Savage attacker oddball idea

[Shuruke]

So ive been looking around at opinions on savage attacker (feat not to be confused with half orc feature.)

Looking at some reddit posts and at a post here in playground

http://www.giantitp.com/forums/showt...avage-Attacker

I noticed no one talks about savage attacker being once per turn (other than a stack exchange)

So with sentinel or polearm master
How does that effect the curve of minimal damage and expected damage?


Things that could trigger this

Pole arm master

Sentinel

Mage slayer

Cavalier fighter

Berserker barbarian (on turns after first bonus action attack hold action reaction attack) not as useful after hitting barb 5

Commanding strike from an ally

Any held action for an attack



I understand alot of this aren't the most optimal but I thought I'd see what everyone thinks, make the savage attacker equivalent of when rogues want to trigger sa as often as possible. Dedicate some of build to triggering it often and see how it goes

[Man_Over_Game]

Weapons that people would use with Savage Attacker would normally range between 1d8 and 1d12. So we'll average it to 1d10.

A 1d10's average damage is 5.5
Savage Attacker bumps that up to 7.15.

So you're looking at about +1.65 damage per turn. Assuming you attacked on 2 separate turns, that's +3.3 damage gain per round.

Compare that with something like Polearm Master, where your modifier is +3, you're using the Bonus Action attack 100% of the time (+1d4+3), the Reaction attack 50% of the time (+[1d10+3]/2), PAM puts you at 18.25 damage per round instead of 8.5 (1d10 +3) for a +9.75 damage gain per round.

[Rukelnikov]

Weapons that people would use with Savage Attacker would normally range between 1d8 and 1d12. So we'll average it to 1d10.

A 1d10's average damage is 5.5
Savage Attacker bumps that up to 7.15.

So you're looking at about +1.65 damage per turn. Assuming you attacked on 2 separate turns, that's +3.3 damage gain per round.

Compare that with something like Polearm Master, where your modifier is +3, you're using the Bonus Action attack 100% of the time (+1d4+3), the Reaction attack 50% of the time (+[1d10+3]/2), PAM puts you at 18.25 damage per round instead of 8.5 (1d10 +3) for a +9.75 damage gain per round.

PAM is one of the most powerful feats in the system, and while I agree SA needs some change i think its unfair to expect it to be as good PAM.

IMO the problem with SA is twofold, on the one hand its damage increase doesn't seem that much, on the other hand, its anything but "savage" it makes attacks far less swingy which is the opposite of what I would imagine.

GWM -5/+10 is what Savage Attacker should be, but thats already used.

[Man_Over_Game]

PAM is one of the most powerful feats in the system, and while I agree SA needs some change i think its unfair to expect it to be as good PAM.

IMO the problem with SA is twofold, on the one hand its damage increase doesn't seem that much, on the other hand, its anything but "savage" it makes attacks far less swingy which is the opposite of what I would imagine.

GWM -5/+10 is what Savage Attacker should be, but thats already used.

A few ideas I had:

• When you miss, you can spend your Bonus Action to attempt another attack. Doing so causes attacks against you to be made with Advantage.
• When you are missing at least half of your HP, attacks you make against adjacent creatures are made with Advantage.
• When you are reduced to 0 HP from an attack, you can first spend your Reaction to attack that creature before falling unconscious. This attack is made with Advantage, and, should it hit, it is considered a Critical Hit.

[Vogie]

I'd either include

• the Monster's Rampage ability, which gives a bonus action attack when you kill a target
• the optional cleave rule from the DMG

[Shuruke]

Weapons that people would use with Savage Attacker would normally range between 1d8 and 1d12. So we'll average it to 1d10.

A 1d10's average damage is 5.5
Savage Attacker bumps that up to 7.15.

So you're looking at about +1.65 damage per turn. Assuming you attacked on 2 separate turns, that's +3.3 damage gain per round.

Compare that with something like Polearm Master, where your modifier is +3, you're using the Bonus Action attack 100% of the time (+1d4+3), the Reaction attack 50% of the time (+[1d10+3]/2), PAM puts you at 18.25 damage per round instead of 8.5 (1d10 +3) for a +9.75 damage gain per round.

Yes that's for average damage and assuming the bonus action isn't already being used on something else

I think the better comparison would be

Great weapon master
Savage attacker
Go to bottom for summary

Assuming both a 1d10 weapon

+6 hit
Against 14 a.c.

And
+6 hit against 18 a.c. (2 sets for gwm one without -5)


To start, dice statistics
https://anydice.com/program/15c35ummary
For one attack
Average on dice gwm(mean) 5.50
Deviation 2.87

Savage attacker
Average (mean) 7.15
Deviation 2.35


Assuming two attacks and one reaction attack

Savage attacker against a.c. 14 +6 to hit
21.052

Gwm against 14 a.c. +1 hit

25.8

Against 18 a.c. +6 to hit

Savage attacker 15.292

Gwm -5+10. 15.075

Gwm 13.575 no -5 hit +10 damage




Now calculating 2d6
https://anydice.com/program/5a4
Savage attacker against a.c. 14 (65%)
24.1195

Gwm -5+10 against a.c. 14 (40%)
27
Gwm normal
19.5

Savage attacker against a.c. 18 +6 hit

17.1125

Gwm -5+10 against 18 a.c. +1 hit
16.05

Gwm without -5+10 against 18 a.c. +6 hit
14.55



All of the above are factoring crit chance for all 3 attacks




20 gwm a hit average before %hit chance , % hit chance 25% lower.
1.35 crit damage added after % hit

Savage attacker
11.37 hit average before % hit on 2 of 3 attacks
.9465 crit damage added after % hit on 2 attacks

3rd attack and no -5+10 attack
10 average before % hit
.85 after %hit for crit





Summary

Assuming savage attacker and gwm are both using bonus action for a non attack (spell, shielding, inspiration)

And both are triggering reaction attack

With 1d10 and a 2d6 weapon savage attacker doesnt pull ahead until 50-55% chance to hit (25%-30% for gwm)


When only assuming 2 attacks (savage attacker only applying on 1) it looks like savage attack catches up at around the 45%-50% to hit (35%-45% gwm)


However at point that the gwm chooses not to use -5+10 for both attacks savage attacker takes lead.

When gwm chooses to not do -5+10 on one attack it falls behind at 70% to hit (45% for gwm)

Also I couldn't figure out how to do math for if u only use savage attacker upon rolling 1 less than average or lower

But based off the thread I originally looked at it seems to go more in favor of the savage attacker in that scenario (my guess would probably be a 5% ish favor meaning savage attacker would be better at 50-55% point)


Warning!! I've been working on this for few hours whenever I had moment thus the message I am quoting is only one that was in the post at the time
Hopefully I didn't get shadpwmonked

[Chronos]

Assuming that you have one attack per turn, and therefore always use Savage Attacker, the average increase is 1.65 damage. But if you have multiple attacks, it's slightly better: You'll be choosing which attack you want to use it on, and you'll probably choose one with below-average damage. But it's difficult to model this, because you'd need in the process to determine the optimal algorithm for how to choose which attack to use it on, and even with such a model, you'd have a very branchy decision tree for calculating probabilities.

[bid]

But it's difficult to model this, because you'd need in the process to determine the optimal algorithm for how to choose which attack to use it on, and even with such a model, you'd have a very branchy decision tree for calculating probabilities.

Not really.

The average of 1d10 is 5.5, if you rolled 5 or less it's worth using savage attacker. This will add (5.5 - 3)/2 ~ 1.25 expected damage.

So, if you rolled 4 your expected gain is (5.5 - 4 ~ 1.5) which is more than the 1.25 expected damage. You use your savage attacker now.

That means your first-of-two use adds (5.5 - 2.5)/3 ~ 1 expected damage and your overall savage attack gain is (5.5 - 2.5)/3 + (5.5 - 3)/2*2/3 ~ 1.83

Which means for 3 attacks with 1d10: 1st attack = use on 3, 2nd attack = use on 4, 3rd attack = use on 5.

[Mjolnirbear]

My current potential house rule for savage attacker:

As-is, plus
* if you bring an opponent to 0 HP, you may make a free attack at another enemy within 5 feet
* when you are at less than half your maximum HP, you may use the damage dice reroll with all your successful attacks.

[Chronos]

Quoth bid:

The average of 1d10 is 5.5, if you rolled 5 or less it's worth using savage attacker. This will add (5.5 - 3)/2 ~ 1.25 expected damage.

It's not that simple. Imagine an extreme case of a 20th-level fighter with haste and using her action surge (9 attacks in a round, and likely 10 if Great Weapon Master triggers and gives her a bonus action attack). Out of that many attacks, there's a pretty good chance that one of those attacks is going to be a 1, and an even better chance that one is going to be at most a 2. That 1 or 2 is going to be the attack you want to use Savage Attacker on. If you happen to roll a 5 on your first damage roll, you probably don't want to use it then, even though it's below average, because there's probably going to be another attack later on which it'll be even more useful.

[bid]

It's not that simple.

Yeah, it's really that simple. Sorry to break your bubble.

Maximizing gain expectancy aint no rocket science. If on your next roll SA offers a 1.25 gain but on this roll it offers a 2.5 gain, the answer is obvious.

And on the extremely rare case that has no impact on the feat's value, you can either extend my 5 minutes job or just WAG it:

Which means for 3 attacks with 1d10: 1st attack = use on 3, 2nd attack = use on 4, 3rd attack = use on 5.

With 10 attacks, you can safely reroll 1-2 on the first 5 and then go 1-3 until the last few hits. I don't think an error of 1 damage is going to matter out of 10 attacks, I'd be way more worried about the 5-ish damage lost to overkill.

[djreynolds]

I have used the feat at my table, its not terrible

This is the issue, it is damage, so unless you have a max attack stat I wouldn't take it

And by the time you get a maxed stat, 8th or 12th level for some classes... you now have other feats needed and damage is not one of them.

It is also once a turn, and doesn't really mesh with fighter's extra attack

If GWM or SS did not exist, players would take this feat and think it was awesome

My recommendation is give out one night and actually see for yourself, how did this feat affect the actual battle.

Did the GWM fighter squeeze out extra BA attacks because the reroll from savage attacker dropped a foe (perhaps they took defensive style over great weapon fighting)... this does come up, need only 4 more HP to kill the BBEG or next turn they drop a big AoE or get away

Feats like GWM, PAM, & SS really make savage attacker not necessary for classes like a fighter, and even a dexterity S&B fighter should take shield master over savage attacker.

But the damage is every turn, give the feat out as blessing from some god and see how it actually performs