Is expertise badly designed?

A partial selection of Bard spells that might not be seen as strictly sitting in back buffing & debuffing for others' glory (not that there's anything wrong with that!): Vicious Mockery, Thunderwave, Heat Metal, Shatter, Stinking Cloud, Polymorph, Animate Objects, Dominate Person, Eyebite, Irresistible Dance, Mass Suggestion, Arcane Sword, Symbol, Dominate Monster, Feeblemind, Power Word Stun, Power Word Kill, True Polymorph.
Then, y'know, sometimes utility spells aren't strictly for others' glory, either.
Yeah, they get a fair number of such options. But they're spell list is far more heavily support oriented than the other casters (arcane casters at least). You can, of course, build classes a wide variety of ways, and a Bard doubly so when they get their magical secrets (if you play that long). But their main damage cantrip has the smallest possible damage die with a debuff rider, and that is the general vibe of the class' magic. The class list is very thematic, and support is a big part of the theme of a class whose iconic ability is to give other people an extra die to roll. It's not that they can't be the MVP in an encounter, it's that nearly any other class will be more often, particularly at lower levels.

In any case, all I'm really getting at is that the particular field where the Bard surpasses all others is in social encounters, and expertise heavily supports this. They are also the class designed to be able to fill in the holes in the party, and being able to invest expertise in skills that nobody else has the ability score to really excel at is part of what makes them suited to this.
 

Beleriphon

Totally Awesome Pirate Brain
Noooo... moving about unnoticed in a crowded urban environment is an incredibly different skillset that relies heavily on things like SEP, social engineering, & more.... None of which is applicable to sneaking into say the den of a mother bear & is largely irrelevant to sneaking into a military fort. Obiwan did the equivalent of casting a charm spell & persuade can't go that far sure, but it can go to absurd levels when you have an absurdly high skill in a sandbox.
Sure, moving unnoticed in a crowd is different, but the principle of not drawing attention to yourself still works. In a crowd as DM if you want to create a distraction that isn't going to involve Stealth proficiency at all, since it isn't about moving stealthily at that point. It isn't about being unnoticed so much as making people look somewhere else.

However, moving around a dark warehouse, streets at night, or another urban non-crowd situation would be different than a forest how? Its about avoiding notice and keeping noise to a minimum. I have no problem with a rogue being very good at being stealthy regardless of environment, since they know how to avoid notice and keep noise to a minimum.
 
Shoving and Grappling aren't. They should have been Saving Throws vs 8+Prof+Str.
It would probably make for better gameplay, but I would caution that the shoving/grappling system is one of the rules I have to explain to people most often (nearly every time it comes up), and that's without them then having to know this "grapple save DC" which unlike a comparable spell save DC is not used for a lot of things. The main problem is that most people (at least most not playing barbarians) shove or grapple once in a blue moon and rarely used rules are the hardest to remember.

Then again nearly every table I've played at had or has someone who can never quite keep straight exactly how bonus actions work and those happen every damned turn.
 

Don Durito

Adventurer
It's a strange thing.

There isn't really a good reason for it to keep getting better. Nothing that Expertise does requires additional scaling. If it's going to be a bonus over proficiency I can't see any good reason for that bonus to change over 20 levels. I can't see, for example why you wouldn't make it a flat +3 (The equivalent of the old Skill Focus feat).

And it doesn't really do it's job that well in many ways. It increases the likelihood of success at a given skill, but there's still a strong possibility of failure (especially early on) at rolls the expert character should still be able to pass). What you really want Expertise to do is give some kind of extra reliability to the roll (the fact that it doesn't do that all that well is why Rogues also need reliable talent).

Really the best alternative would be for Expertise to give Advantage. That's actually the best mechanic to do the job. It doesn't push the results up, and it makes less difference at the top end, but adds a lot of reliability. (4E was bascially doing something like Expertise this way later in it's run). I guess they didn't do that because that way anyone can just get advantage with the Help action, or because it removes that leeway for the GM to just give Advantage for whatever reason. (But of course breaking the bounded accuracy means that the person with Expertise benefits the most from Advantage).

I'm not sure what to do with it - I don't think it's broken - because skills are ultimately a subsystem that it's best to avoid using as much as you can - but I have some issues with it. I'm tempted to put some limitations on what it can be applied to. I don't want anyone having expertise in 'Persuasion' for example, because that just incentivises one person to become the party 'face' while everyone else sits back and stays schtump. (But this is in many ways an issue with the skill itself - nevertheless, where there are issues, Expertise confounds them).
 
Last edited:

Saelorn

Adventurer
Bounded Accuracy for skills is poorly designed. The difference between a master and an untrained novice should be greater than 30% on the binary skill check.

Expertise tries to minimize the damage of Bounded Accuracy for skills, by increasing the difference to 60% for some classes, but that does nothing for any of the other classes.

Reliable Talent is very poorly designed. That's what turns a 45% chance of failure into a 0% chance of failure.
 

Anoth

Explorer
In my mind expertise is for thieves trying to pick locks or remove traps or for stealth like in 1E/2E where those things really need to be near perfect chance of succeed at high levels for those of us that still run games with lots and lots of traps with bad saving throws. I wouldn’t recommend a fighter even trying it in one of my games. The penalty for failing are pretty bad. And it normally means clerics are burning spell slots when they fail. But I am sure the game has changed for many people. I pity not having a trap remover in some games.
 

Mistwell

Legend
It's not just niche protection. I've run into difficulties in my games because of an arcana expertise wizard(who made the artificer look a fool), a stealth expertise AT rogue who never left Sharn* before the campaign (and made the ranger's stealth even in favored terrain look pointless), & a persuade expertise bard(constant "these aren't the droids your looking for" & worse) during different games. All three had 20 in their base stat & all three could expect automatic success against all but plot armor level aided NPCs. People use the arcana expertise example because niche invasion does cause hurt feelings at the table, but it's not the only problems it causes & people have brought up some of them in this thread.

*A place where "nature" basically doesn't exist
It's important to not exagerate in this sort of stuff, and I think you did.

The highest DC (nearly impossible) is 30. A max stat + expertise does not get you "expect automatic success" for that. "these aren't the droids your looking for" & worse" are those nearly impossible checks. I also doubt the artificer "looked like a fool" (sounds like another exaggeration).

Look I know this is the Internet and normally if you don't exaggerate you risk being ignored, but it's not necessary here. You can say what's actually happening without making it seem worse than it really is.
 
Look I know this is the Internet and normally if you don't exaggerate you risk being ignored, but it's not necessary here.
..hmm...? ...what? did somebody say something? ...shrug...

Bounded Accuracy for skills is poorly designed. The difference between a master and an untrained novice should be greater than 30% on the binary skill check.
Expertise tries to minimize the damage of Bounded Accuracy for skills, by increasing the difference to 60% for some classes, but that does nothing for any of the other classes.
Reliable Talent is very poorly designed. That's what turns a 45% chance of failure into a 0% chance of failure.
I think the key thing missing is consistency, and that's kinda the fault of the d20 and it's linear distribution. A master and even a talented novice should perform very differently, but the talented novice might have the occasional stunning success, while the master will have stunning successes so often, they're not stunning at all, and occasionally turn in only an ordinary success, with failure only a rare fluke due to some uncontrollable external factor.

Dice pool systems can capture those sort of things better than d20, if they're done well. But they can be moderately complicated.

One compromise is to roll more dice, but still have a binary success, replace d20 with 2d10 or 3d6, for instance, which have similar averages, but normal distributions.
Conversely, another reasonable compromise is to simply narrate success frequently for trained and expert characters, and call for checks most of the time with untrained ones.
 
Last edited:

ad_hoc

Hero
It would probably make for better gameplay, but I would caution that the shoving/grappling system is one of the rules I have to explain to people most often (nearly every time it comes up), and that's without them then having to know this "grapple save DC" which unlike a comparable spell save DC is not used for a lot of things. The main problem is that most people (at least most not playing barbarians) shove or grapple once in a blue moon and rarely used rules are the hardest to remember.

Then again nearly every table I've played at had or has someone who can never quite keep straight exactly how bonus actions work and those happen every damned turn.
If they are shoving/grappling all the dm needs to do is ask for str mod and proficiency bonus.

If the player is being shoved/grappled all they need to do is make a strength or dexterity saving throw.

It is actually easier for the player and I think neutral complexity for the dm.
 

miggyG777

Villager
..hmm...? ...what? did somebody say something? ...shrug...


I think the key thing missing is consistency, and that's kinda the fault of the d20 and it's linear distribution. A master and even a talented novice should perform very differently, but the talented novice might have the occasional stunning success, while the master will have stunning successes so often, they're not stunning at all, and occasionally turn in only an ordinary success, with failure only a rare fluke due to some uncontrollable external factor.

Dice pool systems can capture those sort of things better than d20, if they're done well. But they can be moderately complicated.

One compromise is to roll more dice, but still have a binary success, replace d20 with 2d10 or 3d6, for instance, which have similar averages, but normal distributions.
Conversely, another reasonably compromise is to simply narrate success frequently for trained and expert characters, and call for checks most of the time with untrained ones.
How about this system

Unskilled: disadvantage ( d20 + Ability Mod )
Novice: d20 + Ability Mod
Proficient: d20 + Ability Mod + Proficiency Mod
Expert: advantage ( d20 + Ability Mod + Proficiency Mod)

Emphasizes the difference between no training and some training. Proficiency means that someone actually is well trained in something and expertise just means he is even more reliable in achieving the results.
Also prevents the players being able to crack very high DCs too early without expertise giving 2x proficiency.

Additionally the varying levels of consistency can be further utilized by using degrees of success.
 
Last edited:

Don Durito

Adventurer
In my mind expertise is for thieves trying to pick locks or remove traps or for stealth like in 1E/2E where those things really need to be near perfect chance of succeed at high levels for those of us that still run games with lots and lots of traps with bad saving throws. I wouldn’t recommend a fighter even trying it in one of my games. The penalty for failing are pretty bad. And it normally means clerics are burning spell slots when they fail. But I am sure the game has changed for many people. I pity not having a trap remover in some games.
I think I'd probably tend to limit Expertise to a range of classical Rogue things. Say any of:

Stealth
Sleight of Hand
Thieves Tools
Disguise Kit
Poisoner's Kit
Acrobatics
 
How about this system

Unskilled: disadvantage ( d20 + Ability Mod )
Novice: d20 + Ability Mod
Proficient: d20 + Ability Mod + Proficiency Mod
Expert: advantage ( d20 + Ability Mod + Proficiency Mod)

Emphasizes the difference between no training and some training. Proficiency means that someone actually is well trained in something and expertise just means he is even more reliable in achieving the results.
Also gets rid of the weird double proficiency progression.
Sensible. I like the Unskilled/Novice distinction.

The last part (Expertise grants Advantage) has been brought up before. The downside is it removes Advantage as a circumstance modifier. For instance, it never helps to Help an Expert (which may explain the attitudes some of 'em have, I guess). That kinda thing. Certainly doesn't kill the idea, but it's a consideration.
 

miggyG777

Villager
Sensible. I like the Unskilled/Novice distinction.

The last part (Expertise grants Advantage) has been brought up before. The downside is it removes Advantage as a circumstance modifier. For instance, it never helps to Help an Expert (which may explain the attitudes some of 'em have, I guess). That kinda thing. Certainly doesn't kill the idea, but it's a consideration.
Yep, have thought about that as well. It adds a bit of complexity but in the case of skill checks I could see that allowing to stack this special "unskilled or expert" disadvantage / advantage with a common RAW advantage / disadvantage. Perhaps one would just have to name it differently.
I believe there even are instances in RAW, where people can throw 3 x d20, such as the "lucky" or "elven accuracy" feat.
 
Last edited:

Todd Roybark

Explorer
[QUOTE="lowkey13, post: 7893329,
*By the way, saying that something has a design principle in mind without stating the design principle in the object itself is hardly ... groundbreaking. Here, let's try it.

"Did you know that U.S. Constitution never, not even once, says anything about separation of powers ... or checks and balances! How you like dem apples????"
[/QUOTE]

Or cynincaly one could say it allows others to posit other design principles not intended, but can not be refutted due to lack of designer notes.

John Marshall was a thief, yo!
 

Don Durito

Adventurer
Yup. Add perception to that list. Possibly investigate.
Investigate yeah. Makes sense. Perception I'm more wary of - everyone wants it and if it's traditionally anyone's niche it's the Ranger's.

I feel like Athletics should be on that list (because of climbing) but then that would eventually make them better at a whole lot of other things than the Fighter or Barbarian.
 

Anoth

Explorer
Investigate yeah. Makes sense. Perception I'm more wary of - everyone wants it and if it's traditionally anyone's niche it's the Ranger's.

I feel like Athletics should be on that list (because of climbing) but then that would eventually make them better at a whole lot of other things than the Fighter or Barbarian.
Eh. Detect noise was a thief skill in ad&d. So i like perception.
 
I'm not sure what to do with it - I don't think it's broken - because skills are ultimately a subsystem that it's best to avoid using as much as you can
How is that not broken?

Heh.... I'm sorry, it just sounded funny... "it's not broken, but don't use it" … er… OK?

But it's not wrong. The "play loop" works very smoothly when the DM judges & narrates success/failure and only calls for checks when it's best for the game. Which, yeah, is using it as little as possible.

'Persuasion' for example, because that just incentivises one person to become the party 'face' while everyone else sits back and stays schtump. (But this is in many ways an issue with the skill itself - nevertheless, where there are issues, Expertise confounds them).
Sure, always been an issue right down to reaction adjustments back in the day. Skill Challenges tried to address it by bringing the whole party into it. A vestige of that remains in the handy Group Check (everyone rolls, if half of 'em pass the DC, success) you can afford to set a less-than-challenging-to-the-Expert DC, since his auto-success won't succeed by itself, the test then turns on rolls of the just-OK folks.
 

Advertisement

Top