Not-so-savage attacker

[LVOD]

Upon first reading, i thought the savage attacker feat sounded amazing. No longer would i experience the moments where an attack connects only to deal unimpressive damage due to a bad roll.

Then i actually crunched the numbers...

For a d4, it nets an extra 5/8 of a point.
For a d6, its 1.172222 points, and for a d8 its a 1.3125 point boost.

Even if your damage die was a d20, its only a benefit of 3.325 points (which, by the way, is the benefit you get from having advantage on a roll).

Is there any redeeming quality to this feat at all? Am i missing something?

[JakOfAllTirades]

Upon first reading, i thought the savage attacker feat sounded amazing. No longer would i experience the moments where an attack connects only to deal unimpressive damage due to a bad roll.

Then i actually crunched the numbers...

For a d4, it nets an extra 5/8 of a point.
For a d6, its 1.172222 points, and for a d8 its a 1.3125 point boost.

Even if your damage die was a d20, its only a benefit of 3.325 points (which, by the way, is the benefit you get from having advantage on a roll).

Is there any redeeming quality to this feat at all? Am i missing something?

It's most useful with the Barbarian Brutal Critical class feature,especially when using both Extra Attack and Reckless Attack. It gets even better if you resolve all your attack rolls (including damage) at once, so you know which one will benefit most from the re-roll.

So yeah, if you're a crit-fishing Barbarian it's pretty good, otherwise not so much.

[bid]

It's once per turn.
anydice: output [lowest 1 of 3d12], average 3.52

If we're generous, that means +3 damage per turn while boosting your main stat would yield +2 damage and improve hit.

The only way to redeem it would be to house rule it to use the same mechanic as durable: "every attack, you can replace one of you damage dice by your attack modifier".

[Giant2005]

To be fair, people rave about the amazingness of GWM's +10/-5 effect, yet when fighting level appropriate foes, that effect tends to increase one's DPR by a much smaller margin than Savage Attacker does.
For example, a level 20 Fighter with GWF and GWM, using the +10/-5 effect against AC 20 (the average AC of a CR 20 creature), increases his DPR by 0.67.

[Specter]

To be fair, people rave about the amazingness of GWM's +10/-5 effect, yet when fighting level appropriate foes, that effect tends to increase one's DPR by a much smaller margin than Savage Attacker does.
For example, a level 20 Fighter with GWF and GWM, using the +10/-5 effect against AC 20 (the average AC of a CR 20 creature), increases his DPR by 0.67.

Yep, and Great Weapon Fighting follows the same line, and yet it's rated gold in many guides. Go figure.

[krugaan]

Yep, and Great Weapon Fighting follows the same line, and yet it's rated gold in many guides. Go figure.

That's because spike damage is psychologically more awesome and sometimes more useful. Plus with advantage the numbers work out different?

[jas61292]

Yep, and Great Weapon Fighting follows the same line, and yet it's rated gold in many guides. Go figure.

Its a case of thinking too much about numbers and not enough about actual play. Or, more specifically caring too much about average values, and not real values.

I DMed a barbarian that used a greataxe and took savage attacker, and it was very effective. Sure, sometimes it didn't do anything, but many times it would increase the damage that he would do by 5-10. An increase off 7 damage one round that allows you to finish off an enemy (even if that is the only increase you get all battle) is far, far more important than any increase to average DPS. Yeah, it being valuable is luck based, but so much of the game is. It is actually very similar to the difference between a greatsword and a greataxe. People will always (and I mean always) talk about how superior the greatsword is as far as the numbers. But when you come up to a situation where getting a 12 will kill, while getting a 11 will not, the greataxe is the far better weapon.

Higher average DPS certainly has its advantage, but it is not the be all and end all of success. What matters is never how much you can do to a punching bag over the course of 10 minutes. Its what you can do in whatever situation you are presented with. Sometimes the higher average DPS option is the best. But sometimes, have picked up Savage Attacker will save your life in a way no other feat would.

[Rysto]

GWF doesn't cost you an ASI, so that's a poor comparison.

If the level 20 Fighter has advantage on all of his attacks, +10/-5 from GWM gains +9.5 DPR.

[MaxWilson]

Is there any redeeming quality to this feat at all? Am i missing something?

Well, basically no, but it scales pretty okay if you turn into an animal with higher base damage dice. For example, you could have some fun as a Mobile Savage Attacker rhinoceros doing 4d8+5 (with reroll) each round by level 6. Depending on how your DM interprets the "statistics" clause of Polymorph you might also have some fun as a Polymorphed T-Rex.

[bid]

Well, basically no, but it scales pretty okay if you turn into an animal with higher base damage dice. For example, you could have some fun as a Mobile Savage Attacker rhinoceros doing 4d8+5 (with reroll) each round by level 6. Depending on how your DM interprets the "statistics" clause of Polymorph you might also have some fun as a Polymorphed T-Rex.
anydice: output [highest of 4d8 and 4d8], average 20.6

That's still only +2 per turn.

[hymer]

boosting your main stat would yield +2 damage and improve hit.

How do you figure +2?

[Skylivedk]

Upon first reading, i thought the savage attacker feat sounded amazing. No longer would i experience the moments where an attack connects only to deal unimpressive damage due to a bad roll.

Then i actually crunched the numbers...

For a d4, it nets an extra 5/8 of a point.
For a d6, its 1.172222 points, and for a d8 its a 1.3125 point boost.

Even if your damage die was a d20, its only a benefit of 3.325 points (which, by the way, is the benefit you get from having advantage on a roll).

Is there any redeeming quality to this feat at all? Am i missing something?

It's one of the bottom-tier feats from a mechanical point of view. I've not smallest inkling of an idea why anyone would defend it. In my campaign, I've turned it into a half-feat. It's still not picked much.

[Skylivedk]

To be fair, people rave about the amazingness of GWM's +10/-5 effect, yet when fighting level appropriate foes, that effect tends to increase one's DPR by a much smaller margin than Savage Attacker does.
For example, a level 20 Fighter with GWF and GWM, using the +10/-5 effect against AC 20 (the average AC of a CR 20 creature), increases his DPR by 0.67.

GWM also adds to options for granting bonus attacks, plus, as shown by another poster, adds a whole lot more damage if you tilt the ac/to hit a little bit.


Yep, and Great Weapon Fighting follows the same line, and yet it's rated gold in many guides. Go figure.

GWF is mediocre/horrible if it doesn't apply to rider effects. If people rate it gold without applying it to riders, they should have their DnD Mechanics Diploma revoked.

[bid]

How do you figure +2?
4d8 average = 18, vs anydice average 20.6

Ok, that's +2.6.

[Spacehamster]

Yep, and Great Weapon Fighting follows the same line, and yet it's rated gold in many guides. Go figure.

Well you do not use the -5/+10 against high AC enemies or mitigate the minus with bless or advantage and its all good. :)

[BigONotation]

It's a sub-standard feat. As a DM, I offer + 1 to str or dex with this feat.

[MrStabby]

If sneak attack counts as weapon damage then it is ok on a rogue. Even then I find better things to take as a feat.

I am not sure how people are calculating damage bonuses for GWM for level appropriate encounters. Is there an assumption that a level appropriate encounter is only ever against one creature at a time and no more? Is there also an assumption that you keep using the feat even if it is disadvantagious?

[DeAnno]

Don't forget the assumption that you never have advantage or other accuracy boosts like Bless, Superiority Dice, etc.

[Ruslan]

Upon first reading, i thought the savage attacker feat sounded amazing. No longer would i experience the moments where an attack connects only to deal unimpressive damage due to a bad roll.

Then i actually crunched the numbers...

For a d4, it nets an extra 5/8 of a point.
For a d6, its 1.172222 points, and for a d8 its a 1.3125 point boost.

Even if your damage die was a d20, its only a benefit of 3.325 points (which, by the way, is the benefit you get from having advantage on a roll).

Is there any redeeming quality to this feat at all? Am i missing something?
You are missing something. It's actually worse than you think.

It's not the same benefit as having Advantage on the roll. If you have Advantage, you roll both dice at the same time, choose the higher one. Even if one d20 die is 19, the other die could improve it to 20, right?

But with Savage Attacker, you roll the first damage die, then choose whether to reroll and use the second roll. So if your first roll is 19 (or even 11), the second roll for sure cannot improve it. Because there is no second roll. Because having rolled a 19 (or even 11) on d20, the percentage play is to keep the roll and not to reroll.

So, with that in mind:
For a d4, it's a 0.5 points boost
For a d6, it's a 0.75 points boost
For a d8, it's a 1 point boost
For a d20, it's a 2.5 points boost

Yeah, it's a pretty bad feat. What I would recommend to salvage it would be to slap "Increase your Strength score by 1, to a maximum of 20" onto it.

[MrStabby]

You are missing something. It's actually worse than you think.

It's not the same benefit as having Advantage on the roll. If you have Advantage, you roll both dice at the same time, choose the higher one. Even if one d20 die is 19, the other die could improve it to 20, right?

But with Savage Attacker, you roll the first damage die, then choose whether to reroll and use the second roll. So if your first roll is 19 (or even 11), the second roll for sure cannot improve it. Because there is no second roll. Because having rolled a 19 (or even 11) on d20, the percentage play is to keep the roll and not to reroll.

So, with that in mind:
For a d4, it's a 0.5 points boost
For a d6, it's a 0.75 points boost
For a d8, it's a 1 point boost
For a d20, it's a 2.5 points boost

Yeah, it's a pretty bad feat. What I would recommend to salvage it would be to slap "Increase your Strength score by 1, to a maximum of 20" onto it.

Even if you roll twice you can use either total.

[hymer]

@ bid: Thanks, but that's not what I ment. You wrote that boosting your main stat gives +2 to damage in addition to the to-hit boost. I'm wondering what's the thinking behind that?

[Zalabim]

Oh, that's simply from two attacks per round.

[FightStyles]

Yeah, it's a pretty bad feat. What I would recommend to salvage it would be to slap "Increase your Strength score by 1, to a maximum of 20" onto it.

Dang you. Reading your post, I thought you were going to improve it. Instead you just made it a half feat.

I do think you made a good point that you the re-roll is not advantage, but simply a re-roll. That should be changed simply for the fun of it.

Essentially, savage attacker just gives you advantage on the damage die of an attack. Usable once per turn. Also give them +1 to str or dex because it fits and is probably needed still as others have mentioned.

I could get down on that.

[smcmike]

As it is currently written, it is slightly better than advantage once per turn on your weapon damage dice. It's advantage that you can choose to use after the first roll, or to save for your second attack.

I'm not saying it's good, but it's better than 1/turn damage advantage.

[RulesJD]

You are missing something. It's actually worse than you think.

It's not the same benefit as having Advantage on the roll. If you have Advantage, you roll both dice at the same time, choose the higher one. Even if one d20 die is 19, the other die could improve it to 20, right?

But with Savage Attacker, you roll the first damage die, then choose whether to reroll and use the second roll. So if your first roll is 19 (or even 11), the second roll for sure cannot improve it. Because there is no second roll. Because having rolled a 19 (or even 11) on d20, the percentage play is to keep the roll and not to reroll.

So, with that in mind:
For a d4, it's a 0.5 points boost
For a d6, it's a 0.75 points boost
For a d8, it's a 1 point boost
For a d20, it's a 2.5 points boost

Yeah, it's a pretty bad feat. What I would recommend to salvage it would be to slap "Increase your Strength score by 1, to a maximum of 20" onto it.

Damn, even on the best class to use it on (Rogue with Sentinel to grant off-turn attacks and lots of dice) it's a garbage feat.

[Kryx]

For example, a level 20 Fighter with GWF and GWM, using the +10/-5 effect against AC 20 (the average AC of a CR 20 creature), increases his DPR by 0.67.
Let's compare at all tiers

Barbarian GWM and Fighter BM GWM

Level 5 vs AC 14
Barbarian: 36.6 DPR vs 25.4 DPR, or 44%
Fighter: 28.2 DPR vs 24.2 DPR, or 16.5%

Level 11 vs AC 17
Barbarian: 35.5 DPR vs 29.4 DPR, or 20.7%
Fighter: 44.5 DPR vs 40.4 DPR, or 10.1%

Level 17 vs AC 19
Barbarian: 45.2 DPR vs 38.3 DPR, or 18%
Fighter: 47.9 DPR vs 43.7 DPR, or 9.6%

Level 20 vs AC 20
Barbarian: 51.6 DPR vs 43.5 DPR, or 18.6%
Fighter: 54.1 DPR vs 52.0 DPR, or 4%

Now to be fair to my numbers both of these builds use advantage quite a lot. The same is true for Bladelock 17+ and OoV Paladin. These numbers also assume that the bonus attack is given, it isn't compared to not taking the feat at all which I did here: https://docs.google.com/spreadsheets/d/1d-9xDdath8kX_v7Rpts9JFIJwIG3X0-dDUtfax14NT0/edit#gid=275734750 Which shows it's the far superior option until ~15 at which point it's about even and then falls below at 20.
For builds like EK, Champion, non-OoV Paladin, Bladelock 1-16 it's not quite as valuable, but throughout their whole career usually worth a fair amount more than .67 DPR.

But lets not turn this into a -5/+10 debate thread - we've done it many times.



Fixing Savage attacker:
I would probably remove the "Once per turn" and make it a half feat. That way it can give a boost to a stat (or use half feats directly) and the effect is actually more worthwhile.
It turns a longsword from 4.5 to 5.81, a polearm from 5.5 to 7.15, a greataxe from 6.5 to 8.49, and a greatsword from 7 to 8.37. It is a much better choice then.

[MaxWilson]

Dang you. Reading your post, I thought you were going to improve it. Instead you just made it a half feat.

I do think you made a good point that you the re-roll is not advantage, but simply a re-roll. That should be changed simply for the fun of it.

What's the difference, in practice? Ruslan misread the feat--since you are permitted to "use either total" you can still attempt the reroll over if you roll high on your initial hit. So it might as well be "damage advantage" even by RAW, no change needed.

[smcmike]

What's the difference, in practice? Ruslan misread the feat--since you are permitted to "use either total" you can still attempt the reroll over if you roll high on your initial hit. So it might as well be "damage advantage" even by RAW, no change needed.

It's better than damage advantage, though, if you have multiple attacks.

[MaxWilson]

It's better than damage advantage, though, if you have multiple attacks.

That's true (it's better than damage advantage on one roll, but worse than damage advantage on all rolls), but I don't think that could have been FightStyles' point, since he was speaking as if it were worse than damage advantage on a single roll, not better.

I'm curious whether I misunderstood his writing.

[bid]

What's the difference, in practice? Ruslan misread the feat--since you are permitted to "use either total" you can still attempt the reroll over if you roll high on your initial hit. So it might as well be "damage advantage" even by RAW, no change needed.
Savage is not a advantage on every damage roll but only once. This means you have to decide to "waste" it on the first roll before seeing the next damage.

Maybe that was his muddled POV and you belief he misread the feat is just miscommunication.

[RulesJD]

The best setup to take advantage of this Feat would be a class that can reliably have Off-Turn attacks that involve a lot of dice, the bigger the better.

Presumably the best class then is a Rogue with Sentinel and some levels of Fighter for Riposte.

Best case scenario, the Rogue is pumping out ~10d6 per hit of dice (counting weapon), that's 7.5 extra damage per hit (~19% increase). So ~15 per turn, if you get to sneak attack twice.

Now that's not too bad, but that's at level 20. It gets dramatically worse the further you go down as a Rogue.

I could see it as a higher level feat choice on an extremeeeely limited number of builds.

[bid]

Best case scenario, the Rogue is pumping out ~10d6 per hit of dice (counting weapon), that's 7.5 extra damage per hit (~19% increase). So ~15 per turn, if you get to sneak attack twice.
anydice: output [highest of 10d6 and 10d6], average 38.05

That's +3.05 damage above average (8.7%), I wonder how you got your number.

[RulesJD]

anydice: output [highest of 10d6 and 10d6], average 38.05

That's +3.05 damage above average (8.7%), I wonder how you got your number.

Went with prior poster's 0.75 per d6 and multiplied by 10 *shrug*.

Never mind then, still garbage.

[bid]

Went with prior poster's 0.75 per d6 and multiplied by 10 *shrug*.
Yeah it's tricky. You can't multiply when you add more dice to the reroll.
{ 2+5 vs 4+3 = 7} vs { 2 vs 4 + 5 vs 3 = 9}, and it keeps diverging when you take in all the cases.

[MrStabby]

Yeah, roughly speaking you want to multiply by the square root of the number of dice, not by the number.

[bid]

Yeah, roughly speaking you want to multiply by the square root of the number of dice, not by the number.
Really? Never thought it'd be that simple, thanks! Because deviation is squared, I guess.

[BurgerBeast]

Yeah it's tricky. You can't multiply when you add more dice to the reroll.
{ 2+5 vs 4+3 = 7} vs { 2 vs 4 + 5 vs 3 = 9}, and it keeps diverging when you take in all the cases.

Am I misunderstanding your notation or shouldn't these both total 9?

Isn't { 5+4 vs 3+2 = 8 } what you're trying to illustrate?

[Vogonjeltz]

Upon first reading, i thought the savage attacker feat sounded amazing. No longer would i experience the moments where an attack connects only to deal unimpressive damage due to a bad roll.

Then i actually crunched the numbers...

For a d4, it nets an extra 5/8 of a point.
For a d6, its 1.172222 points, and for a d8 its a 1.3125 point boost.

Even if your damage die was a d20, its only a benefit of 3.325 points (which, by the way, is the benefit you get from having advantage on a roll).

Is there any redeeming quality to this feat at all? Am i missing something?

It's ideal for a character with multiple attacks, or many dice with a large range (sneak attack comes to mind). It would also be best reserved for either the first attack output that's below the average or the last attack for the round.

[bid]

Am I misunderstanding your notation or shouldn't these both total 9?

Isn't { 5+4 vs 3+2 = 8 } what you're trying to illustrate?
Not really

Advantage on blue die : roll 2 first, roll 4 second = 2 vs 4. best is 4
Advantage on red die : roll 5 first, roll 3 second = 5 vs 3, best is 5

Rolling both blue and red die at the same time:
- first roll is blue = 2 and red = 5 : total 7
- second roll is blue = 4 and red = 3 : total 7
= 7 vs 7, best is 7


So a sum of advantage is not the advantage of the sum.