D&D 5E Is Expertise too good?

[Ristamar]

... rogues just run faster, every turn, than anyone else (high level monks excluded). If your concept has an aspect to be really good at a certain skill, then rogue (or bard) becomes highly incentivized regardless of the other aspects of your concept. Same with being fast -- rogue has abilities that strangely result in being able to run faster than anyone else.

The Rogue's bonus action Dash on the battlefield is an abstraction of the class' nimble movement, not a representation of pure speed. Like many other aspects of D&D, that abstraction can lead to ridiculous outcomes without proper adjudication. Once an encounter becomes a matter of evasion and pursuit, the bonus Dash should not factor into the resolution mechanics.

 

[iserith]

I'm a design buff, largely because design incentivizes play. When design starts to distort incentives, I dislike that design. Expertise distorts incentives by being so attractive in the all-or-nothing skill arena. When it's a do or don't roll that can have big consequences (not seeing an ambush, tracking successfully, hiding from all the sentries, etc.) then the availability of such a distorting mechanic twists the incentives of play. For example, I have a ranger character in my current game and their player has dipped rogue just for expertise and the bonus action options because both of these things mean they can be a better wilderness warrior. That's just weird that it's the rogues who can be better at any skill than even the "smart" class and that rogues just run faster, every turn, than anyone else (high level monks excluded). If your concept has an aspect to be really good at a certain skill, then rogue (or bard) becomes highly incentivized regardless of the other aspects of your concept. Same with being fast -- rogue has abilities that strangely result in being able to run faster than anyone else. That kind of distorting pressure annoys me. I can easily deal with it in game -- rogues just do these things. But, from a design perspective, expertise shattered bounded accuracy, which is an annoying design choice, especially when the concept is already better done by both the advantage and the reliable skill rules. Expertise is just number inflation, and that's boring design especially in a system whose design goals include tamping down number inflation.

So, in play, whatever, I deal the game is fun and we all have a good time. When I put on my design hat, I hates expertise because it doesn't fit.

I would probably lay some of this at the feet of multiclassing then, which is an optional rule anyway. If I include multiclassing in my game, I just have to assume this sort of dipping will be the norm.

 

[iserith]

Oh, that's an interesting way to think about it.

For me: most characters don't have Expertise, so I don't think about it when setting the DC, and if a character busts out Expertise, it feels a little like they "cheated" somehow. Conversely, if I DO think about Expertise when setting DCs, then I have to expand my range of DCs from 10-20 up to 10-30. This leaves a lot of characters in the dust. Almost everyone has a small chance to hit DC 20; almost no one has a chance to hit DC 30.

I think there is a weird urge for some to boost DCs because of Expertise. There's really no need to though, right?

Question: How do you set DCs?

Thinking about this more, I set DCs based on the expected range of PC modifiers, and nothing more. I don't have any notion of "realism" or any scale independent of the scale determined by the expected range of PC modifiers.

That scale is "supposed" to be -1 through +5 at 1st level, up to -1 through +11 at 20th level, with -1 through +9 as a healthy range at mid-levels. So when I say "DC 15" what I really mean is, "about a 55% chance for a 1st-level specialist to succeed, or an 85% chance for a 20th-level specialist to succeed, and a 25% chance for someone dump-statting to succeed." Expertise distorts this scale; by level 9, an Expert can get this up to 95% success, and by level 13 and up, it's 100%.

Unless it's a contest, for me it's usually just 10, 15, or 20 based on my quick assessment of the approach to the goal with an eye toward consistency with past rulings. If someone has a bonus that makes those three go-to DCs trivial, hey, no big deal in my view. Every now and then, under specific circumstances a particular task might be a DC 25 or 30, but that's rare. And that seems about right.

For your scale, did you set those success benchmarks on anything in particular?

Neither. They just talk to the NPC, and if I think their character would detect something their player doesn't, I tell them to roll Insight. The players just do stuff, and I tell them what happens, and the dice are there to help me determine outcomes with less effort and less boring predictability.

I would not ask for a roll if the players didn't describe what they wanted to do (and I found their approach to the goal uncertain). As a player I actually get kind of annoyed when I'm asked to make a check without declaring an action because the DM just assumed my character did something. I only get one thing to do in the basic conversation of the game as a player after all.

It happens more in other RPGs than in 5E.

But an example in 5E might be an NPC with Perception Expertise and a Passive Perception of 22. Suddenly the mid-level rogue with Stealth Expertise has a chance of sneaking past. But if the whole party wants to slip past, then they are SOL, because it's literally impossible for the clumsy paladin to succeed, even with advantage! Now a clever party can still use magic to make it past (guidance, bardic inspiration, etc.) but they have to know in advance what they are facing, and it's still super not gonna happen. Conversely, without Expertise, the NPC might have a PP of 18. That's still devilishly difficult, but it's possible straight-up, and if they apply all the magic, it might even be likely. Group checks (where only half the party needs to succeed) make this particular situation better, but it's still a little problematic.

Let me turn this around: What if Experties tripled your proficiency? quadrupled? What if it were a flat +10, right from first level? Where do you draw the line? In the first 5E playtest, there was no proficiency; the math was totally flat between characters, and they all felt kinda samey. In 3E, it wasn't hard to get a 20-point difference between characters by mid-level, and that was demonstrably un-fun for lots of people.

One of the things I love about 5E was that it tightened up the quantitative difference between characters in order to focus on the qualitative, and Expertise works against that. I understand the need to make rogues better at skills than everyone else. I just think they should have found some non-success-rate-based way to do it.

Oh okay. I thought you meant the Expert auto-failed somehow and I couldn't figure that out. It's really about everyone else. I'm kind of in the camp of "tough luck" there, "can't win 'em all." Or maybe they can if they spend enough of the right resources. That's their problem, not mine.

 

[iserith]

Can I explain how? It looks obvious to me.

I believe that [MENTION=58172]Yaarel[/MENTION] is working under the impression that his player is suggesting the character's approach is "by reading the guard's demeanour" or some such. Because that implied approach seems obvious to him, I think he's reading your posts as though it's obvious to you, which means that when you talk about needing an approach, you must mean something more than the implied "by reading the guard's demeanour."

In trying to figure out what that something more is, he's guessing you want the player to reference the mechanics to some degree because the difference between what you and Yaarel are each actually saying is so small.

Cause really all you're saying is 'I want the player to explicitly say "by reading the guard's demeanour" while he's okay with the implied declaration.

Yeah, that could be so. "By reading the guard's demeanor" might be viable in some situations and not in others. There are many ways to suss out an NPC's agenda depending on the scene. The bottom line is avoiding situations where the DM is assuming what the character is doing which is often what happens when a player asks a question and the DM narrates the results of the adventurer's action without clarifying the character's approach to the goal. As DM, I don't want to play the characters for the players and I don't want my DM to do that to my character and that is what assumptions of approach are in my view. This is avoided when the players and DM perform their respective roles.

 

[Ovinomancer]

I would probably lay some of this at the feet of multiclassing then, which is an optional rule anyway. If I include multiclassing in my game, I just have to assume this sort of dipping will be the norm.

Banning multicasting because expertise is to attractive to dipping is exactly the kind of distortion in talking about.

 

[Ovinomancer]

The Rogue's bonus action Dash on the battlefield is an abstraction of the class' nimble movement, not a representation of pure speed. Like many other aspects of D&D, that abstraction can lead to ridiculous outcomes without proper adjudication. Once an encounter becomes a matter of evasion and pursuit, the bonus Dash should not factor into the resolution mechanics.

That's a nice house rule, but if doesn't follow the rules. And rogues are still the fastest in combat even if they strangely slow down in other situations.

My fix for that feature is that rogues can't use their bonus action to dash if they use their action to dash. This way rogues go from the fastest class to the class that can run while doing something else.

 

[Satyrn]

That's a nice house rule, but if doesn't follow the rules. And rogues are still the fastest in combat even if they strangely slow down in other situations.

My fix for that feature is that rogues can't use their bonus action to dash if they use their action to dash. This way rogues go from the fastest class to the class that can run while doing something else.

Just out of curiosity, do you still allow the fighter to Dash twice on his turn if he uses an action surge? I think I used my fighter's action surge for that more than anything else (because I found it wildly underwhelming without Extra Attack and my fighter died early into his 5th level)

 

[Ovinomancer]

Just out of curiosity, do you still allow the fighter to Dash twice on his turn if he uses an action surge? I think I used my fighter's action surge for that more than anything else (because I found it wildly underwhelming without Extra Attack and my fighter died early into his 5th level)

I'm not sure how the two are comparable.

 

[FrogReaver]

Anyways, getting back to the thread topic.

Maxing stat and having proficiency in a skill gives you a +11. You can't roll below a 12 in that case. I don't hear complaints about auto successes occurring to much with max stat and skill proficiency. Perhaps it's not that auto successes can happen with skill checks it's that they should be "rare" and hard difficulties like DC 15+ should be autopass even more rarely.

Expertise does become an issue there because you can get over 15+ on your DC.

That said, I think the solution is to count 1's as misses on all skill checks. Then there is no more auto success. Count 20's as auto pass and there is no more auto failure. I think that should take care of most issues. If a 5% guaranteed pass / guaranteed fail rate seems to small then do 1-2 is auto miss and 19 and 20 is auto success.

Let some class abilities or feats play around with the auto success and auto miss numbers. I think something like this would be the easiest solution to adapt to 5e.

 

[GreyLord]

Instead of punishing the player with expertise reward them for having it as a perk. Don't artificially increase the DC's...keep the DC's as they are even if it means the player passes it easily or doesn't even have to try.

This is a PERK of their class that they choose and they should be able to enjoy this perk. Instead, let them know that even greater things are possible. Let the PLAYER decide to try to do these impossible things, and set the DC's accordingly. This is ALSO a reward of the player's class and the perk they choose. They can actually CHOOSE to try to do these impossible things and stand a chance of succeeding.

Rather than punish the player and the party by inflating existing DC's, reward them for their choices and their perks. Give the PLAYER choice and also let it be a reward where THEY CAN CHOOSE to try things not normally possible to others and enjoy the class that they chose.