D&D 5E Sneak Attack: Is it broken?

[Prism]

Honestly, it's kind of rare to miss attacks in 5e, especially after about level 8. AC just never scales up enough to make it a bit deal. Not on the monster side anyway.

I just had a quick scan of monsters in the MM from CR10 upwards. I can't really find one with less than AC17 and many have 18 and 19. A 10th level PC has about +8/+9 to hit. It seems to scale just fine to me. Seems pretty similar to the AC of an average PC, except for a shield user (and many of these monsters could use a shield if they want to I guess)

 

[FormerlyHemlock]

I just had a quick scan of monsters in the MM from CR10 upwards. I can't really find one with less than AC17 and many have 18 and 19. A 10th level PC has about +8/+9 to hit. It seems to scale just fine to me. Seems pretty similar to the AC of an average PC, except for a shield user (and many of these monsters could use a shield if they want to I guess)

Nitpick: you can't necessarily assume that 10th level PCs will be fighting CR 10 monsters. Bounded Accuracy is designed to prevent monsters from going obsolete. It would be more appropriate to average the AC of all monsters under CR 10.

E.g. two Chuuls, two ghasts, and four Ghouls is a Hard fight for four 11th level PCs, but they have ACs of 16, 13, and 12 respectively.

 

[Prism]

E.g. two Chuuls, two ghasts, and four Ghouls is a Hard fight for four 11th level PCs, but they have ACs of 16, 13, and 12 respectively.

Sure but that particular fight presents its own paralyzing difficulties

Having played quite a bit of high level 5e now I think that creatures seem about as easy/hard to hit as at lower levels in general - especially in the important encounters that really matter. Having magical plussed weapons and armour certainly sways this but then the encounters should be higher CR to compensate really. Sure the low levels foes get easier to hit, but the new high level ones seem pretty tough.

 

[FormerlyHemlock]

Sure but that particular fight presents its own paralyzing difficulties

Yes, it does. But hitting the monsters isn't likely to be one of them.

For another, less-paralytic example which still illustrates the point (that high-level PCs don't have to fight high-CR monsters): six cult fanatics (CR 2) and Giant Crocodile (CR 5) are a Hard fight for 4 11th level PCs. AC 14 and 13, respectively, although one of the Cult Fanatics should probably cast Shield of Faith on Sobek, the crocodile. Man, that sounds like a fun fight to run too. Eat Hold Person and Spiritual Weapon x5 while a crocodile claws your face off, PCs! Plus I love the Egyptian flavor of a crocodile cult.

Hmmm, I guess that is still a paralyzing example. Pretend I said "Centaurs" (same CR, AC 12) instead of "Cult Fanatic", although I'd rather run the Cult Fanatics.

 

[aramis erak]

It may be a spotlight effect, where when it does work well, it does work well, but overall it is meh or average. If I would point to anything, it is how easy is it to get advantage in the game. Since advantage is useful for every class, it becomes a primary mechanic that is sought at every turn.

By third level, it's great for getting a caster to fail a concentration save. See, 3d6 extra, plus a typically d6 or d8 for the weapon weapon; it's fine in terms of DPR, but it's still a single hit and so a single save vs the damage.

for comparison:
Fighter level 3: 1 attack at 1d8+3, 2d6+3, or 1d10+3 is typical.
Fighter-battlemaster: may be adding an additional d8 to the above
Rogue 3 (any): 1d6+3 or 1d8+3. If an ally is there, or advantage from any means, it's 3d6+3 or 1d8+2d6+3.

At level 5:
Fighter is 2x (1d8+4 or 2d6+4 or 1d10+4), possibly +1d8 from the battlemaster maneuvers
Rogue is 1x (1d8+4 or 1d6+4), and if an ally engaged, or advantage, +3d6...

the rogue consistently stays slightly above fighters on DPR, but has lower armor, and no dividing that damage up.

 

[77IM]

At level 5:
Fighter is 2x (1d8+4 or 2d6+4 or 1d10+4), possibly +1d8 from the battlemaster maneuvers
Rogue is 1x (1d8+4 or 1d6+4), and if an ally engaged, or advantage, +3d6...

the rogue consistently stays slightly above fighters on DPR, but has lower armor, and no dividing that damage up.

Your formulae do not match your conclusion.

With a greatsword, I see your level 5 fighter dealing 22 damage (if both attacks hit), and that's not even accounting for greatweapon fighting style letting the fighter reroll 1s and 2s, which is worth about +4 overall. Nor does it account for battlemaster maneuvers or champion crit chance or heavy weapon feat.

Meanwhile the rogue with a 1d8 weapon is dealing only 19 damage. In the long run a single attack deals less damage than two attacks at half damage. The rogue's saving grace is that they can often get advantage from stealth, or attack twice with TWF for a possible extra 3.5 damage. But since monster ACs are low this probably amounts to around a +20% expected damage, or +4.

So I see the two classes as about even -- the fighter's probably dishing more damage but the rogue is more likely to hit so the expected damage output is the same, with the fighter having the potential to do even more damage via feats and class abilities.

 

[FormerlyHemlock]

Good point about concentration. I don't think 5th level is going to seriously threaten anyone's concentration, but a 9th level Rogue with 5d6 sneak attack damage plus d8+5 weapon damage is on the verge of becoming a serious threat to concentration, especially if he rolls well. I rolled 5d6+d8+5 ten times and here's what I got:

27, 29, 30, 27, 20, 37, 25, 29, 23, 19

Respective concentration DCs are 13, 14, 15, 13, 10, 18, 14, 11, 10. By the time he hits 20th level he'll be breaking concentration effortlessly on every attack with DCs in the 20 to 30 range.

 

[Dausuul]

The issue is that most of 5E is designed around the principle that attack damage scales by adding more attacks. Therefore, when you have an attack that scales by doing more damage per attack, it doesn't play nice with the rest of the system. In isolation, Sneak Attack is perfectly fine.

 

[Zalabim]

Fighters, rangers and warlocks add more attacks. Monks make a lot of attacks, but reach their max of 4 at level 5. Fighter subclasses, rangers, barbarians, paladins, monks, rogues, and spellcasters in general scale at higher levels by dealing more damage per attack. The fact that fighters get the most martial attacks is part of how they're the class that most benefits from magic weapons, while extra attacks are more noticeable for the classes that deal more damage per attack, like barbarian rage, paladin smite, ranger mark, and rogue sneak attacks.

 

[EzekielRaiden]

Alright. So it seems like the general consensus is that Sneak Attack performs its function as a damage booster, analogous to Smite or Rage damage (but...differently reliable, perhaps is the best way to put it). It's not a broken mechanic, nor does it necessarily out-do equivalents from other classes except very contextually.

People do seem to recognize a certain "fragility" about it though. So, my next question then becomes: Is this "fragility" actually a serious issue for a game like 5e, which explicitly (straight from its designers mouths, as I understood it) doesn't care as much about balance? Or is it just something to remember, e.g. something that should be noted for the DM to try to work around?

Basically I just get suspicious because it seems like every attack feature under the sun gets called "broken" if, and sometimes only if, it gets paired with Sneak Attack. When an ability appears to single-handedly constrain a significant portion of design space, well...it might not be "broken," but its "fragility" starts to sound like a flaw that could have been avoided. Perhaps even should be avoided.