Proficiency based on class level instead of character level?

[Damon_Tor]

It's weird to me that a guy who spends 20 levels as a rogue is just as good at lockpicking and sneaking around as a guy who spends 3 levels as a rogue and 17 levels doing other stuff totally unrelated to rogue stuff.

I was wondering how you guys would feel about tracking proficiency bonuses based on class levels instead of character levels. The bonus you get from one class would only apply to proficiency you could gain from that class.

For example, if you're a level 10 barbarian and a level 3 wizard, your attacks with martial weapons would have a proficiency bonus of +4, and your spell attacks/spell save DC would have a proficiency bonus of +2. But your attacks with a quarterstaff (with which both classes are proficient) would have a +5 bonus (as normal for a level 13 character).

Skill and tool proficiency rise with the levels of classes in which those are available skills, even if that class wasn't specifically the class that gave you said proficiency. For example, a Rogue 3/Ranger 3 who got stealth proficiency when he took a level in rogue would nonetheless have a +3 proficiency bonus for his bonus because stealth is also a class skill for rangers.

Exceptions are proficiency you get from sources other than a class. For example, weapon proficiency you gain from the Weapon Master feat would rise based on your character level, regardless of which classes you take. The same is true for skill and tool proficiency you get from your background.

[OvisCaedo]

Eh. I don't think of it as much of an issue, myself. Just because someone stopped taking rogue levels doesn't mean they necessarily stopped doing rogue stuff. It means they stopped bothering to learn new roguey things and instead just continued to focus on perfecting the skills they had picked up. With skills especially I also... kind of dislike how held-hostage some of them feel already to classes, expertise especially, so I guess I have a pretty natural opposition to this.

It's also just very messy in terms of tracking, which I think is usually against 5e philosophy, and feels like a very sweeping nerf to multiclassing in general when there are usually only a couple of specific things that tend to cause balance issues.

[Yakk]

So your premise is false.

A L 17 fighter/L 3 rogue with expertise in Thieves Tools and Stealth and 20 dex has a +17 bonus to both. As does a L 20 rogue.

The L 20 rogue will never roll under a 27 on either. The 17/3 can. (Reliable Talent) Once per short rest, the L 20 rogue can get a 37 in one or the other. The 17/3 cannot. (Stroke of Luck) The L 20 rogue (thief) has advantage on stealth rolls while moving slowly (Supreme Sneak).

The L 17 fighter/L 3 rogue has a chance to unlock the lockbox of the gods; the L 20 rogue auto-succeeds. The L 20 rogue's stealth cannot be beaten by any monster in the entire published corpus of work; they get 2 rolls and a floor of 27. The L 17/3 will be spotted by the red dragon passively 25% of the time. If the L 20 rogue really, really needs to sneak by the dragon, they just say "I got a 37" and no active perception check by the dragon can beat it.

How 5e represents "getting better at something from your class" is "more class features to support it".

[clash]

Conceptually I wouldn't tie the character so much to class. The class is the mechanics but not who you are. Three levels in rogue and 17 in wizard night be how you build a spell thief. If your are continuing to pick locks while learning spellcasting wouldn't both skills continue to improve? Mechanically I would never do it either. It makes multiclassing much worse as you can't actually be good at anything

[Damon_Tor]

Eh. I don't think of it as much of an issue, myself. Just because someone stopped taking rogue levels doesn't mean they necessarily stopped doing rogue stuff. It means they stopped bothering to learn new roguey things and instead just continued to focus on perfecting the skills they had picked up. With skills especially I also... kind of dislike how held-hostage some of them feel already to classes, expertise especially, so I guess I have a pretty natural opposition to this.

It's also just very messy in terms of tracking, which I think is usually against 5e philosophy, and feels like a very sweeping nerf to multiclassing in general when there are usually only a couple of specific things that tend to cause balance issues.

As an additional concession to multiclasses, this system would allow you to remove or relax the restrictions on learning new skills when you multiclass. Under the current system most first level multiclasses give you no additional skill proficiency, but if we bind skill proficiency to class level instead of character level we could allow for a multiclasser to be the jack of all trades: they could have a whole mess of +2 and +3 proficiencies, while a single-classed character might only have a four skills at +6.

As to 5e design philosophy, I suppose my argument is that 5e took things too far towards simplicity here: compared to the 3.X system of "ranks" in skills, my proposed system is still much simpler. Think of it as the two editions meeting in the middle.

[MaxWilson]

It's weird to me that a guy who spends 20 levels as a rogue is just as good at lockpicking and sneaking around as a guy who spends 3 levels as a rogue and 17 levels doing other stuff totally unrelated to rogue stuff.

Yes, this interaction is weird, but so is the fact that the guy who spends 3 levels as a rogue and 17 as something else can never get any better at lockpicking, ever. And if he had spent 16 levels as something else instead, it would be enormously difficult for him to get better at lockpicking.

I think if you go down this road you're going to wind up rewriting the 5E multiclassing system entirely, and wind up with something more like AD&D dual-classing.

[JackPhoenix]

There's also the whole thing that your starting character can have proficiencies from race, class, background and feats. And in the case of thieves' (or other) tools, for simply dumping sack of gold to a teacher willing to explain how to do it in few weeks.

Why should the level 3 rogue/level 17 fighter be worse at lockpicking than level 5 paladin who grew up on the streets or a wizard who decided he's got time and money, and doesn't want to waste spell slots on Knock when looting ancient tombs? (let's ignore expertise and ability bonuses for the moment)

[Doug Lampert]

Yes, this interaction is weird, but so is the fact that the guy who spends 3 levels as a rogue and 17 as something else can never get any better at lockpicking, ever. And if he had spent 16 levels as something else instead, it would be enormously difficult for him to get better at lockpicking.

I think if you go down this road you're going to wind up rewriting the 5E multiclassing system entirely, and wind up with something more like AD&D dual-classing.

Agreed: Fundamentally, level 20 characters should be doing level 20 stuff, and that means they need a level 20 proficiency bonus to things they are proficient in. AD&D style dual-classing meant that you could be a level 8/8 rather than a level 9, because being able to choose between two level 8 powers is about good enough to be functional at level 9, but anything much less is too little. 3rd ed style multiclassing pretty well requires that the benefits of the classes stack.

Additionally, any skill can come from a background, which implies that any skill can be trained by any class.

I could see limiting the EXPERTISE bonus this way, or things like Jack of All Trades, but the basic proficiency bonus is tied to level because there are no "class skills" in 5th edition, all skills are available to all characters.

[False God]

IMO, one of the big draws of 5E is that the combination of a lower ceiling, reduced number of skills, and more balanced classes, there's a lot more overlap in the skills department. I think this is good.

If you feel like lock-picking is your special bread and butter, tell that to the rest of your group. The Fighter 16/Rogue 3 may think they're just being helpful by also having really awesome (but as pointed out, not as totally awesome as you) lockpicking skills so that if your character is unable to help pick a lock for some reason, they've got the group covered.

I'm running an exploration-based adventure in 5E currently and I explained this in Session Zero. So now 3/4 characters have several overlapping wilderness-exploration/survival skills. This isn't a problem. They're working together to explore and survive. Which, IMO, is another big element of 5E, "working together".

----
If you want to track proficiency to class level, aside from added complication, it's mostly going to limit multiclassing. Bob will just roll a Rogue instead of a Fighter/Rogue because he can still role-play himself to be more fightery as a rogue. Your problem remains, and now it's worse. Because the slow progression of proficiency means that, at best, Bob is only ever a few points off from you. Yeah, you're the best, but a bad roll on your end and a good roll on his will average things out.

[Rp4man]

Part of a game that did this and it's good. Like real good.

Essentialy put proficiency bonus where you gain an asi in a class. Had to keep track of multiple proficiency bonuses if multiclassed and allowed overlapping skills to get full bonuses if 2 classes would have the same skill. Hasn't been a. Issue.

Asi's and feats. were moved to character level instead of class. Feats are not something you build a class around or via versa. You build a concept around it and sometimes concepts multiclass and it makes sense to be character level.

[Pex]

5E divorced skill use from class. That’s a feature. Anyone can pick a lock. You can gain proficiency just by being a human. You can gain proficiency by having the Criminal background. You can gain proficiency by customized background. You can gain proficiency by training. Picking locks is not exclusive to Rogues. No skill is exclusive to anyone. That’s a good thing.

[Sigreid]

I think you're seeing it a bit off. The Rogue 3/Fighter 17 is a meta mechanic to determine what new abilities are gained. The character still sees himself as a thief that due to his life experiences decided to focus more on learning how to win a fight than learn more of the shady tricks of his trade. In other words, he doesn't see himself as a rogue 3/fighter 17, he sees himself as a thief that knows how to fight. Or alternately, a warrior that knows a few tricks to get himself out of a jam. He's an individual with a collection of abilities not a collection of classes.

[gkathellar]

He's an individual with a collection of abilities not a collection of classes.

Pretty much this.

[Anymage]

Let's say that instead of Rogue3/Fighter7, I decided to go straight fighter with the criminal background. Assuming a Dex fighter and ignoring expertise for the moment, the single classed fighter will be as good at sneaking and picking locks as the rogue is. And you can't really dock a single classed character's proficiency bonus.

If you really want to go nuts protecting the niche of skillmonkeys you can say that the expertise bonus is based on the sum of all class levels that grant expertise instead of a flat doubling of proficiency mod. Your rogue5/fighter15 would only get (6 base + 3 for the expertisable levels) instead of the whole +12. That interacts oddly enough with the Prodigy feat and sounds clunky given the slow rate of proficiency gain, but whatever floats your boat. But trying to have different proficiency advancements for each class sounds like it would get very headachey very quickly.

[False God]

But trying to have different proficiency advancements for each class sounds like it would get very headachey very quickly.

I mean fundamentally it would just be the return of BAB and Fort/Ref/Will saves. A class-based progression to reinforce the idea that the class is only good at XYZ instead of being potentially good at whatever the player wanted them to be. Proficiency may be rather standardized but it certainly leaves the door open to players customizing their characters beyond the typical class parameters.

[BloodBrandy]

Think of it this way.

If someone takes 3 levels in Rogue then goes into Fighter for 17 levels, they still have and use the lock picking tools for that entire time, they still get practice as things progress

[Anymage]

I mean fundamentally it would just be the return of BAB and Fort/Ref/Will saves. A class-based progression to reinforce the idea that the class is only good at XYZ instead of being potentially good at whatever the player wanted them to be. Proficiency may be rather standardized but it certainly leaves the door open to players customizing their characters beyond the typical class parameters.

BAB and save stacking in 3.x created a lot of odd breakpoints. Also, unless you hacked the system so that good save levels stacked to create a final "good save level", dipping multiple classes for the +2s to save was appealing to martial characters. At this point, your attempt to recreate this in 5e will have one of two effects:

-Each class can apply the +2 to anything in its wheelhouse. You can quickly add up multiple +2 proficiency mods and break bounded accuracy.

-You stack levels in classes that have a relevant skill, and use them to compute a total final level for the proficiency value. At this point, the fact that proficiency bonus only goes up every four levels makes this sound like more complexity than it's worth at low levels.

Also of note: The Rogue1/Paladin19 who forsook their roguish ways many levels ago would be best served by a retraining option. There's an idea for one in one of the recent UAs. A Rogue1/Paladin19 who kept sneaking around and breaking into evil lairs is still exercising those skills, and has no reason to see said skills languish or even atrophy. Characters moving on from old things is best handled through retraining, not focusing on the particulars of a class. (Granted, an O-Chul like character going for 19 levels as Paladin and then suddenly becoming a master sneak through grabbing a final level of rogue does seem odd. I just don't think that sort of situation will practically come up often enough to require a rules patch.)

[Damon_Tor]

Think of it this way.

If someone takes 3 levels in Rogue then goes into Fighter for 17 levels, they still have and use the lock picking tools for that entire time, they still get practice as things progress

But the guy who's been a fighter for 17 levels is suddenly as adept as the lifelong thief as soon as he drops a level into rogue.

[Pex]

But the guy who's been a fighter for 17 levels is suddenly as adept as the lifelong thief as soon as he drops a level into rogue.

He's 18th level then. He's learned a trick or two. By learning that trick he didn't improve his damage with maneuvers or be able to recover hit points when seriously injured or be able to attack with his weapon after casting a spell. He also doesn't improve his hit point total as he could have. Unless the game goes epic he'll also never learn to be able to get a 4th attack for his Attack Action.

[False God]

But the guy who's been a fighter for 17 levels is suddenly as adept as the lifelong thief as soon as he drops a level into rogue.

You can frame that level as "oh, he took a 1 hour class on being a rogue, he can't be nearly as good!" or you can frame it as "this one level represents the culmination of lots of practice on the side."

And also, as demonstrated above, he's not as adept as a full rogue. He's good, but the rogue is still better.

[kazaryu]

But the guy who's been a fighter for 17 levels is suddenly as adept as the lifelong thief as soon as he drops a level into rogue.

this. this right here. is why i refuse to pretend that classes have any real narrative meaning. because you get people that pretend that 'fighters' only fight. and that rogues are all thieves. if you're playing a character, as opposed to a munchkin, then you are likely picking up rogue to complement the narrative you're already building. for example, perhaps you're playing a dex fighter and you're grabbing the rogue levels for some better acrobatics, and to build on the swashbuckler archetype.

maybe you're a battlemaster and you want to build on that 'warlord' feel by nabbing the mastermind at-will.

point being from the proficiency perspective, all you're getting from rogue is what...1 skill proficiency? beyond that all you're doing is improving on what you already have. or you're picking up a new suite of abilites at their lowest functionality.

now, this *is* a class based game, which means that most of the time when you multiclass for narrative you're going to pick up abilities that don't really fit. (for example, thieves cant). but thats not really related to the way proficiency bonuses work.

going back to your original statement. as has already been pointed out in this thread. no. you're not as 'adept' as soon as you put 1 level into it. you lack literally all of the class features, except for expertise, that make the rogue a rogue.

your sneak attack is much lower. you lack almost all nearly all of the defensive combat abilities. (i say nearly because technically expertise/proficiency in acrobatics/athletics is a combat ability if you were to take it)
you don't have abilities like reliable talent, which is where ALOT of late game rogue power comes from (out of combat)

so no, you're not suddenly 'as adept' as a lifelong anything.

[BloodBrandy]

But the guy who's been a fighter for 17 levels is suddenly as adept as the lifelong thief as soon as he drops a level into rogue.

Given the amount of Exp it takes to get that 18th character level, no, it's not 'sudden' at all.

Also, if you are a Strength based Fighter, you are, at best, as good as a level 1 rogue, a newbie, which you are, and you are nowhere near as good as a Rogue who is at level 17 like you are in Fighter.

If you are a dex based fighter, then you are still crappy compared to a level 17 rogue, who gets features to compliment their abilities. All you are is a fighter who learned to apply your deft hand to picking locks...

...you know, like ANY dex based character who has gotten the thieves tool proficiency through background, through a feat, or through a class feature, like, say, the Battlemaster's Student of War feature.

Yes, you can get Thieves Tools from things OTHER than taking a level in Rogue, surprising I know, but you can. You can take the skilled feat (Or Prodigy if you are a human or half human), you can train with it under Xanathar's Guide rules to gain proficiency, you can get it from the Urchin or Criminal backgrounds (Or a custom background if your DM allows), and guess what? Your Proficiency will also be the same as the rogue. Hell, through the Prodigy feat you can have expertise in it, or if you are a bard.

But you know something? a Rogue will still end up better than you because Proficiency mods are not the be all end all of skills. You won't be getting Reliable Talent so you can still roll under a 10 on the D20, you can crit fail on a lock pick. You can be a wizard or eldritch knight, but won't have the same fine control an Arcane Trickster has with the Mage Hand to use the tools through them.

No, you will NOT be anywhere near as adept as a lifelong thief

[Sigreid]

I've never seen anyone wait until level 17 for their 3 level dip.

[Keravath]

I think the OP is mixing the ideas of mechanics and roleplaying with the idea that proficiency could be based on class level rather than character level.

Could you play the game this way? Sure, a DM can play the game any way they wish, assuming they explain it to their players. However, the "unrealistic" or "how can a 3 rogue/17 fighter be just as good as a 20 rogue at picking locks" are roleplaying arguments which are meaningless in the context of mechanics.

From a mechanical roleplaying perspective, the character is the sum total of their personality and the mechanics granted by their class level choices. However, characters don't fit into neat little boxes especially when they multiclass. A fighter could start with a criminal background, be proficient with stealth, thieves tools, deception and make an excellent thief character. The only thing they really lack is expertise but they can easily be roleplayed as a thief. So can ANY other class level combination. Any dex based character can be played as a rogue/thief character archetype with varying abilities depending on which class combination when in to making up the character. There is no reason to assume that ANY skill they have and could practice would not increase as the character experience level increases no matter which choice of class level they make along the way.

The only time I can see a bit of narrative continuity is if the level 19 character who has focused on being a fighter decides that at level 20 they would like some rogue abilities added to their character mix and take one level of rogue at level 20 with expertise in stealth and thieves tools. Let's assume that they weren't previously proficient with either of these before taking the level of rogue. Their modifier goes from +5 (non-proficient 20 dex) to +17 (expertise 20 dex). However, this is mechanics nothing more. This is a magical roleplaying game and there are many roleplaying or magical narrative explanations that can be used to justify this change in skills from an in-game world narrative perspective. It is a big change mechanically but perhaps the god of thieves smiled on the character granting them the abilities.

Basically, class levels provide features that allow a character to have a wide mix of abilities and characteristics. A level 3 rogue never has more than +2d6 sneak attack, they don't have evasion, uncanny dodge, reliable talent ... however, the pure rogue doesn't have 3 attacks/combat round, indomitable, action surge, second wind and all the other features granted by the fighter class levels. The character is NOT their class. The abilities of the character are defined by the character level choices but it is how the characters chooses to play and be perceived that defines the character, not the class.

So the underlying assumption that a 3 rogue/17 fighter spends 17/20 of their time being ONLY a fighter and so should only get proficiency in fighter skills and abilities for those 17 levels is fundamentally flawed in my opinion.

[Fable Wright]

If you don't want people to multiclass this badly, why not just ban multiclassing and be done with it? :smallconfused:

[Randomthom]

Also, what of the guy who gained lockpicking from his background, does he surpass the rogue who stopped at 3 or is he forever locked at +2 proficiency for the background skills unless he picked a class that also allows that skill?

Leave it, it's fine :)

[Demonslayer666]

If you learn a new skill regardless of level, you should start out as a novice and progress from there, but I don't think 5th is broken because of this. I can overlook it for simplicity's sake.

[Nagog]

As an additional concession to multiclasses, this system would allow you to remove or relax the restrictions on learning new skills when you multiclass. Under the current system most first level multiclasses give you no additional skill proficiency, but if we bind skill proficiency to class level instead of character level we could allow for a multiclasser to be the jack of all trades: they could have a whole mess of +2 and +3 proficiencies, while a single-classed character might only have a four skills at +6.

As to 5e design philosophy, I suppose my argument is that 5e took things too far towards simplicity here: compared to the 3.X system of "ranks" in skills, my proposed system is still much simpler. Think of it as the two editions meeting in the middle.

To be fair, I haven't played 3.X, but I have played Pathfinder which sounds pretty similar in that regard.

Tbh, I think this heavily screws with the Bounded Accuracy system. In many ways, DCs in 5e are incorporeal. They're made up by the DM, but in other, more specific ways, they aren't. For example, many illusion spells require an Investigation check against the caster's Spell Save DC. That is a set DC. Some magic items are the same way. On the same note, save DCs and AC (from attack rolls that also require proficiency) also scale out. So things like a martial caster via multiclassing (such as the popular Paladin/Sorc) would flop at higher levels because you'd have a pitiful to-hit bonus unless you used weapons that both classes are proficient with. Similarly, if you took 2 paladin levels to start then dove right into Sorc for the spell slots, your Wisdom saves would never recover. The only reliable thing to gain from such a multiclass would be Heavy Armor from the first level of Pally.

All in all, this change would negate any usefulness to multiclassing whatsoever, unless the classes were nearly identical in the skills, saves, weapons, etc. that they grant you, and considering those are some of the biggest reasons for class dips, it would kill the prospect of multiclassing in ~80% of cases.

[redwizard007]

As many others have pointed out, the mechanics of this change are do-able, but are needlessly complex and anti-5e in nature. Beyond that, your RP reasons are flawed. Characters aren't "insert class name," they are living, breathing (usually,) individuals.

Example 1: Bob is a priest in the temple of Pelor. He took 20 levels of wizard. That doesn't make him a "wizard." He doesn't wear a pointy hat or carry a staff. He knows all the rituals, and prayers, but he avoids armor, a handful of weapons, and he casts different spells from most of his fellows.

This gets even murkier if Bob took 20 levels in divine soul sorcerer or paladin.


Example 2: Tom is a scout for the army. He has levels in rogue, ranger, and fighter. From session zero this was the plan for his build. Why should his skills be penalized for following a different build than a single class rogue?