Great Weapon Master and Sharpshooter are overpowered feats. It's too east to reach +10 to hit with bonus abilities like Precision Attack or Reckless Attack that offset the penalty. This is the main problem. Next time I start a new campaign, I'll be banning both, as well as Pole Weapon Master, Elven Accuracy, Slasher, Piercer and Crusher.
Id prefer not to hard ban any feats if I have to, but I agree GWM and Sharpshooter can be a bit much once you get into Tiers 2 and above. I wish there was an elegant way to split Sharpshooter into two half feats instead (where you can get the full benefit, but it takes longer and you have to invest one more ASI) where each half feat could also still be beneficial on its own.
I think that the ability to not have disadvantage on long range attacks coupled with the ability to ignore most cover is enough to be its own feat.
Tasha's already split up Crossbow Expert for us (apparently the bonus action attack is half of it). I don't think it's that hard to split SS in half, given that -5/+10 on all attacks made with ranged weapons you're proficient with is easily worth a half feat. Just split it into that half and the rest of it, so the other half is ignoring 1/2 and 3/4 cover and long range. That would certainly fit the Tasha's mold. If you made ignoring 1/2 and 3/4 cover a feat on its own no sane person would take it - +2 to some ranged attacks that you can usually fix with mobility and +5 to almost none of them isn't remotely as good as +1 to your Dexterity modifier.
I consider GWM more of a feat tax to allow great weapons to be effective (I consider Archery style good even without help). Absent feats:
Dueling style is clearly superior to great weapon fighting -- the +2 AC is more useful than +1.83 damage per attack.
Dueling style is also clearly superior to defensive style with a two-handed weapon -- now we're looking at +1 AC vs 0.5 damage per attack.
Shields get better at high levels. The damage bonus doesn't (this is partially negated by high levels having more attacks that ignore ac).
Using Dexterity gets you Initiative and good ranged weapons, which is higher value than being able to conveniently wear plate. There are no two-handed finesse weapons.
That said, it would obviously be better if the fighting styles were better balanced vs one another, and the overtuned feats either didn't exist or got reduced to reasonable.
Balancing fighting styles means:
Either all fighting styles are dependent on bonus actions, or none are.
Relative value of fighting styles should not vary with level.
I agree that I was a bit shocked when I first played 5e a few years back and I looked at the Feats list. There was such a wide margin between GWM and SS and things like Keen Mind. Even Savage Attacker is pretty good especially at low levels where you might only be rolling 1-2 damage dice. The Feats seemed SO unbalanced that it bothered me. But then I've played MMOs where there was often an optimized way to do something and many players gravitated that way simply because everyone else seemed to be doing it.
When more than a quarter of your players are creating essentially the same characters with all of the thousands of possible combinations we have, then clearly there is a problem.
I consider GWM more of a feat tax to allow great weapons to be effective (I consider Archery style good even without help). Absent feats:
Dueling style is clearly superior to great weapon fighting -- the +2 AC is more useful than +1.83 damage per attack.
Dueling style is also clearly superior to defensive style with a two-handed weapon -- now we're looking at +1 AC vs 0.5 damage per attack.
Shields get better at high levels. The damage bonus doesn't.
Using Dexterity gets you Initiative and good ranged weapons, which is higher value than being able to conveniently wear plate. There are no two-handed finesse weapons.
That said, it would obviously be better if the fighting styles were better balanced vs one another, and the overtuned feats either didn't exist or got reduced to reasonable.
"Shields get better at high levels. The damage bonus doesn't."
I'm curious about this particular part of your comment. Please 'splain.
I think that if everything 5E-wide were errata'd to provide a damage bonus only once per round (except for ability score modifier and magic weapon static bonuses... but heck, maybe even those too?) the damage curve between optimized and unoptimized characters would SIGNIFIGANTLY flatten, as well as causing rogue to actually pick up steam as a DPR leader. While a lot of conventional optimized builds would lose significant strength, I think overall it would be healthier for the state of the game, and easier for players to parse if ALL bonus damage sources played by the same expectation of "once per round."
"Shields get better at high levels. The damage bonus doesn't."
I'm curious about this particular part of your comment. Please 'splain.
A shield is +2 AC. A shield, +3 is +5 AC. No comparable improvements exist for any of the offensive fighting styles.
Well, countering that would be... to-hit bonuses scale faster than AC can keep up, for all but the most ridiculous of Bladesinger builds. Your defensive shield fighter's 21 AC may indeed go as high as 28 when you're decked out with +3 super-items and a +1 misc AC bonus from a cloak or ring or something (though your squishyer teammates are likely to be hanging out in that 17-20 range, questioning what value you're providing to yourself or your team of making yourself less desirable to target).... but in that time, the attack bonuses of your enemies have jumped up from the +4-6 you were facing at low tiers, to +10-12 (or even higher) against end game foes. Chasing AC doesn't really make you get hit less often, just get hit less more often by high level enemies, which is why most characters don't focus on it as much as offense (and again, if enemies just walk around you, what's the point anyway?).
Meanwhile, bonus damage is bonus damage, you have to deal X damage to clear the encounter, and every bonus towards that total is helpful no matter what.
Haven't they said they don't balance the game with magical items as assumed values? Idk, trying to factor in multiplevery-rare and legendary items seems outside the assumptions we should make in general.
It is apparently so hard to program Aberrant Mind and Clockwork Soul spell-swapping into dndbeyond they had to remake the game without it rather than implement it.
Believe me, I have wasted a ton of time today trying to re-invent the wheel and come up with Excel tricks to visualize all this stuff... but I'd hazard that if you think about every combat as a race to deal the enemy team's total HP in damage, every +1 to hit bonus or +1 damage bonus is probably more useful for your team than +1 AC bonuses. What the ratio is? Where the tipping points are? How the DPR value of the teammate impacts the value of +1 AC bonuses on them vs. a lower-output teammate? Oof, you could probably spend years crafting this thesis, I really wish some intrepid nerd out there had already done so...
Haven't they said they don't balance the game with magical items as assumed values? Idk, trying to factor in multiplevery-rare and legendary items seems outside the assumptions we should make in general.
They say that, but it's a bit dumb and you don't need very rare or legendary, the +30% (65% vs 50%) for a Shield, +1 is already basically keeping up with its first level performance. The problem would also be fixed if magic armor and magic shield provided non-stacking bonuses.
Believe me, I have wasted a ton of time today trying to re-invent the wheel and come up with Excel tricks to visualize all this stuff... but I'd hazard that if you think about every combat as a race to deal the enemy team's total HP in damage, every +1 to hit bonus or +1 damage bonus is probably more useful for your team than +1 AC bonuses.
A bonus to offense is generally more useful than a bonus to defense, but that doesn't mean +9% offense beats +62% defense.
"Shields get better at high levels. The damage bonus doesn't."
I'm curious about this particular part of your comment. Please 'splain.
A shield is +2 AC. A shield, +3 is +5 AC. No comparable improvements exist for any of the offensive fighting styles.
But that would suggest Defense is better than Dueling, not that Dueling is better than Defense. If you're emphasizing (correctly) that AC has increasing returns and damage doesn't, a sword and board wielder should always grab Defense over Dueling. +1 AC is way better than +2 damage per attack at the amounts of AC you're talking about.
But that would suggest Defense is better than Dueling, not that Dueling is better than Defense. If you're emphasizing (correctly) that AC has increasing returns and damage doesn't, a sword and board wielder should always grab Defense over Dueling. +1 AC is way better than +2 damage per attack at the amounts of AC you're talking about.
I'm not emphasizing AC having increasing returns. I'm emphasizing shields having increasing returns. The +14% toughness from reducing hit chance from 40 to 35 is not better than the 16% damage increase from going from 1d8+8 to 1d8+10.
But that would suggest Defense is better than Dueling, not that Dueling is better than Defense. If you're emphasizing (correctly) that AC has increasing returns and damage doesn't, a sword and board wielder should always grab Defense over Dueling. +1 AC is way better than +2 damage per attack at the amounts of AC you're talking about.
I'm not emphasizing AC having increasing returns. I'm emphasizing shields having increasing returns. The +14% toughness from reducing hit chance from 40 to 35 is not better than the 16% damage increase from going from 1d8+8 to 1d8+10.
That's ignoring your own math! Lemme go grab the math you did and present it in-context for both styles.
Well, countering that would be... to-hit bonuses scale faster than AC can keep up, for all but the most ridiculous of Bladesinger builds.
True but still doesn't eliminate the relative value.
At level 1, you're looking at going from ac 17 to 19, vs +4 attack bonus (hit chance 40% vs 30% = +33%)
At level 20, you're looking at going from ac 22 to 27, vs +14 attack bonus (hit chance 65% vs 40% = +62%)
Vs, say, great weapon fighting
At level 1 (16 stat), you're going from 10 to 11.33 (+13%)
At level 20 (20 stat, +3 weapon) you're going from 15 to 16.33 (+8.9%)
Now, I don't know where AC 22 is coming from, and +4 attack bonus contradicts the standard assumptions from the DMG, so I'm just going to be explicit about what constants I use, so you can change to your own constants as you like:
Level 1 (assumed base AC 17 in armor, +3 in attack stat, wielding a rapier), target is assumed to have +3 to hit you and to be AC 13:
Sword and Board Dude: Damage out is 5.1, after rolling to hit and such. +2 damage per hit would be +39.22%.
AC is 19, and so is hit 30% of the time (25% for real, but being critted is worth between +0 (if no incoming damage can crit) and +5 (if all incoming damage can crit) percentage points; I'm assuming it can all crit so the math can happen); going to AC 20 would be 25%. In other words, if it takes 10 hits to kill you, you've gone from taking 33.33 attacks to be killed to taking 40 attacks to be killed, or +20% durability.
So at level 1, Dueling is a higher percent increase.
Level 20 (assumed base AC 18 in armor, and all equipment has a +3 bonus, and +5 in attack stat; target is AC 19 and has +10 to hit):
Damage out is now 10.225 base per hit, so +2 damage is worth +19.56% (it's gone down a lot).
AC is 26, which means the math is the same as above - the monster hits you on a 16 at level 1 and at level 20, under these assumptions. So +1 AC remains at +20%.
So now Defense has a greater percent increase.
Shields only have increasing returns because AC has increasing returns, and both Defense and Dueling can use a shield. The higher your AC already is, the more valuable it is to increase your AC by 1. Damage has decreasing returns - the higher your damage per hit, the less valuable +1 damage per hit is.
TWF fix for me is you get to take the extra off hand attack as part of the Attack Action. This leaves your BA open for what ever you need and works extremely well with the Hunter's mark build as you are no longer burning your BA attack to move your mark.
This removes the action economy tax but doesn't address the damage gap. The only way to fix that is to give TWF a damage bonus that scales with your extra attacks, which is why I chose to add a d4 to melee attacks after Extra Attack becomes available.
the other strategy is to allow the TWF to use a reaction to increase their AC.
I added two slightly different items (bucklers and parrying daggers) to my game for this purpose, since it's always bugged me that the game doesn't support actual historical dueling styles like dagger and rapier.
I wish there was an elegant way to split Sharpshooter into two half feats instead...I think that the ability to not have disadvantage on long range attacks coupled with the ability to ignore most cover is enough to be its own feat.
I've settled for replacing the -5/+10 in SS with crits on 19 (avery mild buff, but synergizes well with Elven Accuracy and Piercer) and limiting the other two benefits to once per round, up to your Proficiency Bonus times per short rest. That way it's more of a dramatic moment instead of just constantly ignoring the laws of physics. Even if the -5/+10 weren't broken, it's utterly uninteresting. It's either always the correct move or never the correct move for any given enemy.
I consider GWM more of a feat tax to allow great weapons to be effective...Absent feats: Dueling style is clearly superior to great weapon fighting -- the +2 AC is more useful than +1.83 damage per attack.
One thing people tend to miss is that GWFS is extremely consistent with a greatsword. A longsword with Dueling is equally likely to give you a 1 + 2 as an 8 + 2. A greatsword with GWFS has a 91% chance to give you at least a 6, and a 50% chance to roll 7, 8, or 9. It's especially nice for Action Surge and Sentinel reactions. You pretty much never have a bad turn, and I'd argue that at a subjective level that feels better than raising your already good AC.
Saved this response for last since it'll be the longest and I imagine many people have lost interest in the GWM math.
Okay, I surrender. I guess that "-5 to hit is -25% damage" doesn't actually work, like, at all. We truly live in the worst of all possible worlds, and there is no safe shorthand for GWM math apart from using a DPR calculator spreadsheet for specific ACs and PBs. I am awash in a sea of uncertainty.
You were right in assuming AC is normally irrelevant, but only if you're comparing equal hit rates and there's no conditional bonuses. That doesn't apply here since GWM changes your hit rate with the -5/+10, and the bonus attack depends on your chances of landing at least 1 crit during your turn, which depends on your number of attacks and whether you have advantage. That kind of naive AC-less comparison also doesn't work on Sneak Attack for the same reason.
The thing about GWM is that the lower your hit rate is, the more that -5 penalty hurts you (e.g. going from 30% hit rate to 5% is an 83.3% drop in your damage.) Likewise the higher your base damage, the less of a relative boost that +10 is and the more damage you're throwing away when the -5 costs you a hit. The -5/+10 isn't always a net gain; your hit rate (which depends on the enemy's AC) has to be high enough and your damage can't be too high either. The answer to this rpg.stackexchange question has some graphs that help visualize this.
There's no easy way to know up how much of a benefit or detriment the -5/+10 trade is going to be for a given AC value other than doing the math with and without it and checking both results (though as a rule of thumb, if you have advantage, it's pretty much always a win.) But since the net gain or loss depends on which AC you used, that raises the obvious question of what AC value makes for a fair comparison. If you assume a 10 AC monster, GWM is going to look great. If you assume a 20 AC monster and no advantage, it's probably a net loss, and it wouldn't make sense to use the GWM damage in a comparison because you'd just be using your regular damage.
The most fair comparison I could think of was to do the math for every AC value, only use the GWM result if it's actually a win, and weigh the results against how common each AC number is. The bonus action also needed to be included because if you have 2 attacks and advantage, you've got 4 chances to roll a 20, and if the bonus action does trigger, it can also benefits from GWM. Even if you only wanted to do the math for 15 AC (which seems to be close enough to the weighted average), once you throw in advantage and the crit chance, the formula would've still been pretty unwieldy for a forum post.
Now, I don't know where AC 22 is coming from, and +4 attack bonus contradicts the standard assumptions from the DMG, so I'm just going to be explicit about what constants I use, so you can change to your own constants as you like:
Level 1 (assumed base AC 17 in armor, +3 in attack stat, wielding a rapier), target is assumed to have +3 to hit you and to be AC 13:
Sword and Board Dude: Damage out is 5.1, after rolling to hit and such. +2 damage per hit would be +39.22%.
AC is 19, and so is hit 30% of the time (25% for real, but being critted is worth between +0 (if no incoming damage can crit) and +5 (if all incoming damage can crit) percentage points; I'm assuming it can all crit so the math can happen); going to AC 20 would be 25%. In other words, if it takes 10 hits to kill you, you've gone from taking 33.33 attacks to be killed to taking 40 attacks to be killed, or +20% durability.
So at level 1, Dueling is a higher percent increase.
Level 20 (assumed base AC 18 in armor, and all equipment has a +3 bonus, and +5 in attack stat; target is AC 19 and has +10 to hit):
Damage out is now 10.225 base per hit, so +2 damage is worth +19.56% (it's gone down a lot).
AC is 26, which means the math is the same as above - the monster hits you on a 16 at level 1 and at level 20, under these assumptions. So +1 AC remains at +20%.
So now Defense has a greater percent increase.
Shields only have increasing returns because AC has increasing returns, and both Defense and Dueling can use a shield. The higher your AC already is, the more valuable it is to increase your AC by 1. Damage has decreasing returns - the higher your damage per hit, the less valuable +1 damage per hit is.
AC 22 came from a prior example (+4 to hit comes from the actual average of stuff you fight at level 1), and looks to be plate plus defensive fighting style. In any case, your math is suffering from giving a +2 damage even when you miss, and defense has somewhat diminishing returns with level because monster accuracy tends to outpace ac increases.
Now, I don't know where AC 22 is coming from, and +4 attack bonus contradicts the standard assumptions from the DMG, so I'm just going to be explicit about what constants I use, so you can change to your own constants as you like:
Level 1 (assumed base AC 17 in armor, +3 in attack stat, wielding a rapier), target is assumed to have +3 to hit you and to be AC 13:
Sword and Board Dude: Damage out is 5.1, after rolling to hit and such. +2 damage per hit would be +39.22%.
AC is 19, and so is hit 30% of the time (25% for real, but being critted is worth between +0 (if no incoming damage can crit) and +5 (if all incoming damage can crit) percentage points; I'm assuming it can all crit so the math can happen); going to AC 20 would be 25%. In other words, if it takes 10 hits to kill you, you've gone from taking 33.33 attacks to be killed to taking 40 attacks to be killed, or +20% durability.
So at level 1, Dueling is a higher percent increase.
Level 20 (assumed base AC 18 in armor, and all equipment has a +3 bonus, and +5 in attack stat; target is AC 19 and has +10 to hit):
Damage out is now 10.225 base per hit, so +2 damage is worth +19.56% (it's gone down a lot).
AC is 26, which means the math is the same as above - the monster hits you on a 16 at level 1 and at level 20, under these assumptions. So +1 AC remains at +20%.
So now Defense has a greater percent increase.
Shields only have increasing returns because AC has increasing returns, and both Defense and Dueling can use a shield. The higher your AC already is, the more valuable it is to increase your AC by 1. Damage has decreasing returns - the higher your damage per hit, the less valuable +1 damage per hit is.
AC 22 came from a prior example (+4 to hit comes from the actual average of stuff you fight at level 1), and looks to be plate plus defensive fighting style. In any case, your math is suffering from giving a +2 damage even when you miss, and defense has somewhat diminishing returns with level because monster accuracy tends to outpace ac increases.
Doesn't damage suffer from the same diminishing returns as creatures gain HP as the levels increase? Both AC and Damage tend to hit plateaus (such as Fighters at lvl 10 are often doing the same at level 9, but then the third attack at 11 causes a big jump)
Shields only have increasing returns because AC has increasing returns, and both Defense and Dueling can use a shield. The higher your AC already is, the more valuable it is to increase your AC by 1. Damage has decreasing returns - the higher your damage per hit, the less valuable +1 damage per hit is.
Generally true for bonuses to hit too. Going from a 50 to a 55% hitrate is a 10% increase in dpr. But going from a 90 to a 95% hitrate is only a 5.5% increase. TBH its why the GWF and Sharpshooter are over-performing in some situations but not others.
Rollback Post to RevisionRollBack
I'm probably laughing.
It is apparently so hard to program Aberrant Mind and Clockwork Soul spell-swapping into dndbeyond they had to remake the game without it rather than implement it.
To post a comment, please login or register a new account.
Tasha's already split up Crossbow Expert for us (apparently the bonus action attack is half of it). I don't think it's that hard to split SS in half, given that -5/+10 on all attacks made with ranged weapons you're proficient with is easily worth a half feat. Just split it into that half and the rest of it, so the other half is ignoring 1/2 and 3/4 cover and long range. That would certainly fit the Tasha's mold. If you made ignoring 1/2 and 3/4 cover a feat on its own no sane person would take it - +2 to some ranged attacks that you can usually fix with mobility and +5 to almost none of them isn't remotely as good as +1 to your Dexterity modifier.
I consider GWM more of a feat tax to allow great weapons to be effective (I consider Archery style good even without help). Absent feats:
That said, it would obviously be better if the fighting styles were better balanced vs one another, and the overtuned feats either didn't exist or got reduced to reasonable.
Balancing fighting styles means:
I agree that I was a bit shocked when I first played 5e a few years back and I looked at the Feats list. There was such a wide margin between GWM and SS and things like Keen Mind. Even Savage Attacker is pretty good especially at low levels where you might only be rolling 1-2 damage dice. The Feats seemed SO unbalanced that it bothered me. But then I've played MMOs where there was often an optimized way to do something and many players gravitated that way simply because everyone else seemed to be doing it.
When more than a quarter of your players are creating essentially the same characters with all of the thousands of possible combinations we have, then clearly there is a problem.
"Shields get better at high levels. The damage bonus doesn't."
I'm curious about this particular part of your comment. Please 'splain.
I think that if everything 5E-wide were errata'd to provide a damage bonus only once per round (except for ability score modifier and magic weapon static bonuses... but heck, maybe even those too?) the damage curve between optimized and unoptimized characters would SIGNIFIGANTLY flatten, as well as causing rogue to actually pick up steam as a DPR leader. While a lot of conventional optimized builds would lose significant strength, I think overall it would be healthier for the state of the game, and easier for players to parse if ALL bonus damage sources played by the same expectation of "once per round."
dndbeyond.com forum tags
I'm going to make this way harder than it needs to be.
A shield is +2 AC. A shield, +3 is +5 AC. No comparable improvements exist for any of the offensive fighting styles.
Well, countering that would be... to-hit bonuses scale faster than AC can keep up, for all but the most ridiculous of Bladesinger builds. Your defensive shield fighter's 21 AC may indeed go as high as 28 when you're decked out with +3 super-items and a +1 misc AC bonus from a cloak or ring or something (though your squishyer teammates are likely to be hanging out in that 17-20 range, questioning what value you're providing to yourself or your team of making yourself less desirable to target).... but in that time, the attack bonuses of your enemies have jumped up from the +4-6 you were facing at low tiers, to +10-12 (or even higher) against end game foes. Chasing AC doesn't really make you get hit less often, just get hit less more often by high level enemies, which is why most characters don't focus on it as much as offense (and again, if enemies just walk around you, what's the point anyway?).
Meanwhile, bonus damage is bonus damage, you have to deal X damage to clear the encounter, and every bonus towards that total is helpful no matter what.
dndbeyond.com forum tags
I'm going to make this way harder than it needs to be.
True but with Proficiency Bonus the to-hit numbers get better over time even if the gear stays the same. This isn't true for damage numbers.
True but still doesn't eliminate the relative value.
Vs, say, great weapon fighting
Haven't they said they don't balance the game with magical items as assumed values? Idk, trying to factor in multiple very-rare and legendary items seems outside the assumptions we should make in general.
I'm probably laughing.
It is apparently so hard to program Aberrant Mind and Clockwork Soul spell-swapping into dndbeyond they had to remake the game without it rather than implement it.
Believe me, I have wasted a ton of time today trying to re-invent the wheel and come up with Excel tricks to visualize all this stuff... but I'd hazard that if you think about every combat as a race to deal the enemy team's total HP in damage, every +1 to hit bonus or +1 damage bonus is probably more useful for your team than +1 AC bonuses. What the ratio is? Where the tipping points are? How the DPR value of the teammate impacts the value of +1 AC bonuses on them vs. a lower-output teammate? Oof, you could probably spend years crafting this thesis, I really wish some intrepid nerd out there had already done so...
dndbeyond.com forum tags
I'm going to make this way harder than it needs to be.
They say that, but it's a bit dumb and you don't need very rare or legendary, the +30% (65% vs 50%) for a Shield, +1 is already basically keeping up with its first level performance. The problem would also be fixed if magic armor and magic shield provided non-stacking bonuses.
A bonus to offense is generally more useful than a bonus to defense, but that doesn't mean +9% offense beats +62% defense.
But that would suggest Defense is better than Dueling, not that Dueling is better than Defense. If you're emphasizing (correctly) that AC has increasing returns and damage doesn't, a sword and board wielder should always grab Defense over Dueling. +1 AC is way better than +2 damage per attack at the amounts of AC you're talking about.
I'm not emphasizing AC having increasing returns. I'm emphasizing shields having increasing returns. The +14% toughness from reducing hit chance from 40 to 35 is not better than the 16% damage increase from going from 1d8+8 to 1d8+10.
That's ignoring your own math! Lemme go grab the math you did and present it in-context for both styles.
Now, I don't know where AC 22 is coming from, and +4 attack bonus contradicts the standard assumptions from the DMG, so I'm just going to be explicit about what constants I use, so you can change to your own constants as you like:
Shields only have increasing returns because AC has increasing returns, and both Defense and Dueling can use a shield. The higher your AC already is, the more valuable it is to increase your AC by 1. Damage has decreasing returns - the higher your damage per hit, the less valuable +1 damage per hit is.
My friend the Ph.D. in Math must love these kinds of conversations!
Wow. Lots to catch up on.
This removes the action economy tax but doesn't address the damage gap. The only way to fix that is to give TWF a damage bonus that scales with your extra attacks, which is why I chose to add a d4 to melee attacks after Extra Attack becomes available.
I added two slightly different items (bucklers and parrying daggers) to my game for this purpose, since it's always bugged me that the game doesn't support actual historical dueling styles like dagger and rapier.
I've settled for replacing the -5/+10 in SS with crits on 19 (a very mild buff, but synergizes well with Elven Accuracy and Piercer) and limiting the other two benefits to once per round, up to your Proficiency Bonus times per short rest. That way it's more of a dramatic moment instead of just constantly ignoring the laws of physics. Even if the -5/+10 weren't broken, it's utterly uninteresting. It's either always the correct move or never the correct move for any given enemy.
One thing people tend to miss is that GWFS is extremely consistent with a greatsword. A longsword with Dueling is equally likely to give you a 1 + 2 as an 8 + 2. A greatsword with GWFS has a 91% chance to give you at least a 6, and a 50% chance to roll 7, 8, or 9. It's especially nice for Action Surge and Sentinel reactions. You pretty much never have a bad turn, and I'd argue that at a subjective level that feels better than raising your already good AC.
Saved this response for last since it'll be the longest and I imagine many people have lost interest in the GWM math.
You were right in assuming AC is normally irrelevant, but only if you're comparing equal hit rates and there's no conditional bonuses. That doesn't apply here since GWM changes your hit rate with the -5/+10, and the bonus attack depends on your chances of landing at least 1 crit during your turn, which depends on your number of attacks and whether you have advantage. That kind of naive AC-less comparison also doesn't work on Sneak Attack for the same reason.
The thing about GWM is that the lower your hit rate is, the more that -5 penalty hurts you (e.g. going from 30% hit rate to 5% is an 83.3% drop in your damage.) Likewise the higher your base damage, the less of a relative boost that +10 is and the more damage you're throwing away when the -5 costs you a hit. The -5/+10 isn't always a net gain; your hit rate (which depends on the enemy's AC) has to be high enough and your damage can't be too high either. The answer to this rpg.stackexchange question has some graphs that help visualize this.
There's no easy way to know up how much of a benefit or detriment the -5/+10 trade is going to be for a given AC value other than doing the math with and without it and checking both results (though as a rule of thumb, if you have advantage, it's pretty much always a win.) But since the net gain or loss depends on which AC you used, that raises the obvious question of what AC value makes for a fair comparison. If you assume a 10 AC monster, GWM is going to look great. If you assume a 20 AC monster and no advantage, it's probably a net loss, and it wouldn't make sense to use the GWM damage in a comparison because you'd just be using your regular damage.
The most fair comparison I could think of was to do the math for every AC value, only use the GWM result if it's actually a win, and weigh the results against how common each AC number is. The bonus action also needed to be included because if you have 2 attacks and advantage, you've got 4 chances to roll a 20, and if the bonus action does trigger, it can also benefits from GWM. Even if you only wanted to do the math for 15 AC (which seems to be close enough to the weighted average), once you throw in advantage and the crit chance, the formula would've still been pretty unwieldy for a forum post.
The Forum Infestation (TM)
AC 22 came from a prior example (+4 to hit comes from the actual average of stuff you fight at level 1), and looks to be plate plus defensive fighting style. In any case, your math is suffering from giving a +2 damage even when you miss, and defense has somewhat diminishing returns with level because monster accuracy tends to outpace ac increases.
Doesn't damage suffer from the same diminishing returns as creatures gain HP as the levels increase? Both AC and Damage tend to hit plateaus (such as Fighters at lvl 10 are often doing the same at level 9, but then the third attack at 11 causes a big jump)
Generally true for bonuses to hit too. Going from a 50 to a 55% hitrate is a 10% increase in dpr. But going from a 90 to a 95% hitrate is only a 5.5% increase. TBH its why the GWF and Sharpshooter are over-performing in some situations but not others.
I'm probably laughing.
It is apparently so hard to program Aberrant Mind and Clockwork Soul spell-swapping into dndbeyond they had to remake the game without it rather than implement it.