Difficulty Numbers: Scaling, or Static?

[Reynard]

I'm not really sure what you mean with "the numbers only go up". An RPG could model Vonn's injured ACL as a flat penalty on the die or some equivalent of disadvantage on the check. Various RPGs take things like injuries and deviations from fitness into account, but most don't because most don't want to get that far into the weeds. And I would wager most players don't want to either.

I mean how actual peak capacity works, for both mind and body, and as how we get older both begin to fail us even when our level of experience should see us through.

 

[Whizbang Dustyboots]

Numbers scaling too much is one of the big problems Castles & Crusades has under the hood, where high level characters are sometimes more vulnerable to effects than lower level characters are.

So as a general rule, I'd rather use static numbers so that higher skilled characters can beat them over time.

 

[Staffan]

What do you think? How do you liked games to handle difficulties?

I think DCs should correspond to actual difficulties, but the game should also have guidelines about what sort of DCs correspond to desired challenges at various levels. So part of the game should say "a moderate challenge at level 10 should have DC 25", and another part of the game says "The DC to climb a rough surface like a brick wall is 25". So this would tell me that if I want a moderate climbing challenge, that should be a rough surface like a brick wall.

I'm going to throw in a related concept - if you're an expert in something you should (almost) never fail.

I agree. I think Monte Cook had an idea along those lines for those five minutes he was contracted to work on 5e. I don't recall the specifics, and I doubt I could find it over on the Wizards site given that they hate keeping old stuff around for more than ten minutes, but it was something like this: PCs have skills at various proficiency levels, and the world provides challenges at various difficulty levels. If your proficiency level is higher than the challenge level, you automatically succeed. If it's lower, you automatically fail. And if it's the same, you need to roll against the appropriate stat. Personally I think this is a little too deterministic, particularly when things are too difficult – I'd like to see some chance of rising to the challenge, but I like the basic concept.

PF2 had the opportunity to do something similar with its proficiency levels, but they seem to prefer using them negatively rather than positively. That is, instead of saying "if you're an Expert in Thievery, you can automatically open this lock" they'll say "If you're not an Expert in Thievery, you can't try to open this lock."

 

[Tonguez]

Static.
3e is the best edition for laying out what a difficulty class was and why for specific tasks (which stands to reason since players having well-informed options was that edition's mantra).

I agree DC should be static with clear advice and examples on how to set a DC for any tasks the PCs might face.

I never understood the point of scaling DCs, as PCs get better at doing a task it shouldnt get more difficult by default.

 

[Lanefan]

Static, in the name of setting consistency.

The cliff by the bay that's DC 20 to climb at 1st level doesn't change itself just because the PCs are now 20th level - it's still DC 20.

That said, I prefer if "experts" always have a small chance of failing and "rookies" almost always have a small chance of success. 1/20 of the time (i.e. nat 1 always fails, nat 20 always succeeds) might be too often for this in some cases, but it's still better than absolutes.

 

[djotaku]

Most things in life are not random. Vaugn did not "roll a nat 1" -- she went out there when she was not up to the task. Yes, she was an amazing athlete, but she is past her prime and did not want to accept that. You see that a lot in sport.

It is actually too bad few games capture that part of it. I bet lots of old adventurers would bite it on "one last run into the dungeon before retirement."

One of my favorite Discworld stories is when Cohen the Barbarian and the other geriatric heroes have one last ride

 

[djotaku]

PF2 had the opportunity to do something similar with its proficiency levels, but they seem to prefer using them negatively rather than positively. That is, instead of saying "if you're an Expert in Thievery, you can automatically open this lock" they'll say "If you're not an Expert in Thievery, you can't try to open this lock."

Good point there. I think it's interesting to consider when TTRPG rules tell you what you can do and when they tell you what you can't do.

 

[Mustrum_Ridcully]

Ultimately, this is 5e in a nutshell. This edition encourages the DM to make a determination of Yes/No based on what the PC brings to the table and the nature of the task, and then suggests rolling only when Yes/No is unknown/unclear. So yeah, don't make the priest of Osiris make a roll about stuff they should know about - they just know it.

That reminds me of something from Night's Black Agent (or Gumshoe as underlying core system?) had: You had Investigative skills where you don't roll for. If your have an investigative skill that fits the situation, you tell the GM and he gives you the relevant clues. You may be able to spend a point to get some sort of advantage. So, the Priest of Osiris would know all the relevant stuff about Osiris, but maybe he wants to spend a Religion point to also remember a ritual that allows them to calm down some guardian constructs of an Osiris Template Site or something.

The basic idea here is that the game is about investigation, and investigation adventures where finding the neccessary clue is locked behind skill checks, you might end up with it being impossible to solve the mystery because of bad luck.

But this wouldn't apply to something like Climbing or Skiing. (Though even there, skill checks are roll 1d6 vs some DC, and you can spend points from your non-investigate skills to gain a bonus to the roll, so the Free-Climbing expert can probably reliably scale a bunch of walls before he's out of his depth.)

 

[billd91]

That reminds me of something from Night's Black Agent (or Gumshoe as underlying core system?) had: You had Investigative skills where you don't roll for. If your have an investigative skill that fits the situation, you tell the GM and he gives you the relevant clues. You may be able to spend a point to get some sort of advantage. So, the Priest of Osiris would know all the relevant stuff about Osiris, but maybe he wants to spend a Religion point to also remember a ritual that allows them to calm down some guardian constructs of an Osiris Template Site or something.

The basic idea here is that the game is about investigation, and investigation adventures where finding the neccessary clue is locked behind skill checks, you might end up with it being impossible to solve the mystery because of bad luck.

But this wouldn't apply to something like Climbing or Skiing. (Though even there, skill checks are roll 1d6 vs some DC, and you can spend points from your non-investigate skills to gain a bonus to the roll, so the Free-Climbing expert can probably reliably scale a bunch of walls before he's out of his depth.)

Even with climbing or skiing, there may be some variation - for example, someone with an alpine adventuring background vs a bookworm from the city library. The alpine adventurer should be able to do things the bookworm can't and situations difficult for the bookworm would be routine for the alpinist.
But with respect to the investigative or informational skills, I maintain a list of which knowledges PCs are trained in when I DM 5e. And I use that as the first thing I consult when questions come up about what a PC would know.

 

[stonehead]

Personally, I prefer static DCs as framed by the original post, as it seems like everyone else does.

I do think though that scaling DCs are kind of an innevitable consequence of a high-variance resolution mechanic like a d20, and high power level scaling.

@djotaku makes the important observation that

if you're an expert in something you should (almost) never fail.

Which is true. A professional rock climber will (almost) always succeed at climbing walls that an amateur like me will always fail at. In fact, a professional will probably always succeed at tasks that a hobbyist will always fail at, and the hobbyist will (almost) always succeed at tasks that an amateur will (almost) always fail at.

For a d20 system to accommodate this, a regular professional would need something like a +35 modifier to their roll. If the system wants to facilitate a zero-to-hero story arc, they need to scale their bonuses from around +5 to around +35 as they level up. That means an interesting challenge for a high-level character will be completely inacessible for a low-level character, and an interesting challenge for a low-level character will be trivial for a high-level character. So in order for the game to stay interesting, the DCs will need to scale with the PCs' levels. Hopefully, there's some in-universe justification for this, like low-level characters climbing fences, and high-level characters climbing smooth faced buildings. The low-level thief steals from peasants in the dark, and the high level thief steals from the King in broad daylight.

Some alternatives are narrowing the range of play, using different dice, and auto-succeed.

Some games narrow the range of play, having PCs start out as competant hobbyists, instead of amateurs with potential. This works pretty well as long as you don't care about the zero-to-hero arc. Depending on implementation, it also runs the risk of flattening out character diversity. If every character is at least competant at everything, there's less to differentiate my character from yours.

Other games just use different dice. With a d20, a +5 bonus turns your 50% chance of success to a 25% chance of success. With 3d6, that +5 turns 50% into a ~95% chance of success. Because of the bell curve, you reach "almost always succeed" a lot sooner, and flat DCs stay relevant a lot longer. Some groups like the unpredictability of the d20 though and like seeing exciting crazy results on a nat 1 or nat 20.

Finally, there's the auto-succeed. Some games, explicitly or implicitly, tell the GM to just not roll for easy tasks, regardless of what probability of failure the dice would give. There's always a subjective judgement call here, and in my opinion, it "feels" kind of inelegant, but it works for plenty of groups.