D&D 5E Is Expertise too good?

[Yaarel]

I think I have decided what I am going to do.

• Crits and fumbles possible for any skill check
• Proficiency is normal while leveling across tiers (from +2 to +6)
• Expertise equals advantage (two d20s)
• Expertise plus situational advantage equals Accuracy (can reroll one of the two d20s)

Actually, I suspect the math might be better to add d20s.

• Crit/Fumble possible for any skill check
• Proficiency normal
• Expertise is advantage (two d20s)
• Expertise with situational advantage is double advantage (three d20s)

 

[Yaarel]

Now I am still leery about granting expertise advantage to every skill check because I would be leery if it happens in combat. But probably the game can handle it (I will find out). And at least not every skill check benefits from expertise.

 

[Yaarel]

@Yaarel , Just thinking

Combat AC's range from about 11-25. There may be an outlier or 2 outside that range. Most combat AC's range from 13-20. In other words, there's not really a full 20 AC range in what you fight in combat. It's closer to a +15 AC range. Most AC's in combat come out closer to a range about half that. Skill checks tend to go from DC 5 to DC 30. I think part of the issue is that skill checks DC's have to large a range.

In combat everyone gets proficiency bonus to their attack check. It's a given. So stat differences of maybe +2 to +5 or +7 (with archery style) are the only differences players will ever realistically see in their chance to hit differences. So a range of about +5. Skills have a difference in range of -1 to +11 so a range of about 12. That difference in range needs to decrease.

I would suggest we leave proficiency out of the bonus on skill checks and find some other meaning for it, this way our range on skill bonus differences stay around the +5 that combat has it at. Possibly have proficiency add an extra d20 dice that stacks with advantage (allowing up to 3 dice to be thrown at a time)

I suggest we have expertise give a 3rd dice on skill checks. Or maybe a chance of auto success.

Yeah, that is why I focus on the combat-level DCs of 15 and 20. When adjudicating a challenge, most challenges tend to be either DC 15 or DC 20.

If it is a DC 10, I will probably eschew the roll and decide the narrative, unless alot depends on that roll.

If it is a DC 5, I will treat it as an autowin, unless disadvantage or penalties are in play.

DC 25 is special, for olympic level difficulty.

By focusing on either 15 or 20, small bonuses to skill checks can yield dramatic outcomes.

 

[Swarmkeeper]

RAW is that the player always describes their approach, before you even decide whether or not they should roll.

Agreed. Sorry if you felt I implied otherwise.

If there's a hidden panel behind the throne, then "searching behind the throne for hidden panels" will have a much lower DC than "searching the room for hidden doors or traps or something"; and the balancing factor for that lower DC is that they won't possibly find anything that isn't hidden behind the throne.

I suppose what you describe is a fine way to adjudicate the check, but I’d be more inclined to just maintain the DC and ask the player to be more specific OR perhaps say a success just takes a longer amount of time if they weren't super specific which lends itself to more urgency going forward and/or greater chance of “random” encounters. In the end, I suppose it’s a DM style issue - which is fine any which way you slice it.

I still maintain Expertise is really not that big a deal and doesn’t “break” the math. But, if it is impinging on the fun at your table, go ahead and tweak it.

 

[TheCosmicKid]

The skill bonus you have vs the DC check isn't what creates uncertainty. Uncertainty is a DM call. It's part of how the game is played. The DM, not some numbers decide when something is uncertain.

And the DM decides the numbers. If the DM wants something to be uncertain, they put the DC in range of the PCs; if not, they don't. So what does an auto-success-on-20 rule really accomplish?

 

[iserith]

Is it "boosting" the DCs or is that just the correct range of DCs? Why do 10/15/20 make a good set of go-to DCs, but 10/20/30 make a poor set of go-to DCs?

I choose 10, 15, 20 because that's Easy, Medium, and Hard which is fairly intuitive. I also remember in the playtest (which may or may not appropriate for the finished product, I'm not sure actually) there was a sidebar that said you basically can't go wrong by choosing from those three DCs most of the time. So I've stuck with it. As you can see, I'm highly scientific when it comes to these sorts of things.

I had another insight: (and thanks for your questions, they are very helpful in sorting this out) If Expertise were much, much more common -- let's say every character had 1-2 Expert skills, and rogues had 5-6 -- it would bug me much less as a DM. In fact, as a DM, I'd just use the 10/20/30 scale and not worry too much about it. Characters would fail hard checks more often at low levels, but succeed at easy checks more often by mid levels. BUT it would bother me much, much more as a player. I would feel like, if I didn't have Expertise, a lot of checks wouldn't even be worth attempting.

Why do you suppose you'd feel differently about it as a DM versus as a player?

Yes; those are the success rates that are the most fun, in my experience. Generally, if something only has a 5% chance of success or a 5% chance of failure, then when those thing happen, it feels cheap. So I like systems where the "best of the best" can achieve around an 80% success rate and the "worst of the worst" still has like a 20% success rate. Anything more extreme and I feel like we shouldn't be rolling, we should just take the obvious answer and get on with the game.

Did you play D&D 4e? If you did, how do you feel those DCs worked?

 

[Yaarel]

I choose 10, 15, 20 because that's Easy, Medium, and Hard which is fairly intuitive.

I do pretty much the same thing, but I ignore ‘Easy’. For me, challenges are either moderate or hard. Heh, if they are easy, they arent really challenges.

So, if something seems difficult to do, its likely either 15 or 20.

 

[iserith]

The point of the segue to a conversation about DM style, was to give an example of how vital skills are for the DM to adjudicate the freedom and improvisation of players.

Skills (and passive skills) are ubiquitous, continual, and impact the outcome of combat.

Skills are the skeleton that supports the personalities of the adventure.

To wreck the math of skills can disrupt and distort every aspect of the D&D game, from player choices to DM adjudication.

Expertise needs to respect bounded accuracy and stay true to combat math.

Personally, I'm not convinced on that. Even the folks that have an issue with it from a design elegance standpoint don't seem to have any real issue with it in the play experience. From a pragmatic standpoint, I would say the play experience is what matters.

Though I wouldn't have any particular objections on it being changed to something else. It's just not that big a deal to me. The game works fine for me so far.

 

[FrogReaver]

And the DM decides the numbers. If the DM wants something to be uncertain, they put the DC in range of the PCs; if not, they don't. So what does an auto-success-on-20 rule really accomplish?

That's only half the rule. There is also the part where it's auto-fail-on-1.

So let's look at an example. Strong Fighter has 8 dex so a -1 dex mod and no proficiency in stealth. Stealthy Rogue has max dex, max proficiency and expertise for a +17 in stealth. The only DC I can set them at is 19 for the Strong Fighter and Stealthy Rogue to both have a chance for failure. If both characters were trying to perform essentially the same task and you decided you want the outcome uncertain is setting the DC to 19 in every case like that really what you want to do?

The auto-success on 20 allows me to set the DC to whatever I feel is most appropriate and still have every character I want to have a chance at success to have a chance at success. In this example I could set the DC at 25 making the check a little harder for the rogue while still leaving the fighter a lucky chance of success.

You could get the same effect by setting different DC's for different characters depending on their skill bonus but that is generally frowned upon.

 

[iserith]

That's only half the rule. There is also the part where it's auto-fail-on-1.

So let's look at an example. Strong Fighter has 8 dex so a -1 dex mod and no proficiency in stealth. Stealthy Rogue has max dex, max proficiency and expertise for a +17 in stealth. The only DC I can set them at is 19 for the Strong Fighter and Stealthy Rogue to both have a chance for failure. If both characters were trying to perform essentially the same task and you decided you want the outcome uncertain is setting the DC to 19 in every case like that really what you want to do?

The auto-success on 20 allows me to set the DC to whatever I feel is most appropriate and still have every character I want to have a chance at success to have a chance at success. In this example I could set the DC at 25 making the check a little harder for the rogue while still leaving the fighter a lucky chance of success.

You could get the same effect by setting different DC's for different characters depending on their skill bonus but that is generally frowned upon.

I think this goes in part back to my comment about where we think about uncertainty and automatic success in the resolution process. Let's break it down.

1. The DM describes the environment. He or she describes something ahead that can detect PCs if they are not sufficiently stealthy.

2. The players describe what they want to do. The fighter and rogue want to sneak past and describe an approach to that goal.

3. The DM narrates the result of the adventurers' actions. But before I can do that, I have to decide on certainty (auto-success or auto-fail) or uncertainty (roll). Because I have already established that there's something ahead that can detect PCs who are not sufficiently stealthy, I decide it's uncertain. I set a reasonable DC and call for a check. The players roll and the results determine the outcome I can then narrate.

Now, the determination of uncertainty is on the task or, as I would say, the approach to the goal - without reference to the PCs' abilities. Sneaking past this thing by the approach offered by the players is uncertain. It remains so until after the dice fall and a result is determined, regardless of whether the rogue ultimately taps the DC. Similarly, auto-success is when the DM determines that the approach to the goal offered to the players works with certainty, not after the DM has determined it's uncertain and the rogue taps the DC. Resolving the result for the rogue happens further downstream in the adjudication process than when certainty/uncertainty is determined.

In this way of thinking, I don't have to give even a single flumph about who I want to succeed or don't. It's the approach I'm judging and nothing else. Does that make sense?