D&D 5E Sneak Attack: optional or mandatory?

[Ratskinner]

My preference for the base rogue would be this: clerical attack and hit dice. (At the moment the attack bonus is the same, but hit dice are too low and there is no extra attack at 8th). The rogue would then rely on stealth (and resultant bonuses/advantage) to get chance to-hit onto a par with the fighter. Damage would be lower than the fighter, but that is part of a "three pillars" trade off; in social or exploration the fighter would be weaker than the rogue.

Turning the rogue into a 4e-style swashbuckler would certainly be one viable sub-class, and at that point damage might be on a par with the fighter. And an assassin, who can surpass fighter damage under stealth conditions, is another viable sub-class. But I don't think these should be the default.

Along similar lines, I actually think that the Warlock(s) would fit as a magic-using rogue sub-class(es), rather than under Wizard. ....but maybe I shouldn't bring that up as some folks feel kinda touchy about that.

 

[N'raac]

I don't have a definite answer to that question. But as I indicated to @ThirdWizard upthread, more often than "never" would be a good start.

That frequency would determine how costly it is to make a poor choice. If a Rogue who trades away Sneak Attack (or, for that matter, a Wizard who decides he picked the wrong specialty) is locked in for 2 or 3 levels, that's too long. Retraining, to the extent it has ever been allowed historically, has been slow, with pretty limited changes at each step.

How do you use expertise dice to get reliable benefits? Via social rules. Via stealth rules. Via distraction rules. Via evasion rules (of the classic D&D variety). The same sort of way that a wizard uses Charm Person, or Fog Cloud, to avoid violence or to bring violent clashes to a close.

COMBAT reliability. I'd be fine with options that allow a rogue to use social skills to get a significant advantage in combat, but that means accepting rogues can use social skills not only on hostile, unwilling targets, but on targets actively attacking them. It also means, that PC and NPC rogues must be allowed to use the same skills - with binding results - on PC's. Just like a wizard casts Charm Person on a PC and, assuming he fails his save, the PC is bound by with the result.

Stealth and distraction? You mean like getting the opponent to momentarily drop his guard enabling a devastating blow to penetrate his defenses? In any case, we're back to combat uses. My point was simply that an expertise die for noncombat skills does not make the rogue combat-viable.

My preference for the base rogue would be this: clerical attack and hit dice. (At the moment the attack bonus is the same, but hit dice are too low and there is no extra attack at 8th). The rogue would then rely on stealth (and resultant bonuses/advantage) to get chance to-hit onto a par with the fighter. Damage would be lower than the fighter, but that is part of a "three pillars" trade off; in social or exploration the fighter would be weaker than the rogue.

By getting rid of the d4 (bumped to d6) for HD, and bumping all d6 to d8,Pathfinder got the first half. Three BAB progressions, each linked to a HD (6/8/10, with the Barbarian the unusual d12).

Note that I also believe the fighter should not be able to trade away his non-combat abilities for better combat skills. Setting the bar that the difference should be more or less the same, and both should have significant utility in both combat and non-combat (or all three pillars, or also ranged and melee combat, or whatever we consider a significant part of the game) would be an excellent goal.

Until and unless there are other equally viable combat abilities for the Rogue, and it sounds like there will be none in the core/default, I remain with the view the Rogue needs Sneak Attack. I would much rather he got a choice of a variety of combat abilities, but the designers do not appear to be pursuing that approach, unfortunately.

 

[Starfox]

A different approach to answering that question focuses on the metagame, thematic/aesthetic framing of the classes. A rogue is a backstreet thug. Therefore, when you play a rogue, your PC will play as a backstreet thug - including sapping people from behind. A fighter is a front-rank gloryhound. Therefore, when you play a fighter, your PC will stand in the front-rank, never falling, stopping the baddies from breaking through to your friends. Try to play a fighter as a dishonourable backstree thug and the rules will push back against you (eg you won't, under the rules, be able to knock people out with saps).

Basically, I agree with all of this.

The quoted line is the crux of it - I namely that it changes what a class is from a building block for character design to a trope made flesh.

The "villain" in this case is the hp mechanic. In a system where defense is handled through parry rolls and very few attacks connect but those that do are lethal (Rune Quest et al), there is no need for sneak attack damage - an attack that strikes an unaware opponent cannot be parried, and thus becomes lethal naturally. This is the effect Sneak Attack goes for. Really, everyone should be able to Sneak Attack, because any surprising attack cannot be parried. But that would make hp meaningless, and so DnD makes a special role/class out of it.

Now, I personally think hp works out decently and see no need to change them - and such a change is not on the table for Next anyway. Therefore I am willing to accept Sneak Attack. The parry mechanic of RQ has it's own problems, and never become quite as heroic as hit points. If you want grim and gritty, parry as a defense works pretty well. If you want heroic play, something more predictable, like hit points, works better.

 

[Starfox]

That frequency would determine how costly it is to make a poor choice. If a Rogue who trades away Sneak Attack (or, for that matter, a Wizard who decides he picked the wrong specialty) is locked in for 2 or 3 levels, that's too long. Retraining, to the extent it has ever been allowed historically, has been slow, with pretty limited changes at each step.

I am all for liberal respecification options.

Personally, I have always been very liberal about character respecification when I DM. If something does not work out or does not live up to expectations, I always allow retraining. Especially in a game like 3E that requires you to plan several levels ahead it is really easy to build your character into a dead end. But many DMs do not. Having a rule about it changes nothing at my table, but it does give players a lever with DMs that are less liberal than I am about these things.

But then maybe this belongs more in DM advice than in rules.

In a group with many min/maxers, this can of course be a problem - players may want to respec from one build that excels at level 5 to another that excels at level 7, say. But these are corner cases, and when optimization has gotten to this kind of extreme, it becomes a game all its own. No rules can probably stop it anyway, because whatever rules are adopted becomes the rules to optimize around. Actually, in this case, I feel rare but large respec (radiation accident in HERO system terms) work better than small tweak respecs at many points.

* Fondly recalls some of my own level-morphing builds from 4E maximizing use of the every level respec option - builds than never got into play.

 

[pemerton]

that means accepting rogues can use social skills not only on hostile, unwilling targets, but on targets actively attacking them. It also means, that PC and NPC rogues must be allowed to use the same skills - with binding results - on PC's. Just like a wizard casts Charm Person on a PC and, assuming he fails his save, the PC is bound by with the result.

I think PC/NPC symmetry is inherent to D&Dnext's design, for better or worse.

And I agree that for social skills to work they have to be able to affect people attacking you.

 

[N'raac]

In a group with many min/maxers, this can of course be a problem - players may want to respec from one build that excels at level 5 to another that excels at level 7, say. But these are corner cases, and when optimization has gotten to this kind of extreme, it becomes a game all its own. No rules can probably stop it anyway, because whatever rules are adopted becomes the rules to optimize around. Actually, in this case, I feel rare but large respec (radiation accident in HERO system terms) work better than small tweak respecs at many points.

I'm not sure building retraining into the character design is a corner case. One example of hard-coded retraining is the ability of a sorcerer to swap out a spell every few levels. How many sorcerer build notes fail to take a spell like Sleep at 1st level with the stated intention of swapping it out at a later level, when the real opposition is no longer likely to be affected? Commonly the swap in would be Magic Missile, which is a lot more impressive doing 6 - 15 damage than 2-5. Seems like that's using retraining to convert from a build that worked well at L1-2 to one that works better at L5-6. And I don't think anyone ever called that out as min-maxing. Where one crosses the line is a very subjective determination.

 

[N'raac]

I think PC/NPC symmetry is inherent to D&Dnext's design, for better or worse.

And I agree that for social skills to work they have to be able to affect people attacking you.

I'm good with that symmetry. I think for social skills to be considered a viable inclusion or replacement for a combat ability, they have to affect people attacking you. I don't think this is necessary for them to be a useful noncombat skill, though. There are two schools of thought and I have seen some players who are very opposed to the thought a raging Barbarian can be talked down by a smooth talking con man. Of course, there are also a lot of players very opposed to any ability of NPC social skills impacting their PC decisions, so some controversy is inevitable.

Is it explicit somewhere in Next to date that social skills particularly affect PC's just as much as NPCs?

 

[Minigiant]

All believe the real culprit for the sneak attack issue is our own imagination. This game fueled by imagination would have player imagine characters other than the urban thuggish thief, the disciplined and trained warrior, the combat ready holyman, and the scholarly spellcaster.

So when a player asks to play the acrobatic performer, a charming upperclassman, a man of science. a wizard school dropout, or many of the other guys in the world. It is up to the game to decide on if it has ways to support these people and how or if it is up to the players themselves to design them.

 

[N'raac]

All believe the real culprit for the sneak attack issue is our own imagination. This game fueled by imagination would have player imagine characters other than the urban thuggish thief, the disciplined and trained warrior, the combat ready holyman, and the scholarly spellcaster.

So when a player asks to play the acrobatic performer, a charming upperclassman, a man of science. a wizard school dropout, or many of the other guys in the world. It is up to the game to decide on if it has ways to support these people and how or if it is up to the players themselves to design them.

I think it is up to the game to provide a broad and diverse array of classes, races and abilities so that choosing these suggested characters does not place the player with a character far weaker than his teammates. How did one play these choices in prior editions? I suggest in 1e, we took a class and lived with the abilities it provided -so all those guys could Backstab and Climb Walls. In 2e, maybe we lucked out and had a kit. In 3e, we still picked a class, and if it was rogue, gained Sneak Attack. 4e? Not my area.

So what SHOULD Next do? I'd like to see robust choices so that we can take:

- Sneak Attack for our urban thuggish thief
- an equivalent ability based on acrobatics for our acrobatic performer, perhaps distraction and damage avoidance
- social skills usable in combat for that charming upperclassman
- gadgets and devices, perhaps alchemical bombs, for a man of science
- minor magical spells and abilities for that wizard school dropout

But if the choice is "sneak attack or better noncombat skills", I think the game fails to provide a platform for these concepts to be effective and efficient, so perhaps it is time to seek out a more robust, option-filled system such as Pathfinder's archetypes, HeroQuest's free form abilities or Hero System's granular point-based system.

 

[Paraxis]

But Indiana Jones is mostly a fighter, and Sherlock Holmes would not be a D&D adventurer.

When the party gathers at the Inn of [Adjective Noun] and meets for the first time by taking up the quest of the week, and they introduce each other, the guy who says "I suck at combat and can't contribute at all to one of the key areas of the adventure." gets left behind. His player then spends the next 30 minutes making a character that doesn't suck.

The game is about big damn heroes going around killing things and taking their stuff, if you want to roleplay something else use a different system than D&D they do that better, don't try and turn D&D into something it is not.

The rules should not be balanced around people making Mary Sue characters because you can't balance that.