Breaking the Impossible Skill DC

[carrdrivesyou]

So a player in my campaign managed to roll a 41 on their skill check...What does that constitute in terms of actual success? Also, this raises the question of how often does this come up? How do you all handle this in your games?

[JackPhoenix]

It means they've succeeded on whatever that check was. That's it. Whatever they were trying to do wasn't impossible, otherwise, there wouldn't be a roll in the first place, as attempt to do something impossible automatically fails, no matter how high can you get your modifier.

[Contrast]

DC30 is nearly impossible. DC40 wouldn't be impossible but rather 'no but seriously this is really really nearly impossible guys'.

It isn't going to happen a lot (you're generally looking at some combination of expertise and either Pass Without Trace or bardic inspiration).

I can't really think of a situation in which I would realistically set a test at DC40, but if you want some ideas:
Tiamat rolling a nat20 on perception would need a 47 stealth to sneak past.
A roll of 43 on Arcana means that you understand magic better than Acererak possibly can.
A 39 on Intimidation means you're scarier than its possible for Zariel to be.
A persuasion roll of 40 means you're more silver tongued than any of the famous devils/demons/anything.
And so on.

Basically, if you're rolling in the 40s, you're probably the very best on any plane at what you're doing or at least in the top few.

[Naanomi]

Isn’t there a door in one of the early 5e dragon modules that takes a DC 40ish check to break down? (Edit: DC 70!)

[Battlebooze]

There are impossible skill checks. Player: "I want to make a running jump across the Grand Canyon! I got a 40 on my athletics check!"
GM: "You make a leap for the record books but you don't make it to the other side, you are short by a half mile. Still, everyone watching is deeply impressed by your effort. Take 20d6 damage."

There are "Impossible" skill checks. Player: "I'm going to convince Asmodeus to stop the Blood War. I roll a 40 on my persuasion check."
GM: "Asmodeus is so amused and impressed by your argument that he agrees to a one day cease fire. He also invites you to a hellish party, hoping you entertain other devils with your impressive and amusing wit."

There are "Beyond normal measure" skill checks. Player: "Before the red dragon finishes off my character, I recite the poem we found, the one about the regrets of a dying flame. I roll a 40 on my performance."
GM: "The dragon's great maw is ready to devour you but she stops and listens. The dragon then strikes and everything fades to black. You wake up in the middle of a wheat field, bruised and hurt like before, but otherwise unharmed. There is no sign of the Dragon."

There are things that no roll can overcome, and things that perhaps... with a roll that the gods would appreciate, might get you somewhere.

[Unoriginal]

So a player in my campaign managed to roll a 41 on their skill check...What does that constitute in terms of actual success? Also, this raises the question of how often does this come up? How do you all handle this in your games?

The highest DC possible is 30. Succeeding DC 30 by 1 or by 11 doesn't change much, overall.

Now, if it was an opposite check, that would matter, because plenty of PCs and NPCs can have +11 to a roll, allowing them to roll above 30 if they get lucky.

DC: Impossible is impossible, there is no way to do it. In fact a DM shouldn't ask for it.

[Naanomi]

The highest DC possible is 30.
We have examples of modules going higher than this (and there is DCs set by other people’s skill checks or the like)

[OldTrees1]

DC 30+ are things that are nigh impossible in the campaign world, including many things that are impossible IRL.

What does that mean? Well that is up to the DM. Is a DC 30 meant to be a Tier 1 or a Tier 4 feature? If you limit checks to Tier 1, then look at superhero movies and real life. If you want them to be Tier 4 features, then assume it is something that is impossible IRL but possible for extraordinary skill masters in the campaign world.


So what were they trying to do? How powerful / weak do you want high level skill use to be? Rogue's signature high level ability is reliably hitting DC 20-30. That is what they get in place of 9th level spells. So how significant should DC 30 at will be?

[Unoriginal]

DC 30+ are things that are nigh impossible in the campaign world, including many things that are impossible IRL.

What does that mean? Well that is up to the DM. Is a DC 30 meant to be a Tier 1 or a Tier 4 feature? If you limit checks to Tier 1, then look at superhero movies and real life. If you want them to be Tier 4 features, then assume it is something that is impossible IRL but possible for extraordinary skill masters in the campaign world.


So what were they trying to do? How powerful / weak do you want high level skill use to be? Rogue's signature high level ability is reliably hitting DC 20-30. That is what they get in place of 9th level spells. So how significant should DC 30 at will be?

DCs results don't change in function of tiers, in 5e. What is DC 30 at lvl 1 is DC 30 at lvl 20, all circumstances equivalent.

[Zhorn]

I can't really think of a situation in which I would realistically set a test at DC40, but if you want some ideas:
Tiamat rolling a nat20 on perception would need a 47 stealth to sneak past.
A roll of 43 on Arcana means that you understand magic better than Acererak possibly can.
A 39 on Intimidation means you're scarier than its possible for Zariel to be.
A persuasion roll of 40 means you're more silver tongued than any of the famous devils/demons/anything.
And so on.

Basically, if you're rolling in the 40s, you're probably the very best on any plane at what you're doing or at least in the top few.
Best to not think of the check results as absolutes. More reasonable to treat them as "the best possible attempt anyone could have made"
A roll of 43 on Arcana doesn't means that you understand magic better than Acererak possibly can, but rather you had a genius moment of understanding on a thing at the time.
A 39 on Intimidation doesn't means you're scarier than its possible for Zariel to be all the time, just circumstances of the moment meant even the bravest/arrogant of foes would still give pause and act with caution towards you.
A persuasion roll of 40 isn't that you're more silver tongued than any of the famous devils/demons/anything, you just have a very convincing argument that's hard to dispute.


I think the wording choice was a mistake. Getting a 30+ isn't 'impossible' or 'nearly impossible' under 5e's math, it's just in a masterful range of result that cannot be achieved without some degree of specialisation or external aid, but once you have either close-to-max stats and/or expertise, you'll see results of 30+ start to happen more and more frequently.
Be it in the books or listening to others talk about DCs, I can't help but roll my eyes a little when I hear "a 30+ check is impossible"

[carrdrivesyou]

To add context, the artificer was trying to make something from raw materials. She guidances herself, then gave herself a flash of genius whilst working. They rolled high and the end result was a 41. So, the device she made was inhumanly perfect. The DC for success was 20, very easy to do for her, but she more than doubled that. In the spirit of rewarding effort and character investment, how would you lot handle that? Improve the device? Increase it's efficiency? Leave it alone?

[Mellack]

She passed the DC. I wouldn't change anything for beating the DC by 1 or by 21. Just give whatever you would have for a success and maybe a statement about how easily everything worked out for her. Same thing if they were trying to shoot a basketball into the hoop. You might just make it or you might swish it hitting nothing but net, it still gets the same points.

[Unoriginal]

To add context, the artificer was trying to make something from raw materials. She guidances herself, then gave herself a flash of genius whilst working. They rolled high and the end result was a 41. So, the device she made was inhumanly perfect. The DC for success was 20, very easy to do for her, but she more than doubled that. In the spirit of rewarding effort and character investment, how would you lot handle that? Improve the device? Increase it's efficiency? Leave it alone?

Are you talking about the partially finished device you can find in one of the quests in Rime of the Frost Maiden, and the PCs have to finish it?


Crafting doesn't require ability checks, usually, so I am asking as it is the only instance of this I can recall.

If the DC was 20 and she had at least +21 for that check, it was an auto-success, no roll needed. Personally I would have said something like "you don't need to roll. [artificier] immediately accomplishes what would take most experts hours if not days to figure out".

If you want to reward the PC, I would say giving her the magic item's formula or just say she has enough knowledge to replicate the engine (if not magical) could be appropriate.

[OldTrees1]

DCs results don't change in function of tiers, in 5e. What is DC 30 at lvl 1 is DC 30 at lvl 20, all circumstances equivalent.

Agreed, yes the DC 30 is DC 30 regardless of your level. That is not what I was talking about.

How strong of an effect is the result? Is the result comparable to a Tier 2 effect or a Tier 4 effect? When Rogues get the ability to reliably hit DC 20-30, is their reward a Tier 1 effect or a Tier 3 effect? Is the DM saying skills are high DC for weak effects, or high DC for strong effects?


To add context, the artificer was trying to make something from raw materials. She guidances herself, then gave herself a flash of genius whilst working. They rolled high and the end result was a 41. So, the device she made was inhumanly perfect. The DC for success was 20, very easy to do for her, but she more than doubled that. In the spirit of rewarding effort and character investment, how would you lot handle that? Improve the device? Increase it's efficiency? Leave it alone?

Given the DC for the device was a 20, I would peg a masterwork version as a 25. Anything above that is still just a masterwork version, but demonstrates the crafter could have made an even harder project. Alternatively have the 40 meant the crafter found a way to bend one of the natural laws during the construction. For example a battery powered watch that runs without a battery.

[Contrast]

Best to not think of the check results as absolutes. More reasonable to treat them as "the best possible attempt anyone could have made"
A roll of 43 on Arcana doesn't means that you understand magic better than Acererak possibly can, but rather you had a genius moment of understanding on a thing at the time.
A 39 on Intimidation doesn't means you're scarier than its possible for Zariel to be all the time, just circumstances of the moment meant even the bravest/arrogant of foes would still give pause and act with caution towards you.
A persuasion roll of 40 isn't that you're more silver tongued than any of the famous devils/demons/anything, you just have a very convincing argument that's hard to dispute.

I mean Zariel literally can't roll a 39 on Intimidation. If they tried to make the same threat in the same context, they'd be less intimidating.
I couldn't find anything with a high enough persuasion to roll a 40 so yes, if you're hitting 40 you have a flat modifier better than anything else, anywhere. You have made the argument better than its possible for literally anything or anyone else to make.

In practice, no PC will have a consistently better knowledge of magic than Acererak (Expertise at max level with a Tome still only comes out to +18 so you'll need to roll a 20 and have bardic inspiration/Guidance and a mid level artificer/anything else I'm not thinking of to get over the top which fits in with your 'flash of inspiration' justification). But for that moment, you have achieved a feat of understanding that would have eluded Acererak at their best.

The point of these is to give context for what hitting DC40 means - you have skills which stand amoung the greatest of immortals and even potentially the gods.

[Zhorn]

Still attached to a d20 roll. You can get the crazy high rolls and have a good modifier, but you're not up there as a constant. You could still roll low, and while the modifier still gives you a good overall result, it is still possible to miss a high dc or be beaten in a contested check against individuals who also can rate near said results, even if slighting below your max potential. Even masters of a discipline can slip up and be bested by others.

[Mork]

As others have said there are no rules for this instance.
That being said:

-the god of the forge sees this masterworks and blesses it, making it magical (so it will be found by new civilizations thousands of years in the future and it will still work).
-You finish the device in moments notice, and spend the rest of the time making it incredibly beautifull, increasing it's worth 3 fold?
-You make the tool so nice in it's use that you don't need to be trained in the tool to use it?


In my games I like to reward insane high rolls, if it is a skill challange, an 35+ will count as 3 successes. A 40+ stealth makes you stealth without cover. A 40+ arcana check gives you a moment of enlightenment, you are on the radar of a god of magic, as a possible succesor, and will probably learn some magical secret that is relevant to the campaign.

If this is going to happen more often, (if there are high level bardic inspiration for example), and you want to do something with a high roll, maybe make subsequent awesomeness in that roll more difficult. Say on a 40+ you want to do something extra, well the next time that skill is being used the DC for something extra is 42, then 44 etc.

[Contrast]

To add context, the artificer was trying to make something from raw materials. She guidances herself, then gave herself a flash of genius whilst working. They rolled high and the end result was a 41. So, the device she made was inhumanly perfect. The DC for success was 20, very easy to do for her, but she more than doubled that. In the spirit of rewarding effort and character investment, how would you lot handle that? Improve the device? Increase it's efficiency? Leave it alone?

Difficult to say without knowing what they were making exactly. Off the cuff I'd probably just have it be created in the minimum possible amount of time at the maximum possible quality.

So for example if it would normally take someone X time to make a quick poor quality version and Y time to make a masterwork version, they just created a masterwork version in X time.


Still attached to a d20 roll. You can get the crazy high rolls and have a good modifier, but you're not up there as a constant. You could still roll low, and while the modifier still gives you a good overall result, it is still possible to miss a high dc or be beaten in a contested check against individuals who also can rate near said results, even if slighting below your max potential. Even masters of a discipline can slip up and be bested by others.

And its possible for a level 1 char with 14 Int to beat Acererak on an Arcana contest if Acererak rolls a 1 - my initial comment was phrased with the assumption all those named rolled 20s. My point was that if you're hitting DC40 you're challenging these amazing beings at the things they do best, when they're doing them at their best, and you're still winning.

If you're even capable of rolling a 40 on a crafting test, you're on the radar of the gods (hell, they might come and ask you to craft something for them).

[clearstream]

To add context, the artificer was trying to make something from raw materials. She guidances herself, then gave herself a flash of genius whilst working. They rolled high and the end result was a 41. So, the device she made was inhumanly perfect. The DC for success was 20, very easy to do for her, but she more than doubled that. In the spirit of rewarding effort and character investment, how would you lot handle that? Improve the device? Increase it's efficiency? Leave it alone?
I'd be cautious about over-rewarding such rolls as there are an increasing number of ways for creatures with access to abilities like those in character classes to achieve them.

You mention guidance and flash of genius, and you or others noted bardic inspiration. There are also psionic dice, so one might imagine a 3rd level rogue, 7th artificer throwing those into the mix. A circle of stars druid can apply weal if they like. Access to expertise is becoming increasingly common, e.g. in the Skill Expert feat. Say our 10th level rogue/artificer with 18 Int, goes all in. d4+4+d10+d6+d6+4+4+d20 = 17 to 58 averaging around 38.

So far as I can tell from use in play, the core rule designers mildly low-balled ability DCs, and then over time there has been ability check inflation on top of that. 40 is the new 30.

[JackPhoenix]

So far as I can tell from use in play, the core rule designers mildly low-balled ability DCs, and then over time there has been ability check inflation on top of that. 40 is the new 30.

No, they did not. The ability DC's are made with an average PC in mind, not hyperspecialized, purpose-built character set up to reach arbitrary high ability modifiers. That's what bounded accuracy, the core design principle of 5e, is about. You don't need to have crazy high modifiers to succeed on things, and shouldn't need to.

[Unoriginal]

It is possible to roll a lot more than 30. It's not that DC 40 is impossible mathematicaly, it's just that DC 30 represents the limit of the possible, as an arbitrary limit.

Above 30, the difficulty is "don't bother", not because it's factually impossible to roll it, it is simply a choice they made to avoid escallation.

Though once again, it is true it doesn't apply to opposite checks.

[noob]

It is possible to roll a lot more than 30. It's not that DC 40 is impossible mathematicaly, it's just that DC 30 represents the limit of the possible, as an arbitrary limit.

Above 30, the difficulty is "don't bother", not because it's factually impossible to roll it, it is simply a choice they made to avoid escallation.

Though once again, it is true it doesn't apply to opposite checks.

Then some modules casually add dcs higher than 30.
Bounded accuracy except when it is not.

[clearstream]

No, they did not. The ability DC's are made with an average PC in mind, not hyperspecialized, purpose-built character set up to reach arbitrary high ability modifiers. That's what bounded accuracy, the core design principle of 5e, is about. You don't need to have crazy high modifiers to succeed on things, and shouldn't need to.
I believe they did, notwithstanding the DMG 238 wordings.


DC 5 can be accomplished by "most people" which seems to me to include ordinary inhabitants of the game world, as well as player characters
DC 30 should be something that a "20th-level character with proficiency and a relevant score of 20 still needs a 19 or 20 on the die roll to succeed"
In tier 2, with basic adds like guidance and inspiration, I see values like d4+d8+4+3+d20 or 10 to 39. For experts that becomes 13 to 42.
By tier 3, that becomes d4+d10+5+4+d20 or 12 to 43. For experts 16 to 47.

When I think about the DCs that I observe characters frequently surpassing in groups I play with and hear about anecdotally on forums like this one, I do not find the DC 30 expectation plausible. It looks low-balled, but I could caveat that in a way you might agree with: it could be true of an unbuffed 20th-level non-expert. That isn't how I see the game being played, however. So when I say low-balled, I really mean relative to what I see and hear about players doing at the table. The real experience of play. Not the hypothetical cases captured in the DMG. I feel disappointed when game designers write guidelines that ignore how their game is really played. I would concede the point that "nearly impossible" means - relative to a character who isn't using the tools at their disposal. The way I play it when I DM is that DC 30 is essentially heroic: it is very possible for heroic creatures - like player characters - and nearly impossible for those who aren't.

I guess what I am drawing attention to is that a DM must reconcile "frequently possible" for the people at their table - the ones most involved in the action - with "nearly impossible" for the world at large. An artificer is already doing something nearly impossible - for the world at large - when she makes her gadget. Hence I advise not to over-reward such rolls. If a DM were to want miraculous skill outcomes that feel miraculous, I believe they'd need to set the threshold for those much higher. But they might enjoy a game where miraculous outcomes are actually fairly common outcomes, which is also fine.

[RedMage125]

OP, you've gotten that the official answer is that nothing special happens.

But I agree with the posters who suggest that, if you want to reward such a high check, a good result would be to have the item be constructed of particularly high quality and/or in a remarkably short amount of time.

[Unoriginal]

I'm not aware of any module written after 5e's launch or written by WotC that has a DC above 30.

Rise of Tiamat/Tyranny of Dragons was written by a different team before 5e was finished, and we all know it countains obvious beta testing artifacts like "you just need to know the activation code to use a Teleportation Circle".

[Naanomi]

I'm not aware of any module written after 5e's launch or written by WotC that has a DC above 30.
Some contested checks that can get there (a few pretty easily) depending on the roll

[Unoriginal]

Some contested checks that can get there (a few pretty easily) depending on the roll

As I've acknowledged twice already.

[Naanomi]

As I've acknowledged twice already.
just clarifying we see it in modules, not just theoretically

[Tanarii]

So far as I can tell from use in play, the core rule designers mildly low-balled ability DCs, and then over time there has been ability check inflation on top of that. 40 is the new 30.
They did not. They high-balled them by about 5 points, with the assumption a Pc would only ever make checks for things in their primary or secondary ability score, and would have proficiency.

Unless you're a "One check to Rule Them All" DM, that's not a reasonable assumption.

[clearstream]

They did not. They high-balled them by about 5 points, with the assumption a Pc would only ever make checks for things in their primary or secondary ability score, and would have proficiency.
Are you saying that, that is what you observe and hear about in play? Or do you mean that the designers said something about that, specifically?

I guess we need to define what "high" and "low" would entail in this context. What do you mean when you say "by about 5 points"? And is it right to understand that you define that in the context of a character without buffs or expertise?

[Tanarii]

Are you saying that, that is what you observe and hear about in play? Or do you mean that the designers said something about that, specifically?Im contrasting "Medium" being a DC15 vs someone trying to make a non-proficient ability check for a tertiary or more score.

[clearstream]

Im contrasting "Medium" being a DC15 vs someone trying to make a non-proficient ability check for a tertiary or more score.
I might misunderstand. Earlier you said...

They high-balled them by about 5 points, with the assumption a Pc would only ever make checks for things in their primary or secondary ability score, and would have proficiency.
But now I think you are saying it is high balled against someone who is making a non-proficient check for a tertiary ability score (would that be a 10?) Is it also high balled for someone making a check against their primary ability score with proficiency?

[Tanarii]

I might misunderstand. Earlier you said...

But now I think you are saying it is high balled against someone who is making a non-proficient check for a tertiary ability score (would that be a 10?) Is it also high balled for someone making a check against their primary ability score with proficiency?
It's high balled for anyone making a non-proficient non-primary check, which should be the most common kind of check being made. Because they seem to assume a default that PCs will only be making checks that they are proficient with and are primary or secondary. Which is an utterly ridiculous assumption.

[clearstream]

It's high balled for anyone making a non-proficient non-primary check, which should be the most common kind of check being made. Because they seem to assume a default that PCs will only be making checks that they are proficient with and are primary or secondary. Which is an utterly ridiculous assumption.
I think I might see now our difference in perspectives. For me, wording in the DMG and PHB makes me believe that DCs are designed with characters in mind. And then what I observe in every group I've played with is that the character with the highest proficiency will almost always make the check. So for me it is completely plausible - the norm - that PCs will only be making checks that they are proficient with and are primary or secondary. I call the DCs low-balled because what the designers appear to have neglected are expertise and buffs, which I see used very commonly in play. Characters will help one another, too, so they'll roll toward the high end of the curve more often than not.

From my perspective, the "ridiculous assumption" would be setting DCs based on non-proficient, non-primary checks. There is obviously a perspective from which that can make sense, but I just can't buy it for a system designed with PCs in mind.

[Tanarii]

A system designed where PCs can only attempt maybe two to four things with a 50% or better chance of success doesn't seem to be working as intended to me.

Edit: looking at it another way, having a baseline average 25-35% chance of success for the majority of tasks that require a check, even at very high level, would feel a bit too much like being a low level BECMI or AD&D thief all the time. :smallyuk:

[Dark.Revenant]

A party consisting of classes like Fighter, Barbarian, Monk, Sorcerer, Warlock, etc. don't have guaranteed access to beyond-baseline skill progression. For such a party, 31 might very well be the highest any of them ever roll on an ability check for the entire campaign.

A DC of 30 means "good luck accomplishing this without a specialist." DCs above 30 are simply impossible, straight-out, for a non-specialist, and become increasingly difficult even for someone who IS a specialist. Use them sparingly. I have only seen them used in third-party campaigns, and even then only when the campaign was converted from the original 3.5e or Pathfinder.

[Pex]

It's high balled for anyone making a non-proficient non-primary check, which should be the most common kind of check being made. Because they seem to assume a default that PCs will only be making checks that they are proficient with and are primary or secondary. Which is an utterly ridiculous assumption.

If only there was a standard guideline for things as to what makes something easy or hard instead of only just having numbers and make it up by fiat.

[Tanarii]

If only there was a standard guideline for things as to what makes something easy or hard instead of only just having numbers and make it up by fiat.
There is a standard guideline.

I just think it's too high by about 3-5 pts. :smallamused:

[Contrast]

There is a standard guideline.

I just think it's too high by about 3-5 pts. :smallamused:

I've had this conversation with multiple DMs of mine but it never quite seems to stick.

More than once I've had DMs throw something like a DC10 Acrobatics test at the entire party and then seem surprised when half the party fails this 'easy' test when that is in fact going to be within the expected range of results even for a lot of level 20 parties.

I have one DM in particular who bemoans it being impossible to be truly 'good' at a skill in 5E but also seems unable to use a DC of less than 10 on any test for anything, resulting on one memorable occasion my warforged artificer being unable to recall the symbol of House Cannith.

[Dark.Revenant]

Yeah, as a general rule, those DCs are for when you're performing under great stress/duress. If the pressure isn't that high, slide everything back by 5 (or just don't even call for a check), and it starts to make more sense.

[clearstream]

A system designed where PCs can only attempt maybe two to four things with a 50% or better chance of success doesn't seem to be working as intended to me.

Edit: looking at it another way, having a baseline average 25-35% chance of success for the majority of tasks that require a check, even at very high level, would feel a bit too much like being a low level BECMI or AD&D thief all the time. :smallyuk:
Aren't we only rolling when there is a risk and cost for failure?

Say with animal handling. There's no roll for taking an amicable dog for a walk! Only for dealing with a fierce cur in a foul mood, that has a vicious bite. Or with acrobatics. There's no roll for walking along the wide top of a low wall. Only for dashing along the narrow top of an icy wall upon a daunting precipice.

[stoutstien]

Aren't we only rolling when there is a risk and cost for failure?

Say with animal handling. There's no roll for taking an amicable dog for a walk! Only for dealing with a fierce cur in a foul mood, that has a vicious bite. Or with acrobatics. There's no roll for walking along the wide top of a low wall. Only for dashing along the narrow top of an icy wall upon a daunting precipice.

In principle this actually makes it worse. If a player is only making a handful of ability checks in a session and they have a good chance to fail 3/5 of them even if the checks in question are easy it doesn't really leave anyone in a good spot. If you convert the DC table into % chance of a certain modifier value to succeed it's pretty bad.

[Contrast]

Aren't we only rolling when there is a risk and cost for failure?

Say with animal handling. There's no roll for taking an amicable dog for a walk! Only for dealing with a fierce cur in a foul mood, that has a vicious bite. Or with acrobatics. There's no roll for walking along the wide top of a low wall. Only for dashing along the narrow top of an icy wall upon a daunting precipice.

And part of the issue (for me at least) is that most DMs jump straight from 'don't bother rolling' to DC10, which for a lot of PCs on any random skill check is going to be something like a 40-50% chance of failing which doesn't leave a lot of room for nuance. Theres a lot of tasks that should fall in the category inbetween 'always succeed' and 'succeed 50% of the time' for people with an average human stat and without proficiency.

I don't want to encourage DMs to start asking for rolls on things they don't think need them but I do want to encourage DMs to remember that the scale doesn't start at 10.

[clearstream]

And part of the issue (for me at least) is that most DMs jump straight from 'don't bother rolling' to DC10, which for a lot of PCs on any random skill check is going to be something like a 40-50% chance of failing which doesn't leave a lot of room for nuance. Theres a lot of tasks that should fall in the category inbetween 'always succeed' and 'succeed 50% of the time' for people with an average human stat and without proficiency.

I don't want to encourage DMs to start asking for rolls on things they don't think need them but I do want to encourage DMs to remember that the scale doesn't start at 10.
I agree with that. I've found myself leaning more toward lower DCs as I have become more familiar with 5e. For example, I now go with 8 when before I'd have said 10, and 12 when before I'd have said 15. I feel likely to even go a few points lower for many checks, and more often waive them altogether.

On the other hand - in line with my "low-balled" argument - I do still feel the "nearly impossible" DC is written from the wrong perspective. This thread has made me reflect on it, yet where I land is something like - I think nearly impossible is relative to PCs and so means "nearly impossible for PCs". It should be experientially nearly impossible. On the other hand, my demographics are calibrated at around 1/250 people having tier 1 character-classes or equivalent, and then 1/5 pf that per tier upward. So the tier 3 character for whom 30 is in fact very possible is... 1/6250. And then perhaps only half of such characters are experts and/or have access to buffs. Is 1/12,500 nearly impossible enough? I guess I'd put "nearly impossible" somewhere above 1/100,000 so I'm going with "No". But that's just for my campaign world with my demographic assumptions.

[Thunderous Mojo]

It means they've succeeded on whatever that check was. That's it. Whatever they were trying to do wasn't impossible, otherwise, there wouldn't be a roll in the first place, as attempt to do something impossible automatically fails, no matter how high can you get your modifier.
We are falling into the territory of Zeno's Paradoxes, and paradoxes in general. The word Ineffable means something that is too great or extreme to be expressed in words, and yet we have a word for it.

Indeed we talk about 'ineffable' things all the time. At the extremes, what is "Possible" and "Impossible" becomes blurry.