the chance to hit) - to the point where there is no discrepancy at all in the majority of games.

[mgshamster]

If the enemy AC is ever increasing, yes it's irrelevant. But it's not. There is a hard cap of 25. We can't optimize for infinity cause then the numbers stop making sense. For example against monsters with infinite AC, dw deals 3.5 dpr and rogue deals 4.2. Does this mean that rogue compares favorably to fighter? Hardly.

Really? I've never seen a hard cap. In fact, I've seen optimized posts getting ACs into the 30s and 40s.

Care to cite a rule for the hard cap on AC?

[Finieous]

Where is the majority of the game played? Between 1-15th level; so the discrepancy between them is much much lower than most people claim uses Kryx's and other's analysis (which don't take into account the chance to hit) - to the point where there is no discrepancy at all in the majority of games.

Kryx's analysis certainly does account for chance to hit.

[mgshamster]

Kryx's analysis certainly does account for chance to hit.

Huh. I must have misremembered. Fair enough!

[Gtdead]

I always assume that we are talking about officially released content. Tarrasque peaks at AC 25. Worst case scenario some madman buffs it for a +2.
Still, against infinite AC, GWM fighter has a 5% to hit for an average of 4d6+15 dpr=1.45 dpr , DW has a 5% to hit for an average of 2d8+5 = 0.7 dpr

4 attacks x 1.45 dpr = 5.8
5 attacks x 0.7 = 3.5

GWM deals 65% more damage than DW. As I said, numbers stop making sense when your example shows DW becoming better against high ACs.
And why infinite AC? Why not? Using the most common party setups, we have already approached infinity by calculating against 40 AC.

[NewDM]

Any factors that help the to-hit chance skew the math significantly in favor of GWM.

Precision Attack (Battlemaster maneuver)? It helps the GWF fighter a lot. Sure, it also helps the DW fighter, but not as much.
Reckless Attack (Barbarian dip)? Again, helps GWF more than DW.
Got a Bardic Inspiration die up your sleeve? Same.
Got a Wolf Totem Ally or a Faerie Fire caster? Same.
+1 weapon? Again, the enhanced to-hit chance helps GWF more than it helps DW (and DW will need two weapons anyway)

And the higher you advance in levels, the more common sources of advantage become. Which is why GWF will pull ahead with levels.
Also, as was mentioned, the GWF fighter doesn't have to take -5/+10. Against the occasional high AC opponent, he can fight normally and be at least on par with DW.

All this math is a nice exercise, but in reality most attacks are going to fall into the 60%-80% hit range because of the most common ACs at any given level, barring outliers.

Most common ACs by CR and Fighter attack bonus by level:
level/CR = AC; Fighter Attack bonus assuming +4 ability score (first feat +5) + Proficiency bonus; % chance to hit
1 = 11; +6; %80
2 = 13; +6; %70
3 = 13; +6; %70
4 = 12; +7; %80
5 = 15; +8; %70
6 = 14; +8; %75
7 = 15; +8; %70
8 = 15; +8; %70
9 = 18; +9; %60
10=18; +9; %60
11=17; +9; %65
12=18; +9; %60
13=18; +10; %65
14=18; +10; %65
15=19; +10; %60
16=19; +11; %65
17=19; +11; %65
18=19; +11; %65
19=20; +11; %60
20=19; +11; %65

Feel free to check the math and the numbers yourselves. As far as I know there are not situations where you would get a direct bonus to attack more than half the rounds in a day (even Battle Master dice only last 18 out of 35 average rounds in a day). If you have disadvantage, that's the only time you'll see a percent chance to hit under 50% or some weird outlier like Animated Armor at CR 1 with an 18 AC.

[Saggo]

It depends on the chance to hit - that's been my entire point all along.

In an isolated round, yes. I'm saying if you expand to an entire day it's to-hit and factors that manipulate to-hit and bonus action use. More factors favor GWM and discourage TWF, which means by extension DW.

More DPR is always good, that's why DW is serviceable, but the value added for DW is less than GWM.

[Mith]

Having not seen TWF in action, would a fix to TWF being that you gain one extra pseudo bonus action that allows you to do an offhand attack, basically freeing up the bonus action to be used elsewhere, keeping everything else the same? Or does that break everything to pieces?

[mgshamster]

In an isolated round, yes. I'm saying if you expand to an entire day it's to-hit and factors that manipulate to-hit and bonus action use. More factors favor GWM and discourage TWF, which means by extension DW.

More DPR is always good, that's why DW is serviceable, but the value added for DW is less than GWM.

That's fair - I did address the point of the day by making the claim that t isn't unreasonable to have high AC opponents at high levels - where GWM favors lower AC. To be clear, I do not want to discount your point, you make a very fair point.

Your earlier point about there not being enough to supplement DW is also a consideration, and anyone who wants to fix the issue should be aware that the difference between the two isn't as drastic as it is often claimed. As an example, I'll do some comparisons with the two most common fixes: change GWM to once per round and change it to -5/+5. But not right now, as I have to put the kids to bed.

I can also do some analysis to any DW house rules, if anyone has any ideas. It has to be something I can work with using my equations though - a house rule changing the DW feat to make it so the offhand attack is not a bonus action (so you can use hex, et al) is not something I can work with.

[NewDM]

That's fair - I did address the point of the day by making the claim that t isn't unreasonable to have high AC opponents at high levels - where GWM favors lower AC. To be clear, I do not want to discount your point, you make a very fair point.

Your earlier point about there not being enough to supplement DW is also a consideration, and anyone who wants to fix the issue should be aware that the difference between the two isn't as drastic as it is often claimed. As an example, I'll do some comparisons with the two most common fixes: change GWM to once per round and change it to -5/+5. But not right now, as I have to put the kids to bed.

I can also do some analysis to any DW house rules, if anyone has any ideas. It has to be something I can work with using my equations though - a house rule changing the DW feat to make it so the offhand attack is not a bonus action (so you can use hex, et al) is not something I can work with.

So no comment on the fact that you will almost always be hitting 60%-80% of the time?

[bid]

Still, against infinite AC, GWM fighter has a 5% to hit for an average of 4d6+15 dpr=1.45 dpr , DW has a 5% to hit for an average of 2d8+5 = 0.7 dpr

4 attacks x 1.45 dpr = 5.8
5 attacks x 0.7 = 3.5
Right. So when you choose a situation where the -5 to hit has no impact, GWM does more damage.

You realize there's a point where DW reaches 13 DPR while GWM still has 5.8 DPR?

[EvilAnagram]

All this math is a nice exercise, but in reality most attacks are going to fall into the 60%-80% hit range because of the most common ACs at any given level, barring outliers.

Most common ACs by CR and Fighter attack bonus by level:
level/CR = AC; Fighter Attack bonus assuming +4 ability score (first feat +5) + Proficiency bonus; % chance to hit
1 = 11; +6; %80
2 = 13; +6; %70
3 = 13; +6; %70
4 = 12; +7; %80
5 = 15; +8; %70
6 = 14; +8; %75
7 = 15; +8; %70
8 = 15; +8; %70
9 = 18; +9; %60
10=18; +9; %60
11=17; +9; %65
12=18; +9; %60
13=18; +10; %65
14=18; +10; %65
15=19; +10; %60
16=19; +11; %65
17=19; +11; %65
18=19; +11; %65
19=20; +11; %60
20=19; +11; %65

Feel free to check the math and the numbers yourselves. As far as I know there are not situations where you would get a direct bonus to attack more than half the rounds in a day (even Battle Master dice only last 18 out of 35 average rounds in a day). If you have disadvantage, that's the only time you'll see a percent chance to hit under 50% or some weird outlier like Animated Armor at CR 1 with an 18 AC.

Starting proficiency is +2, so you're assuming that the Fighter starts with an 18 in its attack slot, which is highly unlikely in most games, not to mention impossible with point buys.

[NewDM]

Starting proficiency is +2, so you're assuming that the Fighter starts with an 18 in its attack slot, which is highly unlikely in most games, not to mention impossible with point buys.

Point buy goes to 15 +2 for race. So that would be 17. 17 is +3. So you are correct. I cede the point.

The new chart would look like:

Most common ACs by CR and Fighter attack bonus by level:
level/CR = AC; Fighter Attack bonus assuming +4 ability score (first feat +5) + Proficiency bonus; % chance to hit
1 = 11; +5; %75
2 = 13; +5; %65
3 = 13; +5; %65
4 = 12; +6; %75 Ability Score Increase
5 = 15; +7; %65
6 = 14; +8; %75 Ability Score Increase (maxed Strength at 20)
7 = 15; +8; %70
8 = 15; +8; %70
9 = 18; +9; %60
10=18; +9; %60
11=17; +9; %65
12=18; +9; %60
13=18; +10; %65
14=18; +10; %65
15=19; +10; %60
16=19; +11; %65
17=19; +11; %65
18=19; +11; %65
19=20; +11; %60
20=19; +11; %65

Drops it to max 75%. However this clearly puts GWM ahead since you will almost never see a hit chance go below 50%.

[mgshamster]

Still, against infinite AC, GWM fighter has a 5% to hit for an average of 4d6+15 dpr=1.45 dpr , DW has a 5% to hit for an average of 2d8+5 = 0.7 dpr

4 attacks x 1.45 dpr = 5.8
5 attacks x 0.7 = 3.5

GWM deals 65% more damage than DW. As I said, numbers stop making sense when your example shows DW becoming better against high ACs.
And why infinite AC? Why not? Using the most common party setups, we have already approached infinity by calculating against 40 AC.

What you've shown here is that when you remove the differences in the chance to hit, the larger dice+modifier does more damage. Well, yeah. That's kind of a duh moment. :)

I haven't figured out exactly where the reduced penalty takes effect, but at some point when you start removing those penalties, you'll see DW break back down again. There's a sweet spot - and in that sweet spot GWM needs to change tactics and not use its -5/+10 option - at which point it stays comparable to DW (at least within a handful of points).

Which brings me back to my original analysis - that the two of them are fairly balanced towards each other. The primary difference - that is, the times you're going to see the largest delta cause between the two - is when the AC is low. Actually, I'm going to find out which percent chance you see GWM's -5/+10 start to outdo the difference between the two than without the -5/+10...

So it's when DW needs an 8+ and GWM needs a 13+, that's when you start seeing GWM (-5/+10) have a greater delta value than GWM (no -5/+10) compared to DW. At that point, the delta value is around 8 points of damage and increases as the chance to hit improves.

What about the high end? When the target value is so high that the penalties for GWM become pointless?

Well, the minimum is 16+ for DW and 20 for GWM. Let's set out GWM baseline for 20: 5.28 DPR.

For DW:
16+: 12.1
17+: 9.7
18+: 7.3
19+: 5.0
20: 2.6

This means that DW will out perform GWM (-5/+10) from AC 22-29. DW will perform comparably (within 10 points*) with GWM (-5/+10) from AC 18-31+. From AC 17 and below, GWM (-5/+10) outperforms DW, scaling up to as much as 40 points more DPR for the very low ACs (8-).

This definitely supports the proposition that GWM (-5/+10) is a mob killer, while DW is better for higher AC targets. Once you drop the -5/+10, GWM is comparable to DW, as they have DPR within a handful of points (somewhere between 5-8 points, depending on how you do the calculation).


I always assume that we are talking about officially released content. Tarrasque peaks at AC 25. Worst case scenario some madman buffs it for a +2.

The game is so much more than just what's been published. Barring that, you can still use PC classes as villains. Barring that, you're literally claiming that those builds which have high AC are not using officially released content. Which is kind of funny, considering how insistent they were that what they did was RAW. :)

Regardless, the AC range can have the same range as skills, so I'm assuming a rough cap of 30 as the "impossible" target, even though you always hit on a 20.

Anyways, please don't think my rebuttals to your comments means that I don't value your input. You've made some good points, and it definitely has helped shape the defining characteristics of DW, especially when looking at where its strength lies.

*I'm defining comparable as "within 10 points" because at this juncture, we're getting five attacks per round. That's a difference of 2 points of damage or less per attack, which is less than half the difference between the weapon damage of a d8 weapon compared to a 2d6 weapon.

[NewDM]

The game is so much more than just what's been published. Barring that, you can still use PC classes as villains. Barring that, you're literally claiming that those builds which have high AC are not using officially released content. Which is kind of funny, considering how insistent they were that what they did was RAW. :)

Using PC classes is a good way to cause a TPK. In my most recent game, I as a player challenged two party members to fight in an arena because they were impressed with Polymorph (Mammoth). They gave me the first round and I polymorphed. They then proceeded to destroy the hit points of the Mammoth in 2 rounds (this was at level 8). They then proceeded to continuously flank me (variant rule in the DMG) and slaughtered my character (Blade Singer) who used Greater Invisibility.

The sheer amount of damage a player class can bring to bear leads me to believe that PC classes are not meant to be used by monsters.

Edit: Also I didn't fail a single Concentration check since I was in Blade Song.

[bid]

*I'm defining comparable as "within 10 points" because at this juncture, we're getting five attacks per round. That's a difference of 2 points of damage or less per attack, which is less than half the difference between the weapon damage of a d8 weapon compared to a 2d6 weapon.
I'm not sure this is a good point. IIRC, around 60% hit DPR are something like 36 and 40, that "within 4 points" is a 10% variation.

OTOH, a 10% variation isn't noticeable out of a party of 4 level-20 characters. I'd say mechanical combat is more for the level 5-11 stretch.

[Firechanter]

At least, PC classes should be used very, very sparingly. We recently had a fight against a group of Duergar Fighters with tricks like Shield Master and Sentinel and it was awesome, really good fight. But you really have to be careful not to overdo it. PC classes have much higher DPM than most monsters.

I'll chime in about the actual topic later, it's too cumbersome on the mobile.

[mgshamster]

Common House Rules:

GWM can only be used once per round.

This is a bit tricky, as only one of the four attacks is going to have increased damage and reduced chance to hit. In addition, if you crit, you could use your bonus action to use the -5/+10 of you haven't already used it. I think I'm going to try for a weighted average for each attack to blend in the chance to hit. I think I'll play with the math tomorrow - it's too late for me to think this deep.

GWM is -5/+5

Max DPR becomes 77.26. Still 30 points higher than DW's Max DPR. Let's see how it compares with different hit probabilities:

The point where the two have the same DPR is now 8+ for DW and 13+ for GWM (-5/+5). They're both 30.9. So the balance point is at the 65% mark. At the high end (20), GWM (-5/+5) is 4.25 DPR. DW performs better all the way up to 19+ to hit. This gives DW the advantage over GWM (-5/+5) from AC 19-30, with comparable damage between AC 7-31+.

This could be a very viable house rule to keep the two within 10 points of each other throughout the entire game, with DW getting ahead during certain times, and at lower level. For very low ACs, GWM (-5/+5) does significantly extra damage.

GWM does -2/+5

Our cap doesn't change from the previous house rule, 77.26, but the difference between GWM and DW will. Without running the numbers to check (it's late, give me a break here), I'm guessing that DW will pull ahead of GWM (-2/+5) only when the ACs are in the 20s and that they'll be comparable when the ACs are in the 10s.

DW grants +1 damage per attack

This house rule will effectively eliminate the difference between DW and GWM (no -5/+10). DW's max DPR becomes 52.5 compared to GWM's 53.2. Less than a single point difference.

GWM still has the massive damage increase for lower ACs, but I'm betting that the balance point will be around 60-65%. Actually, it is at 60%; GWM is at ~32 points here and so is DW.

This could be a viable house rule of you wanted DW to be practically equal to GWM before the -5/+10 modifiers kicked in, but still wanted GWM to have massive damage for lower ACs.

[mgshamster]

I'm not sure this is a good point. IIRC, around 60% hit DPR are something like 36 and 40, that "within 4 points" is a 10% variation.

OTOH, a 10% variation isn't noticeable out of a party of 4 level-20 characters. I'd say mechanical combat is more for the level 5-11 stretch.

That's fair. It's a quibble and not necessarily a point that should matter one way or the other.

The key is to recognize whether you and your group care about a few points difference here and there. If you don't, then as long as you're within a couple of points per attack, you're good. Or you do care, then you need to do something to reduce that delta value.


I'll chime in about the actual topic later, it's too cumbersome on the mobile.

Heh. I've done this entire thread on my mobile, including the opening post and including sending it via PM to have someone else check my math before I posted. :)

I may be on my phone too much. :)

[Firechanter]

Alright, now for a few thoughts:

First off, as it has been pointed out, the typical hit chance PC vs Monster is more like 65%, give or take. Any comparisons between combat styles should be run against this hit chance.

Secondly, to calculate crits correctly - I only learned this trick yesterday - you need to reduce the basic hit chance by your crit chance of 5% (10% for Champions) and calculate the DPR, then add the actual Crit chance * Crit damage.
For example: general hit chance 65%, Crit on 20, normal damage 1d8+4=8.5, crit damage 2d8+4= 13: (0,6*8,5)+(0,05*13) = 5,75
With advantage you need to adjust hit and crit chances accordingly. In this case: (0,84*8,5)+(0,095*13) = 8,375

Thirdly, it has been mentioned but then not explored very much, that just looking at DPR is a bit pointless if you don't look at the echo, i.e. damage taken per round. Here I recommend not pitting two PCs against each other, but looking at the typical attack bonuses and damage values of monsters at a given CR.

For example, when you compare GWF and Dueling, and take into account that even with a humble Uncommon item (+1 Shield) the Dueler will have 3 higher AC, you'll find that the GWFer - even with GWM - will finish the fight more quickly, but still take more damage than the Dueler will take in his longer fight.
And here's the crux for TWFers; their Dual Wield AC bonus is fixed at +1 and _never_ goes up, whereas the Dueler's AC bonus starts at +2 and may increase up to +5. Again, magic shields have very favourable Rarity ratings.

The general proliferation (or lack thereof) of magic items can skew the results in any direction. If you use oldschool weapon type tables when generating treasure (from AD&D or 3E), it is more likely to roll up a longsword than a greatsword, and much more likely to roll up any type of sword than a polearm. So under that regime you may get a lot more mileage out of TWF, and PAM may turn out to be a trap option despite mechanical superiority. Also do note that most of the special weapons in the DMG are swords, and exactly zero of them are polearms.

[Kryx]

Kryx's analysis certainly does account for chance to hit.
Yup. It is based off the DMG and actual averages from the MM. To-hit is definitely taken into account.

For RAW numbers on average over 20 levels a TWF Fighter will do 84% of the DPR that a GWM Fighter does. At 11+ it's 75%.


-5/+10 is a concept that massively imbalances many options - not just GWM vs TWF. See the Balance page (https://docs.google.com/spreadsheets/d/1d-9xDdath8kX_v7Rpts9JFIJwIG3X0-dDUtfax14NT0/edit#gid=2141556455) of my DPR of Classes sheet to see the numbers. Bonus attack from Polearm presents a similar problem.

[EvilAnagram]

Drops it to max 75%. However this clearly puts GWM ahead since you will almost never see a hit chance go below 50%.

This is a misrepresentation of the data. Yes, the average chance to hit at each CR level is fairly high compared to Fighters of that level, but that does not make all of the high-AC and disadvantage-inflicting creatures nonexistant. There still dozens of difficult to hit creatures, most of whom stay difficult to hit and can be used at different CR levels. I threw an Oni with blindsight at a fifth level party, and while they survived, a Dual Wielder would have done quite a bit better. Darkness was the real killer there.

Sure, you can say that on average the chance of high AC is lower than the chance of manageable AC, but you can't say that it will be rare, or even close to rare, for characters to run into high-AC threats in a given campaign.

That said, even though the average AC trends more towards GWM territory, what mgshamster does here is show that we've been wrong to dismiss Dual Wielding and two-weapon fighting. They're not miserable wastes as we've tended to sell them, and we should drop that line.

[Giant2005]

Well I have just read this thread, and the only thing I took away from is it the fact that I seem to play a very different game to everyone else.
Eslin's table comparing attack rolls from players against creatures of a CR equal to the player level is meaningless for me as it is a rare day when the DM throws monsters at us that aren't at level 2 CR levels higher than our character levels.
That also highly influences my perception of the statement "characters do more damage than monsters". Probably solely due to fighting monsters that seem to be of higher levels than other groups face, my experience is that monster damage far outweighs player damage.

[mgshamster]

-5/+10 is a concept that massively imbalances many options - not just GWM vs TWF. See the Balance page (https://docs.google.com/spreadsheets/d/1d-9xDdath8kX_v7Rpts9JFIJwIG3X0-dDUtfax14NT0/edit#gid=2141556455) of my DPR of Classes sheet to see the numbers. Bonus attack from Polearm presents a similar problem.

Request: can you put in something which explains what you are doing in this? You have a bunch of numbers and sometimes colors, associated with other numbers, but most of the time I can't track what you're doing or what you mean by things. A guide would be helpful.

If you already have one in the original thread, can you link to that in your sig?

[mgshamster]

Alright, now for a few thoughts:

First off, as it has been pointed out, the typical hit chance PC vs Monster is more like 65%, give or take. Any comparisons between combat styles should be run against this hit chance.

I disagree. You can have enemies which run the entire range of probabilities on the d20; setting a single probability can lead to false conclusions.


Secondly, to calculate crits correctly - I only learned this trick yesterday - you need to reduce the basic hit chance by your crit chance of 5% (10% for Champions) and calculate the DPR, then add the actual Crit chance * Crit damage.
For example: general hit chance 65%, Crit on 20, normal damage 1d8+4=8.5, crit damage 2d8+4= 13: (0,6*8,5)+(0,05*13) = 5,75
With advantage you need to adjust hit and crit chances accordingly. In this case: (0,84*8,5)+(0,095*13) = 8,375

Nice! It's one way to calculate crits correctly. It's not the only way. The equation I presented in my OP works just as well. It works because a crit doubles the dice, so you account for the first use of the dice in the first half of the equation and the second use of the dice in the second half of the equation. If you want to account for the champion's increased chance to hit, change t from 0.05 to 0.1.

Another thing to consider is that the GWM has a chance to get a bonus attack on a crit, so you need to add in another full attack with a 0.05 probability modifier, then calculate the DPR from that according to the chance to hit.


Thirdly, it has been mentioned but then not explored very much, that just looking at DPR is a bit pointless if you don't look at the echo, i.e. damage taken per round. Here I recommend not pitting two PCs against each other, but looking at the typical attack bonuses and damage values of monsters at a given CR.

I really don't like looking at "typical" monsters, because every game is different. All that kind of analysis will do is tell us how these fair against those typical monsters - and nothing outside of it.

The reason I pitted the two against each other was to assess the value of that +1 in terms of DPR, and it's about 5% of the opponent's. Which makes sense when using my equation, because my equation determines DPR based on the probability to hit in 5% increments of the d20. Each time you increase your AC, your opponents DPR will decrease by 5% (roughly; it's skewed a bit from the auto hit at 20).


For example, when you compare GWF and Dueling, and take into account that even with a humble Uncommon item (+1 Shield) the Dueler will have 3 higher AC, you'll find that the GWFer - even with GWM - will finish the fight more quickly, but still take more damage than the Dueler will take in his longer fight.
And here's the crux for TWFers; their Dual Wield AC bonus is fixed at +1 and _never_ goes up, whereas the Dueler's AC bonus starts at +2 and may increase up to +5. Again, magic shields have very favourable Rarity ratings.

Once the Duelers AC is at a certain point, then the GWM will be forced to not use their -5/+10, which means the only difference in their DPR is going to be from the weapon itself. Now, if we have to add in probability to hit.

Doing some back of the envelope calculations...

Let's say our GWM is wearing +1 full plate (AC 19) and our Duelist is wearing +1 full plate and a +1 shield (AC 22).

Our GWM (-5/+10) is doing 24.8 DPR.

Our GWM (no -5/+10) is doing 35.0 DPR

Our Duelist is doing 29.9 DPR.

Using the -5/+10 is what's called the unbalancing portion of the feat, but notice here that the GWM actually loses if they use that! They have to forego the feat to win the challenge, and in doing so, the only reason he wins is because of the damage difference in the weapons - not because of a feat or skill or anything else, it's just the difference between a 2d6 weapon and a 1d8 weapon. That +2 isn't enough to overcome the rerolls on 1 and 2 in combination with the higher damage value.

So what happens when they're forced to use the same weapon? Let's say a longsword. D8 for the duelist, d10 for the GWM.

Average roll for the duelist is 6.5; average roll for the GWM is 6.3. We already know the GWM can't use its -5/+10 from the analysis above, and here their only advantage is rolling a bonus attack on a crit. Let's see what that does... GWM is doing 30.8 (29.8 without the bonus attack on a crit). So without the feat and using the same weapon, duelist comes out ahead by a tenth of a point; with the feat (for the bonus attack), GWM comes out ahead by 9/10ths of a point. That's fairly balanced. It's the difference in weapon damage that is the key component here.


The general proliferation (or lack thereof) of magic items can skew the results in any direction. If you use oldschool weapon type tables when generating treasure (from AD&D or 3E), it is more likely to roll up a longsword than a greatsword, and much more likely to roll up any type of sword than a polearm. So under that regime you may get a lot more mileage out of TWF, and PAM may turn out to be a trap option despite mechanical superiority. Also do note that most of the special weapons in the DMG are swords, and exactly zero of them are polearms.

You make a good point, but I often find that GMs will tailor the magic items given to their players.

Besides, if I'm going to handwave away "common opponents" in the MM, I can't also accept "common magic items" from the DMG. :)

Customizable game after all. :)

Excellent points all around, thank you.

[Kryx]

Request: can you put in something which explains what you are doing in this? You have a bunch of numbers and sometimes colors, associated with other numbers, but most of the time I can't track what you're doing or what you mean by things. A guide would be helpful.
For the balance page?

Essentially I use GWM as a baseline. For example at level 20 RAW they average to 58 DPR. At the bottom I compare fighter longbow DPR vs GWM DPR and you see it is a shade of green which means it's near the range I expect (85%) - more on that later. Heavy Crossbow is similar, but not quite as good, and hand crossbow blows both the other options out of the water.
So to get the expected numbers I have some settings at the top. I expect Ranged attacks to be about 85% as much damage as melee attacks. So fighter Longbow is expected to be 85% of GWM. For Ranger I expect ranger to do about 90% the DPR of a fighter as it has some spells and utility. 90%*85% = 77%.
In this case Fighter hand crossbow is above what I expect by 7% and Ranger hand crossbow is above what I expect by 8%.

Though now looking back I'm going to change it to not compare to the average of Fighter/Barb GWM

[broodax]

For the balance page?

Essentially I use GWM as a baseline. For example at level 20 RAW they average to 58 DPR. At the bottom I compare fighter longbow DPR vs GWM DPR and you see it is a shade of green which means it's near the range I expect (85%) - more on that later. Heavy Crossbow is similar, but not quite as good, and hand crossbow blows both the other options out of the water.
So to get the expected numbers I have some settings at the top. I expect Ranged attacks to be about 85% as much damage as melee attacks. So fighter Longbow is expected to be 85% of GWM. For Ranger I expect ranger to do about 90% the DPR of a fighter as it has some spells and utility. 90%*85% = 77%.
In this case Fighter hand crossbow is above what I expect by 7% and Ranger hand crossbow is above what I expect by 8%.

Though now looking back I'm going to change it to not compare to the average of Fighter/Barb GWM

I've been absent for some time and came back to see this and was about to look up the old DPR of classes work you did and you beat me to it. I have to take a look at your updated sheet and this balancy goodness.

[Oramac]

GWM:......2d6 damage (avg 8.3)

I'm no math genius, so correct me if I'm wrong, but isn't 7 the average for 2d6? 3.5*2 = 7? What am I missing?

[Rysto]

I'm no math genius, so correct me if I'm wrong, but isn't 7 the average for 2d6? 3.5*2 = 7? What am I missing?

GWF allows you to reroll 1s and 2s, which raises the average damage to 8.3.

[Oramac]

GWF allows you to reroll 1s and 2s, which raises the average damage to 8.3.

Ahh! Gotcha. Ok, thanks!

[Kryx]

I've been absent for some time and came back to see this and was about to look up the old DPR of classes work you did and you beat me to it. I have to take a look at your updated sheet and this balancy goodness.
Ya, it's been a while since I touched it as I've been working on the Shaped sheet for Roll20 pretty extensively.

I don't have full confidence in the balance page (refining it right now to be more accurate - the outliers have a big impact currently) and I'm exploring the balance of some new houserules.

EDIT: Solved it by using TRIMMEAN.

[mgshamster]

For the balance page?

Essentially I use GWM as a baseline. For example at level 20 RAW they average to 58 DPR. At the bottom I compare fighter longbow DPR vs GWM DPR and you see it is a shade of green which means it's near the range I expect (85%) - more on that later. Heavy Crossbow is similar, but not quite as good, and hand crossbow blows both the other options out of the water.
So to get the expected numbers I have some settings at the top. I expect Ranged attacks to be about 85% as much damage as melee attacks. So fighter Longbow is expected to be 85% of GWM. For Ranger I expect ranger to do about 90% the DPR of a fighter as it has some spells and utility. 90%*85% = 77%.
In this case Fighter hand crossbow is above what I expect by 7% and Ranger hand crossbow is above what I expect by 8%.

Though now looking back I'm going to change it to not compare to the average of Fighter/Barb GWM

Ok; I can see it. Thank you!

Question: How are you calculating the GWM average? With -5/+10? Without? Both dependent on AC? A single AC or across some range with the average calculated from that?

Earlier you said something about using the average from the MM or DMG. Or are you using the median? While it's good to determine a baseline (which I think is what you've done), it doesn't account for variance, which is what I've done.

I certainly don't think that my analysis proves yours wrong in anyway, but I also don't think yours proves mine wrong, they just look at the situation from different angles and take in to account different variables (assuming what I said above is accurate).

We also have different goals - you want the damage to be within ~95% ish. I'm ok with a bit more variance provided the to hit targets account for it (which it can if we don't abide by the AC limits for high teens/low 20s).

I agree that TWF needs a bit more; I just don't believe it needs a strong push, and I don't believe that it's that much worse than is traditionally claimed. I feel that your analysis, while very good, is limited due to looking only at an average target value as compared to a variable target value. Please correct me if I'm wrong here.

[NewDM]

Well I have just read this thread, and the only thing I took away from is it the fact that I seem to play a very different game to everyone else.
Eslin's table comparing attack rolls from players against creatures of a CR equal to the player level is meaningless for me as it is a rare day when the DM throws monsters at us that aren't at level 2 CR levels higher than our character levels.
That also highly influences my perception of the statement "characters do more damage than monsters". Probably solely due to fighting monsters that seem to be of higher levels than other groups face, my experience is that monster damage far outweighs player damage.

Yes, if you use a CR above the party's level then you are not following the DMG.
DMG page 82:
"When putting together an encounter or adventure, especially at lower levels, exercise caution when using monsters whose challenge rating is higher than the party's average level. Such
a creature might deal enough damage with a single action to take out adventurers of a lower level. For example, an ogre has a challenge rating of 2, but it can kill a 1st-level wizard with a single blow. In addition, some monsters have features that might be difficult or impossible for lower-level characters to overcome. For example, a rakshasa has a challenge rating of 13 and is immune to spells of 6th level and lower. Spellcasters of 12th level or lower have no spells higher than 6th level, meaning that they won't be able to affect the rakshasa with their magic, putting the adventurers at a serious disadvantage. Such an encounter would be significantly tougher for the party than the monster's challenge rating might suggest."


I disagree. You can have enemies which run the entire range of probabilities on the d20; setting a single probability can lead to false conclusions.

My information is from the most common ACs of each level. This means more often than not you will fact creatures with those ACs. The only thing that will affect it is using lower level CRs in larger groups, but this won't change the fact that GWM is the better option because lower CR enemies have lower AC.


Nice! It's one way to calculate crits correctly. It's not the only way. The equation I presented in my OP works just as well. It works because a crit doubles the dice, so you account for the first use of the dice in the first half of the equation and the second use of the dice in the second half of the equation. If you want to account for the champion's increased chance to hit, change t from 0.05 to 0.1.

This will give you different numbers. The best way is to calculate hit and crit separately and then add the result together for each attack. There are less mistakes this way.


Another thing to consider is that the GWM has a chance to get a bonus attack on a crit, so you need to add in another full attack with a 0.05 probability modifier, then calculate the DPR from that according to the chance to hit.

Or you can take the 35 rounds per day average and toss in an extra crit 5% of the time. That means 1.75 crits per day. Divide that damage back out and you end up with 0.05 * crit damage DPR. Or basically you find that it doesn't add up to much at all.


I really don't like looking at "typical" monsters, because every game is different. All that kind of analysis will do is tell us how these fair against those typical monsters - and nothing outside of it.

That's a fair argument, but the ACs don't range by more than a point or two in either direction from the most common so it doesn't really matter.


The reason I pitted the two against each other was to assess the value of that +1 in terms of DPR, and it's about 5% of the opponent's. Which makes sense when using my equation, because my equation determines DPR based on the probability to hit in 5% increments of the d20. Each time you increase your AC, your opponents DPR will decrease by 5% (roughly; it's skewed a bit from the auto hit at 20).

Once the Duelers AC is at a certain point, then the GWM will be forced to not use their -5/+10, which means the only difference in their DPR is going to be from the weapon itself. Now, if we have to add in probability to hit.

Doing some back of the envelope calculations...

Let's say our GWM is wearing +1 full plate (AC 19) and our Duelist is wearing +1 full plate and a +1 shield (AC 22).

Our GWM (-5/+10) is doing 24.8 DPR.

Our GWM (no -5/+10) is doing 35.0 DPR

Our Duelist is doing 29.9 DPR.

Using the -5/+10 is what's called the unbalancing portion of the feat, but notice here that the GWM actually loses if they use that! They have to forego the feat to win the challenge, and in doing so, the only reason he wins is because of the damage difference in the weapons - not because of a feat or skill or anything else, it's just the difference between a 2d6 weapon and a 1d8 weapon. That +2 isn't enough to overcome the rerolls on 1 and 2 in combination with the higher damage value.

So what happens when they're forced to use the same weapon? Let's say a longsword. D8 for the duelist, d10 for the GWM.

Average roll for the duelist is 6.5; average roll for the GWM is 6.3. We already know the GWM can't use its -5/+10 from the analysis above, and here their only advantage is rolling a bonus attack on a crit. Let's see what that does... GWM is doing 30.8 (29.8 without the bonus attack on a crit). So without the feat and using the same weapon, duelist comes out ahead by a tenth of a point; with the feat (for the bonus attack), GWM comes out ahead by 9/10ths of a point. That's fairly balanced. It's the difference in weapon damage that is the key component here.

You make a good point, but I often find that GMs will tailor the magic items given to their players.

Besides, if I'm going to handwave away "common opponents" in the MM, I can't also accept "common magic items" from the DMG. :)

Customizable game after all. :)

Excellent points all around, thank you.

You're welcome.

[Kryx]

Questions *snip*
The balance page is meant to summarize all the other tabs. To see the actual details you need to go to those tabs. For example to see the RAW damage for Fighter GWM you should go to the "PrimaryMeleeRaw" tab where you can see that GWM builds do indeed use -5/+10. In my houserules version they do not.
AC is calculated on the "S" tab. It currently assumes that the average CR fought is about 1 lower than the PCs level as per the DMG recommendations of not using too high of a level. Seme monsters will be above and some below. I use the average. Based on the CR I use AC determined from the DMG. I also compared it against another GiantITP member's spreadsheet, but I also have that AC per CR in my Spells Balance (https://docs.google.com/spreadsheets/d/1N4QC6EmXE0avgk8jK1aubJcaFoZDYw8b_DuPHh8aBTc/edit#gid=1997326283) spreadsheet. It calculates the actual average CR per monsters in the MM and lines up with what I have on the DPR of classes page.

Regarding different angles: They are indeed great. I calculate things over a whole day. Burst is very different than what I'm trying to balance.
Forum math: I never bother to read forum math because it is very difficult to substantiate the assumptions made by the author and is typically wildly inaccurate. For example you state that a Fighter does 99.2 DPR at level 20 which is no where near close to the 62 that I have for RAW GWM Fighter and you're not even using battlemaster tricks or action surge like I am. You seemingly ignore to-hit all together (which most forum math does). Ignoring to-hit while comparing a feature that is dependent on lowering your to-hit to increase damage is very misleading.

You're welcome to adjust my goals to suit your desires to see how the system stacks up. You could also use the high level of AC that I have on the "S" tab if you want.

TWF is significantly behind. Polearm bonus attack and -5/+10 are significantly ahead of every other build. My math shows it and many other peoples math shows it as well. Though as you stated people have different goals. It is quite clear that WotC wanted TWF to never compete with GWM or Polearm in damage, yet they never gave them added utility. It has been a problem of D&D since the early editions. I prefer TWF actually being able to compete. The same as I prefer the Sorcerer actually being able to compete with a Wizard.

[NewDM]

The balance page is meant to summarize all the other tabs. To see the actual details you need to go to those tabs. For example to see the RAW damage for Fighter GWM you should go to the "PrimaryMeleeRaw" tab where you can see that GWM builds do indeed use -5/+10. In my houserules version they do not.
AC is calculated on the "S" tab. It currently assumes that the average CR fought is about 1 lower than the PCs level as per the DMG recommendations of not using too high of a level. Seme monsters will be above and some below. I use the average. Based on the CR I use AC determined from the DMG. I also compared it against another GiantITP member's spreadsheet, but I also have that AC per CR in my Spells Balance (https://docs.google.com/spreadsheets/d/1N4QC6EmXE0avgk8jK1aubJcaFoZDYw8b_DuPHh8aBTc/edit#gid=1997326283) spreadsheet. It calculates the actual average CR per monsters in the MM and lines up with what I have on the DPR of classes page.

Regarding different angles: They are indeed great. I actually never bother to read forum math because it is very difficult to substantiate the assumptions made by the author and is typically wildly inaccurate. For example you state that a Fighter does 99.2 DPR at level 20 which is no where near close to the 62 that I have for RAW GWM Fighter and you're not even using battlemaster tricks or action surge like I am. You seemingly ignore to-hit all together (which most forum math does). Ignoring to-hit while comparing a feature that is dependent on lowering your to-hit to increase damage is very misleading.

You're welcome to adjust my goals to suit your desires to see how the system stacks up. You could also use the high level of AC that I have on the "S" tab if you want.

TWF is significantly behind. Polearm bonus attack and -5/+10 are significantly ahead of every other build. My math shows it and many other peoples math shows it as well. Though as you stated people have different goals. It is quite clear that WotC wanted TWF to never compete with GWM or Polearm in damage, yet they never gave them added utility. It has been a problem of D&D since the early editions. I prefer TWF actually being able to compete. The same as I prefer the Sorcerer actually being able to compete with a Wizard.

WotC probably meant for TWF to be for Rogues and Rangers and GWM to be for Fighters and Barbarians.

[EvilAnagram]

Yes, if you use a CR above the party's level then you are not following the DMG.
DMG page 82:
"When putting together an encounter or adventure, especially at lower levels, exercise caution when using monsters whose challenge rating is higher than the party's average level. Such
a creature might deal enough damage with a single action to take out adventurers of a lower level. For example, an ogre has a challenge rating of 2, but it can kill a 1st-level wizard with a single blow. In addition, some monsters have features that might be difficult or impossible for lower-level characters to overcome. For example, a rakshasa has a challenge rating of 13 and is immune to spells of 6th level and lower. Spellcasters of 12th level or lower have no spells higher than 6th level, meaning that they won't be able to affect the rakshasa with their magic, putting the adventurers at a serious disadvantage. Such an encounter would be significantly tougher for the party than the monster's challenge rating might suggest."

"Exercise caution," does not mean, "never do this." In fact, the plethora of challenge ratings above 20 suggests that you will fight above your CR at some point.

[NewDM]

"Exercise caution," does not mean, "never do this." In fact, the plethora of challenge ratings above 20 suggests that you will fight above your CR at some point.

"exercise caution...might deal enough damage with a single action to take out adventurers of a lower level...some monsters have features that might be difficult or impossible for lower-level characters to overcome."

Exercise caution or you might kill your players' characters or put them in an impossible to win encounter.

[Saggo]

Question: How are you calculating the GWM average? With -5/+10? Without? Both dependent on AC? A single AC or across some range with the average calculated from that?

Earlier you said something about using the average from the MM or DMG. Or are you using the median? While it's good to determine a baseline (which I think is what you've done), it doesn't account for variance, which is what I've done.
Might be able to add some.

If you look on the S tab, you'll see the ACs being used (normal line by default). He started with DMG recommendations and I confirmed the DMG matched the average AC of a similar CR. You can look at the link in my signature. Deviation is low and the median is similar to the average which means there's not a lot of flux in each CR. There is the assumption that you won't see many mobs with a CR higher than level by much.

Basically means that while you can never guess the exact mobs a given table will face, you know the expected AC for a level appropriate CR.

Each build that says GWM is almost always using the power attack. You can confirm it by look for a line of -5 in each build. S tab again will show you the number of combat rounds a day and how it was calculated. Balance is, at least now, the average daily DPR for each level taken straight from the DPR tabs.

[Bladeyeoman]

TWF is significantly behind. Polearm bonus attack and -5/+10 are significantly ahead of every other build. My math shows it and many other peoples math shows it as well. Though as you stated people have different goals. It is quite clear that WotC wanted TWF to never compete with GWM or Polearm in damage, yet they never gave them added utility. It has been a problem of D&D since the early editions. I prefer TWF actually being able to compete. The same as I prefer the Sorcerer actually being able to compete with a Wizard.

It seems like most of the homebrew solutions I've encountered have tried to balance the damage by decreasing GWM damage or increasing TWF damage. You make an interesting point with the utility comment. Are there modifications to the TWF feat that enhance it without just increasing damage across the board? How would a bonus to-hit work out, mathwise? This could strengthen the tradeoff between GWM for devastating easily-hit targets and TWF for wittling down high AC targets. TWF feat offering a parry ability as a reaction (though this might step on the toes of monks, I think)? Is there something else that makes sense and balances well?

[Kryx]

"Exercise caution," does not mean, "never do this." In fact, the plethora of challenge ratings above 20 suggests that you will fight above your CR at some point.
And the many CRs lower than 1 shows that the PCs are also meant to fight lower level CRs. It's a mix.
Now if a DM throws higher CRs against you all the time then a feature like -5/+10 becomes a trap option.

It either shines brightly, is awful, or is unused. 2/3 of those means the feat was a trap and 1/3 of those means the PCs finish a normal encounter quicker.
Overall I think it's a horribly designed feat in that regard.

[Bladeyeoman]

Although it seems like one of the biggest issues with TWF is that it uses your bonus action every round, and so doesn't synergize well with other spells/abilities that compete for bonus actions.

[Kryx]

It seems like most of the homebrew solutions I've encountered have tried to balance the damage by decreasing GWM damage or increasing TWF damage. You make an interesting point with the utility comment. Are there modifications to the TWF feat that enhance it without just increasing damage across the board? How would a bonus to-hit work out, mathwise? This could strengthen the tradeoff between GWM for devastating easily-hit targets and TWF for wittling down high AC targets. TWF feat offering a parry ability as a reaction (though this might step on the toes of monks, I think)? Is there something else that makes sense and balances well?
Thematically TWF should be all about more attacks or some feature like parry imo. That's why I chose to give TWF an extra bonus attack at level 11 - it fits the thematics. Even then it still falls short of GWM and Polearm+GWM when those have had the polearm bonus and -5/+10 removed. That tells you how poor the balance was to start with.

There are plenty of options to explore. I tried the +1 to attack for a while and some other balance options, but thematics brought me back to an extra bonus attack. Parry or AC bonus or something like that would be the next best thematic option, but is difficult to balance.

[Giant2005]

And the many CRs lower than 1 shows that the PCs are also meant to fight lower level CRs. It's a mix.
Now if a DM throws higher CRs against you all the time then a feature like -5/+10 becomes a trap option.

It either shines brightly, is awful, or is unused. 2/3 of those means the feat was a trap and 1/3 of those means the PCs finish a normal encounter quicker.
Overall I think it's a horribly designed feat in that regard.

I don't think it is horribly designed - it has its uses and it is an incredible asset to have during those uses. The issue is that people keep wanting it to do more than it was designed for, and worse than that, that want compels them to insist it is powerful.
The truth is that its usefulness is far more limited than what numbers alone can express.
The fights where it is an asset (the easier fights with low ACs) are the fights where you don't need it to be an asset. The fights where it is of no use (the harder fights with high ACs) are the fights where you need all of the help you can get.

[RulesJD]

One reason I can see going DW over GWM:


1. If you're in a campaign that will likely have 2+ 1h magical weapons. Once you start adding magical weapons into the mix, suddenly the math changes a lot. Other than that, Polearm/Greatsword wins out.

[EvilAnagram]

"exercise caution...might deal enough damage with a single action to take out adventurers of a lower level...some monsters have features that might be difficult or impossible for lower-level characters to overcome."

Exercise caution or you might kill your players' characters or put them in an impossible to win encounter.

Right. Exercise caution. I used an Oni against a fifth level party. They killed it. I made sure not to use a creature that could kill them in a single blow or render the situation unwinnable, but was still of a very challeging CR. This is how you exercise caution. You are reading the blurb as saying, "never use higher CR or everyone will die," which isn't what it's saying and isn't true.

[Kryx]

The fights where it is an asset (the easier fights with low ACs) are the fights where you don't need it to be an asset. The fights where it is of no use (the harder fights with high ACs) are the fights where you need all of the help you can get.
I agree and wrote so above. Keep in mind those negatives for fighting above your CR don't exist for the Barbarian who has constant advantage or other cases like bless.
It's a trap option that is horribly designed imo.

[Nu]

Although it seems like one of the biggest issues with TWF is that it uses your bonus action every round, and so doesn't synergize well with other spells/abilities that compete for bonus actions.

Personally, this is my problem with it. There are so many neat things you can do with your bonus action, like Shield Mastery, Cunning Action, Bardic Inspiration, some battlemaster maneuvers, any spell with a cast time of a bonus action... and you definitely feel like you lose a lot more with TWF if you use your bonus action to do any of those things (though it's mutually exclusive with Shield Mastery at least).

That's kind of why I don't even like to dual wield on my rogues, though I realize sometimes it's beneficial to do so--I want that Cunning Action dammit, it's a core feature of the class!

[EvilAnagram]

Thematically TWF should be all about more attacks or some feature like parry imo. That's why I chose to give TWF an extra bonus attack at level 11 - it fits the thematics. Even then it still falls short of GWM and Polearm+GWM when those have had the polearm bonus and -5/+10 removed. That tells you how poor the balance was to start with.

Honestly, with this math it seems fairly balanced still.

[Giant2005]

One reason I can see going DW over GWM:


1. If you're in a campaign that will likely have 2+ 1h magical weapons. Once you start adding magical weapons into the mix, suddenly the math changes a lot. Other than that, Polearm/Greatsword wins out.

Needing two magic weapons to do the same thing that others can do with one isn't an advantage, it is a disadvantage.

A better advantage is the fact that having more, less powerful hits means you are better able to spread your damage out and not waste it on overkill damage. The guy that has 3 attacks of 1D8+5 is far superior to the guy with two attacks of 2D6+15 when you are going up against three guys with 6 hit points remaining each.

[mgshamster]

The balance page is meant to summarize all the other tabs. To see the actual details you need to go to those tabs. For example to see the RAW damage for Fighter GWM you should go to the "PrimaryMeleeRaw" tab where you can see that GWM builds do indeed use -5/+10. In my houserules version they do not.
AC is calculated on the "S" tab. It currently assumes that the average CR fought is about 1 lower than the PCs level as per the DMG recommendations of not using too high of a level. Seme monsters will be above and some below. I use the average. Based on the CR I use AC determined from the DMG. I also compared it against another GiantITP member's spreadsheet, but I also have that AC per CR in my Spells Balance (https://docs.google.com/spreadsheets/d/1N4QC6EmXE0avgk8jK1aubJcaFoZDYw8b_DuPHh8aBTc/edit#gid=1997326283) spreadsheet. It calculates the actual average CR per monsters in the MM and lines up with what I have on the DPR of classes page.

Ah, thank you!


Regarding different angles: They are indeed great. I calculate things over a whole day. Burst is very different than what I'm trying to balance.

Very true - I'm not calculating burst damage, either, but rather a standard DPR.


Forum math: I never bother to read forum math because it is very difficult to substantiate the assumptions made by the author and is typically wildly inaccurate. For example you state that a Fighter does 99.2 DPR at level 20 which is no where near close to the 62 that I have for RAW GWM Fighter and you're not even using battlemaster tricks or action surge like I am. You seemingly ignore to-hit all together (which most forum math does). Ignoring to-hit while comparing a feature that is dependent on lowering your to-hit to increase damage is very misleading.

I did provide the formula. :)

....And now I'm going to have to do some dressing down, unfortunately.

If you read past the first paragraph (which your comment here strongly suggests that you didn't), you'll see that I am very much incorporating the attack requirements by looking at the number on the d20 which is required to roll. It's built right in to the formula. This also means that you're critiquing my math in very bad faith, as you didn't even read it and you're making false assumptions about my analysis and claiming it's invalid because it didn't match your previous analysis - without even checking to see if it did!

Please do not converse in bad faith, you've done some excellent work and I'd really hate to see someone with quality work have to be dismissed because they're blinded by their own ego. I'm giving you the benefit of the doubt here, and I'm going to assume that this is simply a mistake, but if you continue to discuss in bad faith, I'll no longer respect anything you have to say.

Now that this bit unpleasantness is over, let's get back to the fun.

The 99.2 is before adjusting for hit probabilities. It's just a straight calculation of average damage; (8.3+15)*4 = 93.2 It's actually a bit higher, because with that I didn't include crits or the bonus attack on a crit, which adjusts it by a few points.

After that, adjust for probability to hit (h term in my equation). If you need a 12 on the d20 to hit, that's a 45% chance to hit. 0.45*93.2, then add in the crit (0.05*8.3)4, and add in the bonus attack (0.05(0.45(8.3+15)+0.05*8.3)) for total DPR against whatever AC requires a 12 or higher to hit.


You're welcome to adjust my goals to suit your desires to see how the system stacks up. You could also use the high level of AC that I have on the "S" tab if you want.

Thanks! What formula are you using to calculate DPR? Looking at the sheet on my phone doesn't show me the command lines, just the final figures in the cells.


TWF is significantly behind. Polearm bonus attack and -5/+10 are significantly ahead of every other build. My math shows it and many other peoples math shows it as well. Though as you stated people have different goals. It is quite clear that WotC wanted TWF to never compete with GWM or Polearm in damage, yet they never gave them added utility. It has been a problem of D&D since the early editions. I prefer TWF actually being able to compete. The same as I prefer the Sorcerer actually being able to compete with a Wizard.

How are you defining "significantly"?

I agree that it's behind, it's just not that far behind when looking at higher AC than +/-1 CR or assuming an average AC for a CR. I don't believe significant adjustment is needed, only minor adjustment. My anaysis also showed where the balance point was - from what range of ACs that DW did better and what range of ACs did they perform comparably (which I defined as within 10 DPR, which very well may exceed your percentage based specifications, due to the nature of percentage analysis).

My analysis of some house rules from last night show that a minor adjustment to GWM (-5/+5) or DW (+1 to damage) can bring it in line, and give them different strengths. My analysis with the duelist from this morning shows that when you take away the -5/+10, the primary difference isn't with the feat, it's with the weapon used. Change the weapon damage and you bring them more in line with each other.

On a side note, I've also seen analysis that GWF and TWF are on par with each other when you don't use the optional feats. That analysis compared two non-berserker barbarians.

[Giant2005]

Personally, this is my problem with it. There are so many neat things you can do with your bonus action, like Shield Mastery, Cunning Action, Bardic Inspiration, some battlemaster maneuvers, any spell with a cast time of a bonus action... and you definitely feel like you lose a lot more with TWF if you use your bonus action to do any of those things (though it's mutually exclusive with Shield Mastery at least).

That's kind of why I don't even like to dual wield on my rogues, though I realize sometimes it's beneficial to do so--I want that Cunning Action dammit, it's a core feature of the class!

That is my biggest aversion to it too, but more specifically my aversion is the fact that I can easily get another attack in with Polearm Master.
Why dual-wield when I can simply take a feat and be able to do the same thing without having to forgo a shield?

[Rysto]

A better advantage is the fact that having more, less powerful hits means you are better able to spread your damage out and not waste it on overkill damage. The guy that has 3 attacks of 1D8+5 is far superior to the guy with two attacks of 2D6+15 when you are going up against three guys with 6 hit points remaining each.

That would be true if GWM didn't already have overkill protection built it. Don't forget that it gives you a bonus attack when you kill anything.

[Kryx]

*snip*
You caught me! As I said I never read forum math.
Please excuse any perceived ego. I have no desire to "win arguments" like most of the internet - I'm here to learn things and share those learnings
Forum math: I have never once seen forum math that is complete and presents a whole picture. Every single case I've ever seen is misinformation at best.

No offense to you, but the same is true in the very situation. Your math on the first page (now that I've read it) is deceiving by omission of features and choosing an abnormally high AC. Features like Reckless Attack, Bless, Battle Master trip, Paladin OoV, etc have a huge impact and have been ignored. As pointed out by Saggo and myself the AC per CR has little variance overall. Taking that and average CR enemies the normal to hit is 65%.
But even if we used your higher AC items those outlined features still come out ahead.

To see how DPR is calculated you'll have to look at the spreadsheet. It's fairly straightforward when you can see the cells referenced.

By how far TWF is behind you can see that on the balance tab. By RAW:

Barb TWF does 75% of the damage a Barb GWM does
Fighter TWF does 84% of the damage a Fighter GWM does (same as S&B)
Paladin TWF does 83% of the damage a Paladin GWM does (less than S&B)
Paladin OoV TWF does fine with 91% of the damage a Paladin OoV GWM does. Less than I'd expect, but at least within 10% of GWM
Ranger TWF does 84% of the damage a Fighter GWM does (pretty close to expected)
Rogue TWF does 78% of the damage a Fighter GWM does (trash)
Bladelock TWF does 90% of the damage a Bladelock GWM does (a bit low)

TWF doing the same (or less) damage as a Sword & Board character is laughable as S&B gets many other benefits.

GWF and TWF are closer without feats, definitely. That's because the horribly unbalancing -5/+10 and polearm bonus attack are gone.

[RulesJD]

Needing two magic weapons to do the same thing that others can do with one isn't an advantage, it is a disadvantage.

*snip*

Not necessarily, and that's my point.

If you somehow know that 2 1h weapons will drop that have additional damage riders/health absorbs/etc., the math can change in favor of being able to wield both weapons. This is especially true if there are benefits to just holding the weapon (damage resistance, increased AC, etc).

As an example, the Curse of Strahd hardcover has more (and better) 1h weapons than 2h Heavy weapons. Is this enough to make DW better than GWM? Hell no. But it is a consideration if you know your campaign will have lots of really good 1h weapons.

[mgshamster]

If DW gave you an extra attack on the Bonus Action (two attacks instead of one), it would do slightly more damage than GWM (no -5/+10). Right around 57 DPR vs 53.

This means that a house rule giving an additional attack on TWF can bring it better in line with GWM when it's not using its power attack feature. The increase in damage should also bring the balance point down a little bit on the probability axle for the d20, more in line with NewDM's AC analysis provided earlier. I'd have to check, but I've got a lot at work right now. I'll try at lunch or maybe this evening.

[mgshamster]

You caught me! As I said I never read forum math.
Please excuse any perceived ego. I have no desire to "win arguments" like most of the internet - I'm here to learn things and share those learnings
Forum math: I have never once seen forum math that is complete and presents a whole picture. Every single case I've ever seen is misinformation at best.

Quick response, but thank you very much for not taking offense. I apologize for having to say that, I've just had to deal with some people who do not covmerse in good faith recently, and it's been getting on my nerves.

I'm happy we can move forward with a positive attitude. Thanks for being jovial about it. :)

I'll try to respond a bit later about your other insights and see where we can take it.

[Bladeyeoman]

Guys! This is the internet! You're not supposed to be mature adults about this!

...but seriously, as a by-stander, it's great (and, sadly, relatively unusual) to see disagreements on an internet forum discussed reasonably. Thanks.

[Kryx]

If DW gave you an extra attack on the Bonus Action (two attacks instead of one), it would do slightly more damage than GWM (no -5/+10).
By RAW it seems you are correct. Perhaps the Rend feature I suggested while discussing this long ago would be a better fit for more RAW games. Rend = if both mainhand and offhand hit you do additional damage equal to your proficiency bonus.

For my own Houserules (without -5/+10, without polearm bonus, with TWF getting ability scores by default) you can see that TWF is pretty much within the range I expect it to be.

I actually don't care to discuss this too much to be honest. Even beyond the math I find the -5/-10 feature to be highly undesirable as I've discussed above. I'd personally much rather remove it and balance without it. But that bias came after I played with it, not from the math. It is too highly abusable (also for Hand Crossbows).

[R.Shackleford]

Pretty much every time I see someone talk about wanting to play a DW character, the general recommendation is to avoid it because it's that much worse than GWM. My point is that it's not that much worse than GWM. Without the -5/+10, they're about on par, with DW slightly better at lower levels and GWM slightly better at higher levels. With it, GWM is better when it's easier to hit and DW is better when it's harder to hit.

Typically people believe in this white room where monster AC and Monster HP are at such a level you need to be doing max damage and have max accuracy to take them out.

This is not the case. DMs throw in random monsters with random ACs and with random HP.

The only time these numbers really matter (:) hit and damage) is when compared to the creature you are attacking.

Of a GWM and DW are each taking on a critter by themselves (which happens... Not all that often in a team game). The GWM can deal 50 damage but the DW deals 30 damage (each per round consistently).

The critters each have 30 hp (or less)... On average they will kill the critter at the same time. If the critters have 31 - 50 HP then the GWM wins.

If the critter has 51 - 60 then they are equal again! It will take two rounds for both of them to kill the critter.

So just doing more damage doesn't really matter all that much unless one goes way beyond the target HP.

200 HP takes the GWM 4 rounds to kill.

200 HP takes the DW 6 rounds to kill.

However in those rounds the enemy will be doing stuff to. That +1 AC could be protecting you enough to where you last longer than your friend (personally I think the +1 AC is a joke and they should have made that part better but then again +1 AC on a fighter means potentially one of the highest ACs on the team).

I'm not saying DW IS better than GWM, but the idea of "more damage = Win" is false in real gameplay because of the variable enemy HP.

You also take into account the dodge action and dynamic environments. My critters love the dodge action and their dynamic environments.

[EvilAnagram]

Since mgshamster mentioned my formula, I suppose I'll bring it in with the caveat that I find his formula much more elegant.

Note that this is specific to raging Barbarians who did not take the Berserker path.

Defining Terms:
n = number of d20 results that provide a non-critical hit
A = average damage of non-critical hit
C = average damage of a critical hit
B = average damage of non-critical bonus action hit
Cb = average damage of critical bonus action hit
D = average damage per turn
x = average damage of weapon die
At level 20: A=x+9, C=5x+9, B=x+4, Cb=5x+4

Formula for GW Barbarians:
D=2([nA+C]/20)
or
D=(n[x+9]+5x+9)/10
with a greataxe
D=(15.5n+41.5)/10


Remember that with Brutal Critical, the greataxe tends to outperform the greatsword.

Formula for TW Barbarians:
D=2([nA+C]/20)+([nB+Cb]/20)
or
D=([n{x+9}+5x+9]/10) + ([n{x+4}+5x+4]/20)
assuming d6 weapon dice
D=([12.5n+26.5]/10)+([7.5n+21.5)/20)


At a 50% chance to hit (n=9), TWB is ahead of GWB by less than a point (18.3, 18.1). With only a 5% miss chance, D=32.05 while for TWB D=32.98, and at no point is there a significant disparity in damage.

[mgshamster]

However in those rounds the enemy will be doing stuff to. That +1 AC could be protecting you enough to where you last longer than your friend (personally I think the +1 AC is a joke and they should have made that part better but then again +1 AC on a fighter means potentially one of the highest ACs on the team).

What brought this whole thing up for me was a conversation with a friend - I asked, "If we keep giving the TWF an increasing bonus to AC, at what point would it be balanced with the GWM?"

I never actually analyzed that question; instead I got caught up in the DPR calulations. But if my initial and unexplored hypothesis of the +1 AC is equal to approximately 5% of the DPR (of the opponent), then pitting it against GWM could tell us how much of an AC bonus is needed to make it balanced. Quick estimate: about a +2 AC is needed to balance against GWM (no -5/+10) in general or GWM (-5/+10) at AC 19. About 5% is worth around 4.7 points of damage vs the GWM, and the GWM (no -5/+10) is doing about 6-8 DPR above the DW across the entire range of ACs. Likewise, with the GWM (-5/+10) is doing 7 damage more at AC 19.

It's shaky ground to equivalate AC bonus to damage output, but I'm not sure how else to do it. Suggestions welcome.

[Giant2005]

Paladin TWF does 83% of the damage a Paladin GWM does (less than S&B)

How do you figure that?
Against AC 20, a TWF Paladin has two attacks with a 55% chance of a hit for 1d12+1d8+5, with a 5% chance of a crit for 2d12+2d8+5. It also has a third attack with a 55% chance of a hit for 1d12+1d8 and a 5% chance of a crit for 2d12+2d8.
2(0.55(6.5+4.5+5))+2(0.05(13+9+5))+0.55(6.5+4.5)+0 .05(13+9) = 27.45 DPR.

Against AC 20, a GWM Paladin that is using the -5/+10 feature, has two attacks with a 35% chance of a hit for 2d6+1d8+15, with a 5% chance of a crit for 4d6+2d8+15 with a 5% chance of a bonus attack that inflicts +0.3(2d6+1d8+15)+0.05(4d6+2d8+15).
2(0.3(8.33+4.5+15))+2(0.05(16.67+9+15))+0.0975(0.3 (8.33+4.5+15)+0.05(16.67+9+15)) = 21.78 DPR.

Against AC 20, a GWM Paladin that isn't stupid enough to be using the -5/+10 feature, has two attacks with a 55% chance of a hit for 2d6+1d8+5, with a 5% chance of a crit for 4d6+2d8+5 with a 5% chance of a bonus attack that inflicts +0.55(2d6+1d8+5)+0.05(4d6+2d8+5).
2(0.55(8.33+4.5+5))+2(0.05(16.67+9+5))+0.0975(0.55 (8.33+4.5+5)+0.05(16.67+9+5)) = 23.79 DPR.

In all cases the TWF guy comes out ahead not just by having higher DPR, but by having 2 more AC, the ability to spread his damage among more enemies if necessary, and more chances of smiting.

[Rysto]

Against AC 20, a TWF Paladin has two attacks with a 55% chance of a hit for 1d12+1d8+5, with a 5% chance of a crit for 2d12+2d8+5. It also has a third attack with a 55% chance of a hit for 1d12+1d8 and a 5% chance of a crit for 2d12+2d8.

Shouldn't the 1d12s be 1d10? Unless the Paladin is mounted and using lances, you can't dual wield a 1d12 weapon.

[Finieous]

Maybe you should look at his DPR of Classes doc. He's not calculating for "AC 20" and he's calculating average DPR over the course of a typical adventuring day using all available resources (including OoV advantage).

[Giant2005]

Shouldn't the 1d12s be 1d10? Unless the Paladin is mounted and using lances, you can't dual wield a 1d12 weapon.

I was using lances (the highest non-mounted, 1h weapon is only d8 btw)

[Saggo]

How do you figure that?

You're still using isolated rounds. 83% is the mean of each level's per adventuring day ratio of TWF to GWM, or to put it naturally the balance of Paladin across all levels.

[Giant2005]

You're still using isolated rounds. 83% is the mean of each level's per adventuring day ratio of TWF to GWM, or to put it naturally the balance of Paladin across all levels.

That makes sense. The Paladin is a terrible dual-wielder prior to level 11, so levels 1-10 would bring the average down considerably.

[Saggo]

That makes sense. The Paladin is a terrible dual-wielder prior to level 11, so levels 1-10 would bring the average down considerably.

That's true. Paladin TWF does improve from 76% to 91%, 1 to 20 respectively. He specifically uses a trimmed mean though, excluding a portion of the extremities.

[OldTrees1]

Time to restate the accurate math: (yes, this does compare to AC)(those that don't read the math ...)

4 attacks + 1 bonus attack(3 different triggers)
I am presuming 20th level Fighters with the corresponding fighting style, feats(Dual Wielder, Great Weapon Master, GWM + Polearm Mastery), and weapon(1d8, 2d6, 1d10/1d4)

Obviously first comparing without the -5/+10 and then with the -5/+10 if necessary

1) Weapon damage on a hit
Great Weapon Style rerolls 1s and 2s once
1d8+5 -> 9.5
1d4+5 GWS -> (2.5+2.5+3+4)/4 +5 -> 8
2d6+5 GWS -> 2*(3.5+3.5+3+4+5+6)/6 +5 -> 13 + 1/3
1d10+5 GWS -> (5.5+5.5+3thru10)/10 +5 -> 11.3

2) Great Weapon Master bonus attack permission chance
GWM permits a bonus attack if you got a crit(reducing to 0 is not included in this math) chance of permission varies with advantage:
Chance of 1+ crit = 1-(19/20)^4 = 18.5494% (no rounding)
Adv: Chance of 1+ crit = 1-(19/20)^8 = 33.658% (no rounding)
Dis: Chance of 1+ crit = 1-(399/400)^4 = 0.9963% (no rounding)
So GWM gets 4.185494, 4.33658(Adv), or 4.009963(Dis) attacks to the DW's 5 attacks. PAM obviously get the haft attack for rounds they don't get this attack (0.814506, 0.66342, 0.990037).

3) Summary so far:
Normal:
5 attacks, +11 atk, 9.5 dam, 20/x2
4.185494 attacks, +11 atk, 13+1/3 dam, 20/x2
4.185494 attacks, +11 atk, 11.3 dam, 20/x2 +0.814506 attacks, +11 atk, 8 dam, 20/x2

Adv
5 attacks, +11 atk, 9.5 dam, 20/x2
4.33658 attacks, +11 atk, 13+1/3 dam, 20/x2
4.33658 attacks, +11 atk, 11.3 dam, 20/x2 +0.66342 attacks, +11 atk, 8 dam, 20/x2

Dis
5 attacks, +11 atk, 9.5 dam, 20/x2
4.009963 attacks, +11 atk, 13+1/3 dam, 20/x2
4.009963 attacks, +11 atk, 11.3 dam, 20/x2 +0.990037 attacks, +11 atk, 8 dam, 20/x2

4)Without -5/+10 (here AC is included by leaving accuracy as a variable)
DPR = Sum for each attack (Probability to Hit * Damage per Hit + Probability to Crit * Damage per Hit)
This is equivalent to and the simplified form of "Sum for each attack ( (Probability to Hit - Probability to Crit) * Damage per Hit + Probability to Crit * 2 * Damage per Hit )"
With only 1 unknown(Probability to Hit) we can identify the points of equality also known as the breakpoints where both are equal

Normal:
DW DPR = 5 * (9.5X + 0.475) = 47.5X + 2.375
GWM DPR = 4.185494 * (40X/3 + 2/3) = 55.806586666X + 2.7903293333
PAM GWM DPR = 4.185494 * (11.3X + 0.565) + 0.814506 (8X + 0.4) = 53.8121302X + 2.69060651
As we can see in normal non -5/+10 case the DPR for DW is strictly less than for either GWM style. However the difference is smaller the lower the probability to hit

Adv:
DW DPR = 47.5X + 4.63125
GWM DPR = 57.82106666X + 5.637554
PAM GWMDPR = 54.310714X + 5.295294615
Again the pattern is that DW is strictly less.

Dis:
DW DPR = 47.5X + 0.11875
GWM DPR = 53.466173333X + 0.13366543333
PAM GWMDPR = 53.2328779X + 0.133082195
Again the pattern is that DW is strictly less.
Since there is no point at which DW eclipses GWM(without -5/+10), there is no point at which DW eclipses GWM(that can choose to -5/+10 or not). I am slightly surprised that PAM GWM(without -5/+10) is strictly inferior DPR to GWM(without -5/+10).

[PoeticDwarf]

I did not fail to account for that - it's in my spoiler notes. DW is dealing about 47.5 damage for its five attacks, which is within a handful of points of GWM.

This means that they're relatively comparable - at the end game they're only a few points part from each other. Which is not nearly as bad as you're making it out to be.

In early game, DW is ahead of GWM. Somewhere in mid game they cross, and they're never that far apart. And that's if you don't use the -5/+10. If you do, then they're balanced right around the 50% chance to hit mark.

Does this include PAM? And in 5e most monsters have low AC anyways

[RulesJD]

Shouldn't the 1d12s be 1d10? Unless the Paladin is mounted and using lances, you can't dual wield a 1d12 weapon.

What 1h weapon are you using that does 1d10? It should be 1d8.

[Rysto]

What 1h weapon are you using that does 1d10? It should be 1d8.

Yeah, that was completely wrong. Not sure what I was thinking of.

[mgshamster]

Does this include PAM? And in 5e most monsters have low AC anyways

I did a quick PAM analysis later in the thread. Not detailed.

It seems the biggest critique of this analysis is that it's unlikely to have to roll an 11+ to hit bad guys in this edition. It seems as if most are calculating that it's normal to need a 7+ or 8+ for level appropriate encounters.

Just by raising the AC by 4-5 points across the board (making it around 40-50% chance to hit) you can reduce a lot of the imbalance of GWM (-5/+10). Doesn't help with GWM (no -5/+10). However, giving DW an additional attack on the Bonus Action will help that. Or use Kryx' Rend idea. If this also drops the balance point to 7-8+ on the d20, that will match where people claim the average AC is, which will balance out DW with GWM.

Another idea is to drop weapon damage; for example, the majority of the difference between duelist and GWM (no -5/+10) comes from the damage dice. To fix this, give duelist +4 damage instead of +2. That will balance them in the same way that giving DW an extra attack on the bonus will.

[Kryx]

How do you figure that?
At 20 as per my spreadsheet:


Paladin GWM
Normal hit chance is 65%. Greatsword is 8.33 + 5 + 4.5 = 17.83*.65 = 12.2 with crit. Twice is 24.5 DPR. Add on 7.3 for smite. Overall crit chance is 10% with an 11% chance to kill (total damage vs average HP) = 19% chance of GWM (ignore the cross over of crit and kill). GWM does 2.4, Opportunity attack does 1.5. 24.5+7.3+2.4+1.5 = 35.7 DPR

Paladin S&B + dueling
Normal hit chance is 65%. Adv hit chance is 88%. With shield master happening 48% of the time that is on average throughout the day 76% chance to hit [.48*.88 + (1-.48)*.65]. Crit chance 7% done via the same method. Each longsword is 4.5(longsword) + 5(str) + 4.5(improved divine smite) + 2(dueling) = 16*.76 = 12.1 add in crit to get 12.8. Two attacks is 25.6 DPR. Opportunity attack chance of 12.5% standard adds on 1.4 damage. All spells used toward smite adds on average 7.6 damage. 25.6+1.4+7.6 = 34.6 DPR
Dueling adding 2 damage is pretty silly..

Paladin TWF
Normal hit chance is 65%. Each longsword is 4.5 + 5 + 4.5 = 14*.65=9.6 after crit. 9.6*2 = 19.1. Add on 4.4 for the offhand bonus. 1.2 from opportunity attack. 7.6 from Smite = 32.3 DPR

The math is all available on my spreadsheet. I skipped some of the smaller details here as you can see the full details there, but this should give an overall impression.



Maybe you should look at his DPR of Classes doc.
People don't like to look at spreadsheets. :(
Side note: Normal paladin doesn't have OoV advantage. With that calculated in TWF does just fine.

[mgshamster]

32.3 - 35.7 is within a hand full of points from each other. Less than 5 points. That's a very small amount. I think reflecting that as a percentage (87% of the damage of the other) is a bit deceiving, it feels like we're hiding small absolute differences behind large relative percentages.

It's one of the reasons why I do not like reporting information with percentages unless I absolutely have to.

[Kryx]

32.3 - 35.7 is within a hand full of points from each other. Less than 5 points. That's a very small amount. I think reflecting that as a percentage (87% of the damage of the other) is a bit deceiving, it feels like we're hiding small absolute differences behind large relative percentages.
Up to 5DPR is a huge range in 5e. 32.3 vs 35.7 looks small, but that's 3.4 DPR. More than 10% is huge.

When comparing a Fighter build your metric would likely change from "less than 7 points" or something similar. You'd essentially be using percentages then too.
Percentages are the only meaningful way to compare them.

If you don't think percentages are sufficient then lets use total damage during a day.
35.7*25.7 = 917.49
32.3*25.7 = 830.11

Now you're looking at 87 less damage in a day. That is significant. It's all the same level of inequality, just different ways of measuring.

[mgshamster]

Up to 5DPR is a huge range in 5e. Percentages is the only meaningful way to compare them. 32.3 vs 35.7 looks small, but that's 3.4 DPR. More than 10% is huge.

When comparing a Fighter build your metric then changes from "less than 10 points" or something similar. You'd essentially be using percentages then too.

If you don't think percentages are sufficient then lets use total damage during a day.

35.7*25.7 = 917.49
32.3*25.7 = 830.11

Now you're looking at 87 less damage in a day. That is significant. It's all the same level of inequality, just different ways of measuring.

Ok. Fair point.

If you don't like small numbers, just add a multiplier. The difference is the same.

[Gtdead]

Right. So when you choose a situation where the -5 to hit has no impact, GWM does more damage.

You realize there's a point where DW reaches 13 DPR while GWM still has 5.8 DPR?

That's an odd response. There is a point where GWM does 100 dpr and dw does 50 too. My point (as I mentioned in the post you quoted) was that optimization needs a target AC/Save. If we break the confines suggested by the books and calculate against 35 ac, then what's stopping us from calculating against 45 AC? or 55?

[Skylivedk]

Up to 5DPR is a huge range in 5e. 32.3 vs 35.7 looks small, but that's 3.4 DPR. More than 10% is huge.

When comparing a Fighter build your metric would likely change from "less than 7 points" or something similar. You'd essentially be using percentages then too.
Percentages are the only meaningful way to compare them.

If you don't think percentages are sufficient then lets use total damage during a day.
35.7*25.7 = 917.49
32.3*25.7 = 830.11

Now you're looking at 87 less damage in a day. That is significant. It's all the same level of inequality, just different ways of measuring.

Thank you, Kryx. So we're back at DW being mechanically weak in 5e (if rend, or something else isn't added).

[mgshamster]

That's an odd response. There is a point where GWM does 100 dpr and dw does 50 too. My point (as I mentioned in the post you quoted) was that optimization needs a target AC/Save. If we break the confines suggested by the books and calculate against 35 ac, then what's stopping us from calculating against 45 AC? or 55?

It's not an odd response, I think you might be missing what he's trying to say. The supposed balance point for GWM is that it gets a -5 penalty to attack, but when you remove that penalty, you remove the difference in damage.

By going to the ultra high ACs (31+), there is no difference between a +6 to attack and a +11 to attack. You've removed the -5 penalty from the equation, which removes the balance point. Now you're just looking at who does more flat damage when comparing identical attack modifiers, and it's obviously the one who's going to have a higher damage die and a higher damage modifier. 2d6+15 > 1d8+5.

My analysis was trying to determine at what point they are balanced, and based on yours and others analysis of common ACs, that balance point is above what's to be expected for ACs in the game. How far above does it go is a valid question, and the answer to that is "around AC 29-30."

This makes me wonder about a different house rule: how do they compare when -5/+10 is just someone anyone can do at any time, no feat required?

The answer, I suspect is the same as simply not choosing to use it, which means that giving DW an extra attack and Duelist an extra +2 to damage (maybe a feat for one-handed weapons, and give it some ther goodies?) will balance them out with GWM. But then you might not even have to give DW an extra attack, because the standard Bonus Action attack may be enough to make up for the difference when they're getting an extra 10 damage on that last hit. I'm not exactly sure - I'd have to run the numbers.

[EvilAnagram]

It's not an odd response, I think you might be missing what he's trying to say. The supposed balance point for GWM is that it gets a -5 penalty to attack, but when you remove that penalty, you remove the difference in damage.

By going to the ultra high ACs (31+), there is no difference between a +6 to attack and a +11 to attack. You've removed the -5 penalty from the equation, which removes the balance point. Now you're just looking at who does more flat damage when comparing identical attack modifiers, and it's obviously the one who's going to have a higher damage die and a higher damage modifier. 2d6+15 > 1d8+5.

My analysis was trying to determine at what point they are balanced, and based on yours and others analysis of common ACs, that balance point is above what's to be expected for ACs in the game. How far above does it go is a valid question, and the answer to that is "around AC 29-30."

This makes me wonder about a different house rule: how do they compare when -5/+10 is just someone anyone can do at any time, no feat required?

The answer, I suspect is the same as simply not choosing to use it, which means that giving DW an extra attack and Duelist an extra +2 to damage (maybe a feat for one-handed weapons, and give it some ther goodies?) will balance them out with GWM. But then you might not even have to give DW an extra attack, because the standard Bonus Action attack may be enough to make up for the difference when they're getting an extra 10 damage on that last hit. I'm not exactly sure - I'd have to run the numbers.

Sorry, I might not be keeping up, but I thought the balance point was around 22 AC?

[mgshamster]

Sorry, I might not be keeping up, but I thought the balance point was around 22 AC?

It does. That's the point where DW is dealing more damage than GWM (-5/+10). If you keep increasing the AC, where does it stop dealing more damage? Around AC 30.

Of course, that goes away when you remove the -5/+10 from GWM. Adding in an extra attack on DW's bonus action balances them back in.

[Gtdead]

It's not an odd response, I think you might be missing what he's trying to say. The supposed balance point for GWM is that it gets a -5 penalty to attack, but when you remove that penalty, you remove the difference in damage.

..

I understand that -5 is the balance point. My original response was about the irrelevancy of bless. If the AC is ever increasing then yes, DW will outperform GWM till the last number before infinite AC, whatever that is under the particular 5e system. But if we assume that the highest AC we are going to see is the one suggested by the books, then there are enough common buffs/ expected equipement that will push the hit chance to levels where GWM still outperforms DW. I've created the spreadsheets and I know where the breakpoints are. They favor GWM. Not by much against the highest of ACs mind you, but they do.

Your point of DW not being that awful as people think is not lost on me. My argument is against the superiority of it against high CR monsters.

[mgshamster]

I understand that -5 is the balance point. My original response was about the irrelevancy of bless. If the AC is ever increasing then yes, DW will outperform GWM till the last number before infinite AC, whatever that is under the particular 5e system. But if we assume that the highest AC we are going to see is the one suggested by the books, then there are enough common buffs/ expected equipement that will push the hit chance to levels where GWM still outperforms DW. I've created the spreadsheets and I know where the breakpoints are. They favor GWM. Not by much against the highest of ACs mind you, but they do.

Your point of DW not being that awful as people think is not lost on me. My argument is against the superiority of it against high CR monsters.

I'm in agreement with you. I'm a bit saddened that the AC caps out so low, so it'll be a potential house rule for me until new monsters come out.

What did you think of the house rule for just letting anyone "power attack" whenever they like? Unbind it from the feats. Would that be enough to bring DW in line with GWM? Or would an extra attack on the bonus action still be needed for DW?

I probably won't be able to run these numbers until later tonight.

[Gtdead]

I'm in agreement with you. I'm a bit saddened that the AC caps out so low, so it'll be a potential house rule for me until new monsters come out.

What did you think of the house rule for just letting anyone "power attack" whenever they like? Unbind it from the feats. Would that be enough to bring DW in line with GWM? Or would an extra attack on the bonus action still be needed for DW? Also make sure to add it after lvl 5.

I probably won't be able to run these numbers until later tonight.

Letting fighter and ranger use power attack on DW is ok (you should probably remove the +1 AC on DW for the power attacking round), for the sole reason that if they decide to go DW, they lose their main feature (taking advantage of GWM)
Applying this change universally will break some things.

It's is shaky territory because if it wasn't for GWM/SS, martials could never hope to surpass casters in dealing damage. So essentially, it's a pretty important thing to them.
See how various class" raw output compare without GWM.

A lvl 20 Fighter with Greatsword deals 4*(8.7+5)=54.8
A lvl 20 Oathbreaker/Devotion with PAM deals 2*(6.3+1d8+10)+(3+1d8+10)=59.1
A lvl 20 Hunter with Mark deals 3*(3.5+5+3.5)+1d8+5=45.5
A lvl 20 Rogue with Haste and Booming Blade (1d8+5+3d8)+(1d8+5)+(1d8)+10d6=72
A lvl 17 Sorcerer by quickening a firebolt deals 2*(4d10+5)=54
A lvl 17 Sorcerer by quickening GFB deals 2*(1d6+10+3d8)=54
A lvl 16 Bladesinger with haste deals 3*(1d8+10)+(1d8+5)=53
A lvl 17 Arcana Cleric with BB,SG,SW deals (1d8+3d8+10)+(1d8+5)+3d8=51

Do you really want to add extra bonus attacks to DW and letting everyone power attack? Bards, Clerics, Wizards and Rogues will see a power spike.
Casters will never need to fireball again (not that they need now tbh but anyway). Lvl 1 fighter dips will become very strong.
Monks will never need to flurry again, and will be able to power attack after stunning strike for huge damage.

[mer.c]

I agree that nerfing GWM is pretty harsh on martials. It seems like just helping DW play to its strength would be a better way of balancing it. What would it look like to allow multiatrackers to get a second off-hand attack at 11, while decoupling one or both off-hand attacks from the bonus action?

Or possibly make it a feat to do that, to require some additional investment and line up with PAM+GWM.

[mgshamster]

I agree that nerfing GWM is pretty harsh on martials. It seems like just helping DW play to its strength would be a better way of balancing it. What would it look like to allow multiatrackers to get a second off-hand attack at 11, while decoupling one or both off-hand attacks from the bonus action?

Or possibly make it a feat to do that, to require some additional investment and line up with PAM+GWM.

It effectively makes DW have the same damage output as GWM when they're not using the -5/+10, and it shifts the balance point down a few ACs for when they are using -5/+10.

[Gtdead]

DW actually competes with the other melee styles if we forget about GWM for a moment. At lvl 11, it's just a 4 dpr difference with PAM (although PAM makes for easier AoOs), but you gain a +1 AC, which is better than it looks. If you add a +1 dmg/attack on THF, in addition to it's current bonus, it will be the same.

[Giant2005]

People don't like to look at spreadsheets. :(

I don't mind looking at spreadsheets, it is just that yours is trying to do too much and doesn't have a lot of explanations involved. Without explanations, it is really hard to figure out where all of those numbers are coming from; and without being able to figure out where those numbers are coming from, it is of very limited use to me.

Now that you offered a little bit of an explanation, there are some things you included that I don't much agree with. The first is the choice to have a level 20 character go up against AC 18 - you have to look as low as CR 16 to find a creature with an AC of 18, and there aren't any creatures of CR 18 or higher that have an AC that low. It just seems like a really dubious choice.
The other aspect is including limited use, unreliable abilities like Smite, Opportunity Attacks, or the assumption that you will kill something. Variables like those are far too heavily dependent on individual tables and even individual sessions at those tables to ever be a reliable consideration that could be included on a universal basis.
I for one have only ever smited to my maximum capacity in a day once, and that is only because I was about to die. I went balls-to-the-wall on my last harrah because there was no point in conserving any kind of resource for future encounters. For me, OAs are so rare that the amount of damage you included from them could only happen in a dream. Without having something like Polearm Master to gain OAs more reliably, I'd consider myself lucky to have one OA triggered every 30 encounters. As for killing things, killing something before the battle is over isn't even something that can potentially happen against a BBEG, when counting every last point of DPR you can muster actually matters. Even then, it relies on having lots of low hit point enemies thrown at you, which I am sure is possibly quite common in a lot of games, but most DMs would rather use a couple of stronger monsters rather than waste their time running an army of little guys.

[Giant2005]

Thank you, Kryx. So we're back at DW being mechanically weak in 5e (if rend, or something else isn't added).

If you came to that conclusion, you would be wrong.
I don't agree with the conditions that Kryx used to come to those numbers, but you obviously do so we will consider them as reasonable. Even with that in mind, the DW guy is not only as comparably effective as a GW guy, but it tends to be superior.
The thing you are failing to consider is that he has +2 AC over the GW guy.
Against a Marilyth (the highest CR creature that qualifies for the AC 18 Kryx was using in his calculations), the GW guy will take an incoming 59.05 DPR. The DW guy only takes 49.75.
The GW guy might deal 10.5% more damage under those conditions, but he also receives 18.7% more damage than the DW guy. The DW guy is in a much stronger position to survive the encounter (although neither is likely to without help).

[Skylivedk]

If you came to that conclusion, you would be wrong.
I don't agree with the conditions that Kryx used to come to those numbers, but you obviously do so we will consider them as reasonable. Even with that in mind, the DW guy is not only as comparably effective as a GW guy, but it tends to be superior.
The thing you are failing to consider is that he has +2 AC over the GW guy.
Against a Marilyth (the highest CR creature that qualifies for the AC 18 Kryx was using in his calculations), the GW guy will take an incoming 59.05 DPR. The DW guy only takes 49.75.
The GW guy might deal 10.5% more damage under those conditions, but he also receives 18.7% more damage than the DW guy. The DW guy is in a much stronger position to survive the encounter (although neither is likely to without help).

Where do you get +2ac from? I presume you mean 1ac.

Also - the DW guy will have to attune to two weapons whereas GWM can use that slot for something else.

Also it's not 10% more, but 20% more in the numbers you mention. That's huge. That's also affecting how much damage you take (which would drop with almost the same %cause you'd kill that much faster).

As a DM, I've included rend in dualwielder and also granted rend to extra attack classes (and the two would stack).

I'm also more interested in game balance around level 11 that at level 20 where I don't think I'd have sessions

[Giant2005]

Where do you get +2ac from? I presume you mean 1ac.
The GWF guy has the GWF fighting style and the GWM feat, the DW guy has the Defense fighting style (it is the only one he has access to that is of any use to him) and the DW feat (which gives another +1). The difference in AC between them is +2.


Also - the DW guy will have to attune to two weapons whereas GWM can use that slot for something else.
Most magic weapons don't require attunement - the only one of note that I can think of that does require attunement is the Flame Tongue. Even if the DW guy uses one Flame Tongue and one magic weapon that doesn't require attunement, he is still getting the same +2d6 damage on two hits as the GWM guy. All it would do is lower the relative difference in their respective DPRs, which is something that actually favors the DW guy.


Also it's not 10% more, but 20% more in the numbers you mention. That's huge. That's also affecting how much damage you take (which would drop with almost the same %cause you'd kill that much faster).
Krix's numbers were 35.7 and 32.3, that is a 10.5% difference in damage. As for it affecting how much damage you take, that can be true, but often isn't. I just made a post in another thread here (http://www.giantitp.com/forums/showsinglepost.php?p=20695267&postcount=18) that is relevant. It compares a Warlock with more AC via Moderately Armored vs one with more damage via Agonizing Blast which is a direct parallel to a Paladin with more AC via defense/dual-wielding vs one with more damage via GWM (also I still debate the conditions required to secure more damage). In the Warlock thread, I showed that even while considering the difference in time to kill, the guy with the AC walked away with less hit points lost.


I'm also more interested in game balance around level 11 that at level 20 where I don't think I'd have sessions
This is the game balance at level 11. Sure the numbers were taken as a level 20, but the Paladin doesn't gain any more AC nor damage in the levels between 12 and 20. The ratios will be exactly the same.

[Skylivedk]

The GWF guy has the GWF fighting style and the GWM feat, the DW guy has the Defense fighting style (it is the only one he has access to that is of any use to him) and the DW feat (which gives another +1). The difference in AC between them is +2.

Most magic weapons don't require attunement - the only one of note that I can think of that does require attunement is the Flame Tongue. Even if the DW guy uses one Flame Tongue and one magic weapon that doesn't require attunement, he is still getting the same +2d6 damage on two hits as the GWM guy. All it would do is lower the relative difference in their respective DPRs, which is something that actually favors the DW guy.

Krix's numbers were 35.7 and 32.3, that is a 10.5% difference in damage. As for it affecting how much damage you take, that can be true, but often isn't. I just made a post in another thread here (http://www.giantitp.com/forums/showsinglepost.php?p=20695267&postcount=18) that is relevant. It compares a Warlock with more AC via Moderately Armored vs one with more damage via Agonizing Blast which is a direct parallel to a Paladin with more AC via defense/dual-wielding vs one with more damage via GWM (also I still debate the conditions required to secure more damage). In the Warlock thread, I showed that even while considering the difference in time to kill, the guy with the AC walked away with less hit points lost.

This is the game balance at level 11. Sure the numbers were taken as a level 20, but the Paladin doesn't gain any more AC nor damage in the levels between 12 and 20. The ratios will be exactly the same.

I think Kryx bases his math on the DW having TWF. Without TWF the DPR difference would be bigger.

Also losing your bonus action to deal less damage is rather painful.

With lower levels come lower AC which makes -5/+10 better, hence my mention of level 11. PAM strikes me as being strictly better for palas - the feat requirement is the same, you get more attacks and you only need one weapon plus you have reach.

[Saggo]

Now that you offered a little bit of an explanation, there are some things you included that I don't much agree with. The first is the choice to have a level 20 character go up against AC 18 - you have to look as low as CR 16 to find a creature with an AC of 18, and there aren't any creatures of CR 18 or higher that have an AC that low. It just seems like a really dubious choice.
It's AC 19 for levels 17-20.


The other aspect is including limited use, unreliable abilities like Smite, Opportunity Attacks, or the assumption that you will kill something. Variables like those are far too heavily dependent on individual tables and even individual sessions at those tables to ever be a reliable consideration that could be included on a universal basis.

It's dishonest to not include them. Any comparison over an adventuring day without accounting for resources is just as skewed as using an isolated round if you're trying to be accurate or relevant with your data.

You're right that any given table will be different, and since we can't poll everyone, you're limited to taking a certain amount of abstraction unless you're applying the comparison to a specific situation and context. But it would be wrong to not account for it at all.

[djreynolds]

TWF isn't bad. GWM is nice. PAM is really good, because you can use GWM with it and it requires no duel wield fighting style to get your strength bonus on the off hand strike.

I think we have to look from another perspective.

A dex based TWF, without multiclassing, has the ability to wield a ranged weapon with real skill and can effectively switch hit.

Without multiclassing, most strength based melee types are stuck with a javelin or hand axe which doesn't have much range and cannot benefit from GWM.

The dex based fighter using TWF can still select sharpshooter for his ranged attacks.

PAM is a powerful feat. But many players have success dueling war hammers and long swords, which should be more plentiful in regards to being magical.

So really the only guy losing out is that strength based duel wielder because he cannot use GWM or GWS with duel long swords and unlike the PAM fighter, he must get the duel wielder fighting style and the duel wield feat.

But in the end, players make choices that are not as much combat or optimized related but are concept related. Dueling battle axes is cool.

[Kryx]

The first is the choice to have a level 20 character go up against AC 18 - you have to look as low as CR 16 to find a creature with an AC of 18, and there aren't any creatures of CR 18 or higher that have an AC that low. It just seems like a really dubious choice.
The Average CR of enemies at level 20 is 19. CR 19 creatures have 19 AC as per the DMG and based on averages from the actual monsters in the monster manual.
As you have previously stated your group consistently fights things that are 2 CRs higher than your group. To do so there is no way your group is fighting 6-8 encounters a day or the recommended XP budgets. Your group simply does not follow the DMG recommendations. The balance of the game, and my sheet, is based on those recommendations.


The other aspect is including limited use, unreliable abilities like Smite, Opportunity Attacks, or the assumption that you will kill something. Variables like those are far too heavily dependent on individual tables and even individual sessions at those tables to ever be a reliable consideration that could be included on a universal basis.
I calculated the mathematical probability of it occurring. The same as people do for crits occurring 5% of the time. For an overall view of an adventuring day (which the game is balanced on) this will be very accurate. For an individual round (which the game isn't balanced on) not so much.


I for one have only ever smited to my maximum capacity in a day once, and that is only because I was about to die. I went balls-to-the-wall on my last harrah because there was no point in conserving any kind of resource for future encounters.
I actually assume that Paladins only use 70% of their spells on smite. The other 30% used on utility. In my experience my Paladins have used nearly 95% on smite as it's extremely efficient.


For me, OAs are so rare that the amount of damage you included from them could only happen in a dream. Without having something like Polearm Master to gain OAs more reliably, I'd consider myself lucky to have one OA triggered every 30 encounters.
I can easily adjust OA's base level to be lower. I am in agreement that 12.5% of the time (1.25 rounds out of 10) is too high. 5% is likely more accurate for a baseline (1 out of every 20 rounds). I have made this adjustment


As for killing things, killing something before the battle is over isn't even something that can potentially happen against a BBEG, when counting every last point of DPR you can muster actually matters. Even then, it relies on having lots of low hit point enemies thrown at you, which I am sure is possibly quite common in a lot of games, but most DMs would rather use a couple of stronger monsters rather than waste their time running an army of little guys.
Everyone builds encounters differently. The game, however, is balanced on their recommendations. If DMs only include very strong monsters in every encounter then features that expect otherwise (GWM) obviously lose value. But again, that's your group differing from the game's expectations.


If you came to that conclusion, you would be wrong.
I don't agree with the conditions that Kryx used to come to those numbers, but you obviously do so we will consider them as reasonable. Even with that in mind, the DW guy is not only as comparably effective as a GW guy, but it tends to be superior.
TWF has been shown to be significantly behind in damage terms since 5e came out. I am not the only one who has shown the math for this.


The thing you are failing to consider is that he has +2 AC over the GW guy.
18 AC from plate vs 12+5+1=18 from Dex at level 12+ using the normal rules without vuman. 2 ability score bumps and a feat vs flat 18 ac from gold which is attainable at level 5. Definitely not in the TWF's favor.
Dex has some advantages and Strength has some advantages. AC is not in favor of Dex.


The GWF guy has the GWF fighting style and the GWM feat, the DW guy has the Defense fighting style (it is the only one he has access to that is of any use to him) and the DW feat (which gives another +1). The difference in AC between them is +2.
Defensive? If you're going to create comparisons then compare offensive vs offensive and defensive vs defensive. This is very misleading.


Against a Marilyth (the highest CR creature that qualifies for the AC 18 Kryx was using in his calculations), the GW guy will take an incoming 59.05 DPR. The DW guy only takes 49.75.
I'm not using AC 18 as the max. Please see the "S" tab.
Those DPR numbers likely don't include sources for advantage (which is what makes GWM amazing).


Krix's numbers were 35.7 and 32.3, that is a 10.5% difference in damage.
Those numbers were for a Paladin which is 83% of the GWM. Fighter is 85%. Barbarian is 76%. None were 10.5% difference. Please do not misrepresent the data.






It's dishonest to not include them. Any comparison over an adventuring day without accounting for resources is just as skewed as using an isolated round if you're trying to be accurate or relevant with your data.
Full agreement. These features are often the ones left out of forum math.


You're right that any given table will be different, and since we can't poll everyone, you're limited to taking a certain amount of abstraction unless you're applying the comparison to a specific situation and context. But it would be wrong to not account for it at all.
Yup, there are many ways to model the balance data (burst damage, certain situations, etc). The way the game is balanced is over an adventuring day where those things average out. Some encounters will be more favorable, others less.

[Giant2005]

I think Kryx bases his math on the DW having TWF. Without TWF the DPR difference would be bigger.
TWF wasn't included in the calculations for obvious reasons (Paladins can't take that fighting style).


Also losing your bonus action to deal less damage is rather painful.
It is only less damage if you are facing an enemy that has relatively low AC and hit points for a character of your level.


With lower levels come lower AC which makes -5/+10 better, hence my mention of level 11.
Kryx's calculation was made against a fairly low AC of 19. The average AC of a CR 11 creature is 17, so you are right in saying that would make a difference, however your proficiency bonus is also 2 points lower, so the net result is exactly the same.


PAM strikes me as being strictly better for palas - the feat requirement is the same, you get more attacks and you only need one weapon plus you have reach.
It is. PAM is simply amazing (a little too amazing imo), although I prefer to use it with Staff+Shield rather than a reach weapon, so you get the best of both worlds. It kind of makes me hate my own characters though - the image I get in my mind when my character uses a staff and shield isn't a particularly impressive one.


It's AC 19 for levels 17-20.
You are right. Either way, that is clearly below the average as only 1 of the CR 20 enemies has an AC that low, and that one enemy isn't even a real CR 20 - it is a CR 18 enemy that has some lair bonuses (none of which effect its AC).

It's dishonest to not include them. Any comparison over an adventuring day without accounting for resources is just as skewed as using an isolated round if you're trying to be accurate or relevant with your data.

You're right that any given table will be different, and since we can't poll everyone, you're limited to taking a certain amount of abstraction unless you're applying the comparison to a specific situation and context. But it would be wrong to not account for it at all.
Its not dishonest, it is called a baseline. Your baseline is what is quantifiable and is used so you can add the variables to it without starting from scratch every time. If you take a baseline, make estimates of all the variables, and add them to it, it is no longer a baseline. It isn't really much use at all at that point as you can no longer adjust for the variables because static assumptions of those variables have already been crammed right in there. It is a very un-scientific approach to things.

[Kryx]

TWF wasn't included in the calculations for obvious reasons (Paladins can't take that fighting style).
Ah, I now see why you made the ac comment earlier. I houserule it so Paladins can TWF, but for RAW what you said makes some sense. However it's different than you presented.

Level 1: 16 (chainmail) AC vs 15 (12+3dex)
level 5: 18 (Plate) AC vs 17 (12+4dex+1 def)
level 12: 18 (Plate) AC vs 19 (12+4dex+1 def+1feat)


Kryx's calculation was made against a fairly low AC of 19.
Please see the DMG page 274. The maximum AC they use is 19. It starts at CR 17 and continues to 30.
The numbers that Saggo averaged (https://docs.google.com/spreadsheets/d/1-eul_N14RmAYuUjvJW4CsYvPmoCeks3LhbgivSM_n2o/edit#gid=0) also confirm this. Of the 3 monsters at CR 20 they average to 19.6666.

[Giant2005]

The Average CR of enemies at level 20 is 19. CR 19 creatures have 19 AC as per the DMG and based on averages from the actual monsters in the monster manual.
As you have previously stated your group consistently fights things that are 2 CRs higher than your group. To do so there is no way your group is fighting 6-8 encounters a day or the recommended XP budgets. Your group simply does not follow the DMG recommendations. The balance of the game, and my sheet, is based on those recommendations.
Firstly I'd like to apologize to you. I feel that I have been coming off as a little too critical of your sheet when that isn't my intention. It is simply a biproduct of people considering it the end of the discussion which forces me to do so. I'd just like to state that I do both appreciate and respect your work, even if I disagree on some aspects of it.
As for some disagreements, the AC value is most certainly one of them. There are 4 CR 20 monsters in the MM, 3 of them have 20 AC and the 4th has an AC of 19 (and is in fact only a CR 18 creature that has its CR artificially boosted by being in its lair). The average AC of a CR 20 creature should be 20, not 19.


I calculated the mathematical probability of it occurring. The same as people do for crits occurring 5% of the time. For an overall view of an adventuring day (which the game is balanced on) this will be very accurate. For an individual round (which the game isn't balanced on) not so much.
I understand that and I commend you for your efforts. I just don't think it is a useful metric, and including it in the baseline is misleading.
Basically my objection comes from one of functionality. By including a chance to kill something in the baseline, you can no longer factor in actually killing something or not in any given round - you can account for neither scenario, only the average of both. If you don't include that estimation, you can make a simple adjustment in the moment and account for either occurrence in any given moment.
EDIT: I decided I had better clarify my point here a little more by throwing in an example to reinforce my point.
A Fighter gets far more use out of a Flame Tongue than any other class. To account for that advantage for the Fighter, you would need to work out how often one would expect to find a loot cache per level, multiply that by the chance of getting a Flame Tongue in a loot cache, multiply that by the average damage of the Flame Tongue, and then add that value to every attack of every class that uses a sword.
Although that would capture one of the Fighter's advantages on an average level, you simply wouldn't do that because it is including a condition into the baseline that might not even be a factor in any given encounter. Yet that treatment is exactly what you have given these other conditions that might not even be a factor in any given encounter. In doing so, you have made the results much less useful. It is a simple task to factor in the damage of a Flame Tongue to a baseline if one is available, but trying to account for that condition being met is far more problematic if your baseline already includes partial credit for using a Flame Tongue. It essentially prevents you from ever accurately including the existence of, or lack thereof, a Flame Tongue.


Everyone builds encounters differently. The game, however, is balanced on their recommendations. If DMs only include very strong monsters in every encounter then features that expect otherwise (GWM) obviously lose value. But again, that's your group differing from the game's expectations.
The game doesn't make recommendations on what monsters to use - it makes allowances for a large array of possible monster combinations (including an army of weaker monsters). There is a very important distinction there and the truth is, even if it does make allowances for legions of weaker monsters, they are far from practical to use at the table.
I have only my own experiences to confirm such a thing, but I think it is a fair assumption to make that most high level encounters, at most tables will consist of few, stronger foes. Sure many weak foes is a possibility but it is a lot more taxing on the DM and takes a lot more time at the table to complete. Those factors should prevent such encounters from occurring at as higher rate as the stronger foe encounters.


TWF has been shown to be significantly behind in damage terms since 5e came out. I am not the only one who has shown the math for this.
It most certainly is for the fighter due to their attack routine coming from having multiple, weaker attacks. It is also conditionally weaker for the Ranger (those conditions again being lots of weak enemies - Hunter's Mark and TWF do not go well together if you have to keep reapplying Hunter's Mark every round or two). Those that gain damage bonuses rather than extra attacks make far greater use of dual-wielding however.


18 AC from plate vs 12+5+1=18 from Dex at level 12+ using the normal rules without vuman. 2 ability score bumps and a feat vs flat 18 ac from gold which is attainable at level 5. Definitely not in the TWF's favor.
Dex has some advantages and Strength has some advantages. AC is not in favor of Dex.
I wasn't talking about Dex.
I was talking about two guys in Plate - one of which has GWF and GWM, the other has the Defense Fighting Style and Dual Wielder. Dual-Wielding does not require going Dex-based.


Defensive? If you're going to create comparisons then compare offensive vs offensive and defensive vs defensive. This is very misleading.
How is that misleading? The Dual-Wielding Paladin has no other Fighting Style to take. It would be misleading to assume that the Dual-Wielder chose GWF or Dueling instead of choosing the one fighting style that he could actually benefit from.


Those numbers were for a Paladin which is 83% of the GWM. Fighter is 85%. Barbarian is 76%. None were 10.5% difference. Please do not misrepresent the data.

I don't even know what you are talking about anymore, it seems like you are throwing out random percentages at me to confuse me...
What I was referring to was your own calculations below:

Paladin GWM
Normal hit chance is 65%. Greatsword is 8.33 + 5 + 4.5 = 17.83*.65 = 12.2 with crit. Twice is 24.5 DPR. Add on 7.3 for smite. Overall crit chance is 10% with an 11% chance to kill (total damage vs average HP) = 19% chance of GWM (ignore the cross over of crit and kill). GWM does 2.4, Opportunity attack does 1.5. 24.5+7.3+2.4+1.5 = 35.7 DPR

Paladin TWF
Normal hit chance is 65%. Each longsword is 4.5 + 5 + 4.5 = 14*.65=9.6 after crit. 9.6*2 = 19.1. Add on 4.4 for the offhand bonus. 1.2 from opportunity attack. 7.6 from Smite = 32.3 DPR

You had the GWM Paladin coming in at 35.7 DPR and the TWF Paladin at 32.3 DPR, hence the 10.5% difference in their DPR.

Actually, looking at those numbers, I'm not really sure how you came by them. The TWF Paladin's off-hand bonus is incorrect unless you are considering something that I can't think of (that lowers its DPR). The off-hand should be 0.6(4.5+4.5) + 0.05(9+9) = 6.3, yet somehow the listed value is 4.4

[Zalabim]

I hardly see this considered or acknowledged. The Dual Wielder feat is pretty bad, but it's probably mechanically balanced for using two lances from horseback. The two feat combo to compare to GWM and PAM is DW with Mounted Combatant. It's also worth mentioning that some of Kryx's damage numbers require the target to be within the size range to be tripped.

It wouldn't be right to fix DW without addressing Dual Lances.


I am slightly surprised that PAM GWM(without -5/+10) is strictly inferior DPR to GWM(without -5/+10).

PAM does a little less damage, but it makes up for it if you get to make an extra OA from the feat.

[Kryx]

I hardly see this considered or acknowledged. The Dual Wielder feat is pretty bad, but it's probably mechanically balanced for using two lances from horseback. The two feat combo to compare to GWM and PAM is DW with Mounted Combatant.
It isn't considered because it is a gimmicky build that works in a very limited amount of circumstances whereas there are 6+ builds that use TWF without lances.


It's also worth mentioning that some of Kryx's damage numbers require the target to be within the size range to be tripped.
BM and S&B are the only ones tripping. Barbarian, OoV, and Bless do not have this size limit issue.
Regarding the size limit issue: The sheet assumes BM and S&B use trip 55% of the time and the other 45% on riposte. For larger creatures he'd be using riposte more.
It's been accounted for.

[Finieous]

Also optional rule in DMG pg. 271 to climb on big creatures. Basically the same opposed Athletics/Acrobatics check as a shove or grapple. Grants advantage on attacks when tripping doesn't work due to size.

[Giant2005]

Also optional rule in DMG pg. 271 to climb on big creatures. Basically the same opposed Athletics/Acrobatics check as a shove or grapple. Grants advantage on attacks when tripping doesn't work due to size.

That doesn't help the Shield Master guy so much. He can shove as a bonus action but he cannot climb on someone as a bonus action.

[Finieous]

That doesn't help the Shield Master guy so much. He can shove as a bonus action but he cannot climb on someone as a bonus action.

I didn't know we were talking about Shield Master, but anyway, it's just another (optional) way for martials to gain advantage on some percentage of their attacks over the course of a typical adventuring day.

[OldTrees1]

PAM does a little less damage, but it makes up for it if you get to make an extra OA from the feat.

Certainly does (I love PAM more than GWM).
Just, when doing the math I was surprised the "gets 5 attacks & has GWM" style had strictly lower DPR than the "gets 4.01-4.34 attacks & has GWM" style. I suspect that it gets more complicated when they are using the -5/+10 because higher base damage favors the more attacks style.

[Coffee_Dragon]

It wouldn't be right to fix DW without addressing Dual Lances.

Would it be enough to say that dual lances is a bizarre concept that wouldn't happen at any sane table?

[Saggo]

Its not dishonest, it is called a baseline. Your baseline is what is quantifiable and is used so you can add the variables to it without starting from scratch every time. If you take a baseline, make estimates of all the variables, and add them to it, it is no longer a baseline. It isn't really much use at all at that point as you can no longer adjust for the variables because static assumptions of those variables have already been crammed right in there. It is a very un-scientific approach to things.

Ignoring Kryx's sheet since he's capable of defending himself, this would be true if we couldn't adjust variables. The baseline is always there, and if you want to see it, you set the variables to 0%. That doesn't tell us anything relevant for the most part, since comparing a Paladin who never Smites to a Fighter who never Action Surges, for a quick example, would show us a wildly different picture of the two classes than is realistic or expected.

I like Kryx's work because it created a framework where we can easily adjust those variables.