What do you do? Give examples using saving throws. Forget saving throws and think about skills:
OK.
In AD&D 1E, the GM says, "Roll to Climb a Hard wall please." The Player asks, "What does hard mean?" and the GM answers, "20". ¹ The player looks at their sheet and sees "Climb 85%". They roll a d%, add 20, and have to get 85 or under. A success requires rolling 65 or under, a 65% chance.
(Unless of course the character is not a thief, in which case they stand at the bottom of the wall looking wisfully upwards.)
In D&D 5E, the GM says, "Roll a Hard save vs Constitution climb check please." The Player asks, "What does hard mean?" and the GM answers, "20". The player looks at their sheet and sees "CON Save STR (Athletics) +6". They roll a d20, add 6, and have to get 20 or over. A success requires rolling 14 or over, a 35% chance.
_____
¹ In practice... the GM answers, "I don't know what hard means, the book doesn't say, umm, let's say a hard wall is +20. Is that too low? Should it be +30?"
And then the player asks, "Do I have to roll under or over?" and everyone has to think about which parts of the game use roll-under and which use roll-over. After all, this is the game where +1 armour subtracts 1 from your armour class.
Functionally these might seem to be the same thing. They are different. In how they represent and resolve skills.
What's the big difference? With a little experience, a DM will have a fair idea of what the odds of 10, 15, 20, etc. mean relative to a given bonus, and can apply them accordingly. There is no way to preemptively assign ratings to the majority of ideas players will come up with, so consequently there's not a good system for preemptive transparency. Plus revealing the DC on certain checks could provide meta knowledge to players, particularly for stuff like social or stealth rolls. Any tabletop system is predicated on the players trusting the GM to treat them fairly, and will have far too many moving and subjective parts to realistically create hard "this DC for this particular circumstance" tables as opposed to "for something about this difficult we recommend DC 15". And it's not that awful hard to work out the general application of 5e DCs. 10 is a token challenge for anyone proficient and usually better than 50/50 otherwise, 15 is a token to experts, challenge to profs, and just possible for everyone else, 20 challenge for experts, just possible for profs, and about 1-in-20 for the rest, if at all.
I've explained more times than I can count how they are different. Read the thread.
I did, and your entire point seems to boil down to “but what if the DM makes the roll too hard?”, which is not actually a concern grounded in what type of dice one rolls and is in point of fact one that exists independently of the dice system. You then follow up by asserting a more elaborate process that still requires the DM to adjudicate an ease/difficulty modifier is superior, even though they retain just as much license to dictate the difficulty to suit their purpose, if they’re so inclined.
So, if I can set up the order of preference, from least preferred to most preferred, for OP it's like this:
- d20 + modifiers, to hit a DC that you learn after you roll (the "DM is in full control" method)
- d20 + modifiers, to hit a DC that you learn immediately before you roll (my usual method)
- d20 + modifiers, to hit a DC that you learn well before you decide whether to roll (the 3.5e method)
- d20 + modifiers that you learn after you roll, to hit a DC written on your character sheet (a method I would be surprised to hear is used by anyone)
- d20 + modifiers that you learn immediately before you roll, to hit a DC written on your character sheet (what I understand to be the old-school method. Also the Numenara method, though they use 2d6 I think)
- d20 + modifiers that you learn well before you decide whether to roll, to hit a DC written on your character sheet (I'll call it the White Wolf method, though I'm not sure that's exactly accurate)
- d20 + modifiers that you learn immediately before you roll, to hit a DC written on your character sheet (what I understand to be the old-school method. Also the Numenara method, though they use 2d6 I think)
This is essentially the method used in Call of Cthulhu, where your skill percentage gets broken down into smaller target numbers depending on difficulty (Hard = half your skill, Extreme = 1/5th of your skill), although I think the GM has some latitude in telling you what kind of success you needed before or after you roll
Rollback Post to RevisionRollBack
Active characters:
Carric Aquissar, elven wannabe artist in his deconstructionist period (Archfey warlock) Lan Kidogo, mapach archaeologist and treasure hunter (Knowledge cleric) Mardan Ferres, elven private investigator obsessed with that one unsolved murder (Assassin rogue) Xhekhetiel, halfling survivor of a Betrayer Gods cult (Runechild sorcerer/fighter)
What do you do? Give examples using saving throws. Forget saving throws and think about skills:
OK.
In AD&D 1E, the GM says, "Roll to Climb a Hard wall please." The Player asks, "What does hard mean?" and the GM answers, "20". ¹ The player looks at their sheet and sees "Climb 85%". They roll a d%, add 20, and have to get 85 or under. A success requires rolling 65 or under, a 65% chance.
(Unless of course the character is not a thief, in which case they stand at the bottom of the wall looking wisfully upwards.)
In D&D 5E, the GM says, "Roll a Hard save vs Constitution climb check please." The Player asks, "What does hard mean?" and the GM answers, "20". The player looks at their sheet and sees "CON Save STR (Athletics) +6". They roll a d20, add 6, and have to get 20 or over. A success requires rolling 14 or over, a 35% chance.
_____
¹ In practice... the GM answers, "I don't know what hard means, the book doesn't say, umm, let's say a hard wall is +20. Is that too low? Should it be +30?"
And then the player asks, "Do I have to roll under or over?" and everyone has to think about which parts of the game use roll-under and which use roll-over. After all, this is the game where +1 armour subtracts 1 from your armour class.
Functionally these might seem to be the same thing. They are different. In how they represent and resolve skills.
What's the big difference? With a little experience, a DM will have a fair idea of what the odds of 10, 15, 20, etc. mean relative to a given bonus, and can apply them accordingly. There is no way to preemptively assign ratings to the majority of ideas players will come up with, so consequently there's not a good system for preemptive transparency. Plus revealing the DC on certain checks could provide meta knowledge to players, particularly for stuff like social or stealth rolls. Any tabletop system is predicated on the players trusting the GM to treat them fairly, and will have far too many moving and subjective parts to realistically create hard "this DC for this particular circumstance" tables as opposed to "for something about this difficult we recommend DC 15". And it's not that awful hard to work out the general application of 5e DCs. 10 is a token challenge for anyone proficient and usually better than 50/50 otherwise, 15 is a token to experts, challenge to profs, and just possible for everyone else, 20 challenge for experts, just possible for profs, and about 1-in-20 for the rest, if at all.
I've explained more times than I can count how they are different. Read the thread.
I did, and your entire point seems to boil down to “but what if the DM makes the roll too hard?”, which is not actually a concern grounded in what type of dice one rolls and is in point of fact one that exists independently of the dice system. You then follow up by asserting a more elaborate process that still requires the DM to adjudicate an ease/difficulty modifier is superior, even though they retain just as much license to dictate the difficulty to suit their purpose, if they’re so inclined.
Did you? Because beyond that point and the thing I've mentioned more than that is how the different systems represent and resolve skills differently.
Go back and read the example given of two systems, one using fixed percentages for skills and the other using increments of percentiles for them instead.
Whether they function mathematically the same even in circumstances where any bonus or penalty is to be applied to the roll would then have to be met with the equivalent target number that would have to be chosen, saying this thief has a # percent chance of doing this thing and saying this thief adds or subtracts this to or from a roll are different.
This is also why I've said much of this is a matter of preference.
I do believe a system that allows a DM to tell a player to add 1 or 2 to or subtract 1 or 2 from a d20 roll against a fixed target number determined by class and level is less likely to be a system prone to impulse than one in which a DM can just tell the player they've failed because what they rolled didn't meet the target number. A target number that may only ever exist in the DM's head if it ever did.
But that comes down, as many said, to a bad DM. Not necessarily the system. I can agree with that.
So, if I can set up the order of preference, from least preferred to most preferred, for OP it's like this:
- d20 + modifiers, to hit a DC that you learn after you roll (the "DM is in full control" method)
- d20 + modifiers, to hit a DC that you learn immediately before you roll (my usual method)
- d20 + modifiers, to hit a DC that you learn well before you decide whether to roll (the 3.5e method)
- d20 + modifiers that you learn after you roll, to hit a DC written on your character sheet (a method I would be surprised to hear is used by anyone)
- d20 + modifiers that you learn immediately before you roll, to hit a DC written on your character sheet (what I understand to be the old-school method. Also the Numenara method, though they use 2d6 I think)
- d20 + modifiers that you learn well before you decide whether to roll, to hit a DC written on your character sheet (I'll call it the White Wolf method, though I'm not sure that's exactly accurate)
Is that right?
The one I've bolded is my preferred method.
The first two methods you've provided don't really account for those DMs who let the DC be known neither before nor after the roll. But never. Unless what you mean when you say "after" is just whether or not you have succeeded or failed. Much like not having the DM tell players an opponent's AC. Which of course is easy enough to figure out if you've hit and missed enough.
That's very common. You'll find blogs and threads and subreddits pages in length with those who use that method saying why it is even "the right way" to approach DCs. Obviously, there is none, and how you say you run it seems perfectly fine.
My objections—at least those that are about a DM putting impulses before rules—probably only concern those who use that method. It is, as many said, a problem that lies with the DM and not really with the system. That is just a matter of prefence.
So, if I can set up the order of preference, from least preferred to most preferred, for OP it's like this:
- d20 + modifiers, to hit a DC that you learn after you roll (the "DM is in full control" method)
- d20 + modifiers, to hit a DC that you learn immediately before you roll (my usual method)
- d20 + modifiers, to hit a DC that you learn well before you decide whether to roll (the 3.5e method)
- d20 + modifiers that you learn after you roll, to hit a DC written on your character sheet (a method I would be surprised to hear is used by anyone)
- d20 + modifiers that you learn immediately before you roll, to hit a DC written on your character sheet (what I understand to be the old-school method. Also the Numenara method, though they use 2d6 I think)
- d20 + modifiers that you learn well before you decide whether to roll, to hit a DC written on your character sheet (I'll call it the White Wolf method, though I'm not sure that's exactly accurate)
Is that right?
The one I've bolded is my preferred method.
Personally speaking, I hate that method and would probably leave a table.
I want to know the chance of success before I choose a path of action. I want to know if the wall is easy or hard to climb before I make the decision to have my character climb. I don't want the GM to suddenly suprise me with, "This wall is really difficult, roll at -50, if you fail you fall and die."
I mean, realistically speaking most events you're going to be rolling a d20 for are too spontaneous and varied for you to have an exact DC level understanding of how difficult they will be. Only physical stuff in the area of jumping, lifting, running, climbing, etc. is something you could plausibly have a consistent baseline for how well you could perform. Keep in mind that to a certain degree the d20 roll itself is meant to represent the various unknowable factors in the attempt, or errors in judgement on your character's part- not in the sense of intruding on player agency, but simply in the human error factor. The hard knowable factor is your modifier to the roll, the DC is approximately how difficult it is on a scale of 1-20 (21+ being ones that only a truly exceptional individual could even hope to accomplish, per the humanoid baseline), and the d20 is the variables of circumstance, judgement, and the spectrum of possibilities within the DC.
I mean, realistically speaking most events you're going to be rolling a d20 for are too spontaneous and varied for you to have an exact DC level understanding of how difficult they will be.
Eh, in most cases your uncertainty about whether you'll succeed is adequately covered by the d20 roll -- you can easily describe rolling a 1 as "it was harder than I expected".
The game gives DMs a set of "difficulties" and assigned DCs for each. As long as they present the difficulty to you, how is that any different than your table? If you know a "difficult roll" is a DC 20, and you know your modifier is +7, then you need a 13 or higher, or a 40% chance of success.
Sure, a DM could use an in-between number, or not tell you the difficulty, but it seems like the other method eliminates a difficulty scale from the situation. If your Character always has a 20% chance of failing a climb check, no matter if the cliff is sloped, vertical, dry, coated in ice, dripping in oil, etc. I prefer this power reside with the DM, not the player (or a fixed table).
There is a legitimate gripe about the swingy-ness of a d20 system for checks, but I don't see why this is an issue for 5e
So, if I can set up the order of preference, from least preferred to most preferred, for OP it's like this:
- d20 + modifiers, to hit a DC that you learn after you roll (the "DM is in full control" method)
- d20 + modifiers, to hit a DC that you learn immediately before you roll (my usual method)
- d20 + modifiers, to hit a DC that you learn well before you decide whether to roll (the 3.5e method)
- d20 + modifiers that you learn after you roll, to hit a DC written on your character sheet (a method I would be surprised to hear is used by anyone)
- d20 + modifiers that you learn immediately before you roll, to hit a DC written on your character sheet (what I understand to be the old-school method. Also the Numenara method, though they use 2d6 I think)
- d20 + modifiers that you learn well before you decide whether to roll, to hit a DC written on your character sheet (I'll call it the White Wolf method, though I'm not sure that's exactly accurate)
Is that right?
The one I've bolded is my preferred method.
Personally speaking, I hate that method and would probably leave a table.
I want to know the chance of success before I choose a path of action. I want to know if the wall is easy or hard to climb before I make the decision to have my character climb. I don't want the GM to suddenly suprise me with, "This wall is really difficult, roll at -50, if you fail you fall and die."
By "immediately before you roll" it doesn't mean you've started to climb that wall and then and only then does the DM get to tell you you're doomed. It means you will know what your chances are just before you start to climb the wall. Should you then choose to. At least that's how I run what is understood to be "the old-school method" with the "DC" recorded on the character sheet.
I think the author of that post simply used the wording "immediately" to make clear the distinguish between that and rolling with any variable level of difficulty being predetermined "well before" as those conditions have already been accounted for in the rules.
But I could be wrong.
No, sorry, what I meant (and I almost clarified this in the post, but I couldn't do it concisely) is that you get a description of the task, decide whether to attempt it, and then the DM tells you the DC, and then you roll. Because this is an analog game (the mechanics are executed by the brains of the participant players), there is a period of potential negotiation there. "Wait, 30? I thought you said it was a regular rock wall, wouldn't a 30 be like, the Hell Climb Pillar from Jojo's?" "Yeah, okay, it's probably more like a 20." And there's also sometimes a period of quarreling over "I wouldn't try to do it if it seemed that impossible," that sometimes results in an agreement to let the player back out of the attempt, or take some action to improve their chances before they make the attempt. But essentially, the DC is only communicated in fictional terms. These intermediate steps are just trying to rectify a failure to achieve clarity in that way.
The point of saying the DC before the roll, in this case, is simply to remove doubt about the impartiality of the outcome. The DM said the target number, and you didn't roll it. Plain and simple. The methods where you know the DC well in advance are referring to the ones where you find out what the DC would be if you tried it, and then you get to decide whether you're going to try it.
No, sorry, what I meant (and I almost clarified this in the post, but I couldn't do it concisely) is that you get a description of the task, decide whether to attempt it, and then the DM tells you the DC, and then you roll. Because this is an analog game (the mechanics are executed by the brains of the participant players), there is a period of potential negotiation there. "Wait, 30? I thought you said it was a regular rock wall, wouldn't a 30 be like, the Hell Climb Pillar from Jojo's?" "Yeah, okay, it's probably more like a 20." And there's also sometimes a period of quarreling over "I wouldn't try to do it if it seemed that impossible," that sometimes results in an agreement to let the player back out of the attempt, or take some action to improve their chances before they make the attempt. But essentially, the DC is only communicated in fictional terms. These intermediate steps are just trying to rectify a failure to achieve clarity in that way.
The point of saying the DC before the roll, in this case, is simply to remove doubt about the impartiality of the outcome. The DM said the target number, and you didn't roll it. Plain and simple. The methods where you know the DC well in advance are referring to the ones where you find out what the DC would be if you tried it, and then you get to decide whether you're going to try it.
I mean, communicating some sense of the scale of the task can be reasonable for the raw physical, but less so for most other instances. And regardless it can provide a certain level of meta knowledge. Really, if you can't trust your DM to assign reasonable DC's without oversight, you should probably just leave the table. The role of DM simply doesn't work if the players can't trust them to run things independently.
Upon reflection, what I usually do is actually the one I labeled the 3.5e method, I just get there in a more abstract way than 3.5e does. I'll say, "the wall is really rough broken stone with iron rebar sticking out, climbing it would be pretty easy, let's say a DC 10." I actually don't have a problem with just putting the mechanics right in the narration like that. There are times when I leave it to the players to deduce, but that's just because it's sometimes fun to use your brain in that way. Like I usually won't give the AC of armored foes, I'll just say what kind of armor they're wearing. The encyclopedias at the table get to benefit from the fact that their brains, like mine, are cursed to carry such information to their grave. But I don't think it breaks immersion to just say the number.
Where I differ from 3.5e is that I don't prefer to have a big list of modifiers to consider all the time. I really appreciate the "5, 10, 15, 20" baseline. We know cover is 2 or 5, which is weird, but not overly involved if it's just cover. Adv/dis if the modifier has something to do with you or your approach, otherwise it just has a DC based on if a regular person tried to do it. Cool, easy.
First off, of course, there is no "wrong" way to play D&D.
The idea that the d20 create 5% increments as good is something not everyone agrees with. I see the intellectual appeal of the tidiness and simplicity, which is cool if you like it or don't care. But for me, and maybe others, it mutes part of the potential flavor--perceived and real--of the game.
The answer to the 2d10 middle bulge is a d100 or E-dice (which aren't as much fun), or make percentage based skill checks reflect that middle bulge: start at 50 as zero, divide the skill score and add one half in either direction (so a 66% chance would equal any number 17-83).
The player is responsible for knowing their skills and dice rolls needed to perform them.
The DM is responsible for making sure dice roles are doing their job--that is making the game 'realistically' challenging and not too easy or too hard nor overly fair or overly unfair.
FACTUALLY, the d20 system entered Dungeons and Dragons with 3.0e in 2000, RIGHT? That was 23 years ago (not 30 years ago), D&D was invented in 1974, so, it will be 4 more years before d20 will have become the longest serving central mechanic of the game. Either way, an argument for it based on its 'age' is a false appeal fallacy. The abacus is 4-5,000 years old, but that's not a reason to keep using them over a microprocessor.
This was a partially muddled discussion because it was started about multiple issues, but it didn't need to be quite so adversarial as it seemed to get at points. Anyone so into the game that they are here is just trying to enhance their experience and we should all be on board for that. If this isn't a topic that interests you, don't read it, or at least don't reply.
Responding to some who disagrees with you about something with the equivalent of "well leave then," is bigotry and weak. Be stronger and better than that.
Here is an articulation of the problem of d20 based systems:
Percentage-based rolls offer 1in100 possible degrees of chance, which feels way more random than the 5 in 100 or 1-20 of the d20.
This allows the game and DM and players to make the mathematical 'contours' of probability something other than 5% increments. Pickpocket/bend bars/ at 67% vs 69% vs 66% chance of success vs d20>12- or >13, aka 65% or 70% chance of success.
It feels more varied, more flavored.
The rounding of the d20 (no pun intended), creates a feeling of less randomness, less probability variation/flavor.
Heck, maybe it is really just the numbers: <65% looks/feels more 'randomy' than <13. It feels like you have more ways/opportunities to succeed because you do, because you have equally more ways/opportunities to fail.
Also, percentages are scientificy.
Again, of course, play the way you that you find fun and don't let anyone tell you that way is wrong unless nobody wants to play with you. Then maybe think about what is wrong with your way of playing.
EDIT: Not to zombify a thread, but I feel like the issue wasn't completely addressed, I just created an account here, and decided to offer this response.
I did, and your entire point seems to boil down to “but what if the DM makes the roll too hard?”, which is not actually a concern grounded in what type of dice one rolls and is in point of fact one that exists independently of the dice system. You then follow up by asserting a more elaborate process that still requires the DM to adjudicate an ease/difficulty modifier is superior, even though they retain just as much license to dictate the difficulty to suit their purpose, if they’re so inclined.
So, if I can set up the order of preference, from least preferred to most preferred, for OP it's like this:
- d20 + modifiers, to hit a DC that you learn after you roll (the "DM is in full control" method)
- d20 + modifiers, to hit a DC that you learn immediately before you roll (my usual method)
- d20 + modifiers, to hit a DC that you learn well before you decide whether to roll (the 3.5e method)
- d20 + modifiers that you learn after you roll, to hit a DC written on your character sheet (a method I would be surprised to hear is used by anyone)
- d20 + modifiers that you learn immediately before you roll, to hit a DC written on your character sheet (what I understand to be the old-school method. Also the Numenara method, though they use 2d6 I think)
- d20 + modifiers that you learn well before you decide whether to roll, to hit a DC written on your character sheet (I'll call it the White Wolf method, though I'm not sure that's exactly accurate)
Is that right?
This is essentially the method used in Call of Cthulhu, where your skill percentage gets broken down into smaller target numbers depending on difficulty (Hard = half your skill, Extreme = 1/5th of your skill), although I think the GM has some latitude in telling you what kind of success you needed before or after you roll
Active characters:
Carric Aquissar, elven wannabe artist in his deconstructionist period (Archfey warlock)
Lan Kidogo, mapach archaeologist and treasure hunter (Knowledge cleric)
Mardan Ferres, elven private investigator obsessed with that one unsolved murder (Assassin rogue)
Xhekhetiel, halfling survivor of a Betrayer Gods cult (Runechild sorcerer/fighter)
Did you? Because beyond that point and the thing I've mentioned more than that is how the different systems represent and resolve skills differently.
Go back and read the example given of two systems, one using fixed percentages for skills and the other using increments of percentiles for them instead.
Whether they function mathematically the same even in circumstances where any bonus or penalty is to be applied to the roll would then have to be met with the equivalent target number that would have to be chosen, saying this thief has a # percent chance of doing this thing and saying this thief adds or subtracts this to or from a roll are different.
This is also why I've said much of this is a matter of preference.
I do believe a system that allows a DM to tell a player to add 1 or 2 to or subtract 1 or 2 from a d20 roll against a fixed target number determined by class and level is less likely to be a system prone to impulse than one in which a DM can just tell the player they've failed because what they rolled didn't meet the target number. A target number that may only ever exist in the DM's head if it ever did.
But that comes down, as many said, to a bad DM. Not necessarily the system. I can agree with that.
The one I've bolded is my preferred method.
The first two methods you've provided don't really account for those DMs who let the DC be known neither before nor after the roll. But never. Unless what you mean when you say "after" is just whether or not you have succeeded or failed. Much like not having the DM tell players an opponent's AC. Which of course is easy enough to figure out if you've hit and missed enough.
That's very common. You'll find blogs and threads and subreddits pages in length with those who use that method saying why it is even "the right way" to approach DCs. Obviously, there is none, and how you say you run it seems perfectly fine.
My objections—at least those that are about a DM putting impulses before rules—probably only concern those who use that method. It is, as many said, a problem that lies with the DM and not really with the system. That is just a matter of prefence.
Personally speaking, I hate that method and would probably leave a table.
I want to know the chance of success before I choose a path of action. I want to know if the wall is easy or hard to climb before I make the decision to have my character climb. I don't want the GM to suddenly suprise me with, "This wall is really difficult, roll at -50, if you fail you fall and die."
I mean, realistically speaking most events you're going to be rolling a d20 for are too spontaneous and varied for you to have an exact DC level understanding of how difficult they will be. Only physical stuff in the area of jumping, lifting, running, climbing, etc. is something you could plausibly have a consistent baseline for how well you could perform. Keep in mind that to a certain degree the d20 roll itself is meant to represent the various unknowable factors in the attempt, or errors in judgement on your character's part- not in the sense of intruding on player agency, but simply in the human error factor. The hard knowable factor is your modifier to the roll, the DC is approximately how difficult it is on a scale of 1-20 (21+ being ones that only a truly exceptional individual could even hope to accomplish, per the humanoid baseline), and the d20 is the variables of circumstance, judgement, and the spectrum of possibilities within the DC.
Eh, in most cases your uncertainty about whether you'll succeed is adequately covered by the d20 roll -- you can easily describe rolling a 1 as "it was harder than I expected".
The game gives DMs a set of "difficulties" and assigned DCs for each. As long as they present the difficulty to you, how is that any different than your table? If you know a "difficult roll" is a DC 20, and you know your modifier is +7, then you need a 13 or higher, or a 40% chance of success.
Sure, a DM could use an in-between number, or not tell you the difficulty, but it seems like the other method eliminates a difficulty scale from the situation. If your Character always has a 20% chance of failing a climb check, no matter if the cliff is sloped, vertical, dry, coated in ice, dripping in oil, etc. I prefer this power reside with the DM, not the player (or a fixed table).
There is a legitimate gripe about the swingy-ness of a d20 system for checks, but I don't see why this is an issue for 5e
No, sorry, what I meant (and I almost clarified this in the post, but I couldn't do it concisely) is that you get a description of the task, decide whether to attempt it, and then the DM tells you the DC, and then you roll. Because this is an analog game (the mechanics are executed by the brains of the participant players), there is a period of potential negotiation there. "Wait, 30? I thought you said it was a regular rock wall, wouldn't a 30 be like, the Hell Climb Pillar from Jojo's?" "Yeah, okay, it's probably more like a 20." And there's also sometimes a period of quarreling over "I wouldn't try to do it if it seemed that impossible," that sometimes results in an agreement to let the player back out of the attempt, or take some action to improve their chances before they make the attempt. But essentially, the DC is only communicated in fictional terms. These intermediate steps are just trying to rectify a failure to achieve clarity in that way.
The point of saying the DC before the roll, in this case, is simply to remove doubt about the impartiality of the outcome. The DM said the target number, and you didn't roll it. Plain and simple. The methods where you know the DC well in advance are referring to the ones where you find out what the DC would be if you tried it, and then you get to decide whether you're going to try it.
I mean, communicating some sense of the scale of the task can be reasonable for the raw physical, but less so for most other instances. And regardless it can provide a certain level of meta knowledge. Really, if you can't trust your DM to assign reasonable DC's without oversight, you should probably just leave the table. The role of DM simply doesn't work if the players can't trust them to run things independently.
Upon reflection, what I usually do is actually the one I labeled the 3.5e method, I just get there in a more abstract way than 3.5e does. I'll say, "the wall is really rough broken stone with iron rebar sticking out, climbing it would be pretty easy, let's say a DC 10." I actually don't have a problem with just putting the mechanics right in the narration like that. There are times when I leave it to the players to deduce, but that's just because it's sometimes fun to use your brain in that way. Like I usually won't give the AC of armored foes, I'll just say what kind of armor they're wearing. The encyclopedias at the table get to benefit from the fact that their brains, like mine, are cursed to carry such information to their grave. But I don't think it breaks immersion to just say the number.
Where I differ from 3.5e is that I don't prefer to have a big list of modifiers to consider all the time. I really appreciate the "5, 10, 15, 20" baseline. We know cover is 2 or 5, which is weird, but not overly involved if it's just cover. Adv/dis if the modifier has something to do with you or your approach, otherwise it just has a DC based on if a regular person tried to do it. Cool, easy.
GENERAL RESPONSE to this thread:
First off, of course, there is no "wrong" way to play D&D.
Here is an articulation of the problem of d20 based systems:
Again, of course, play the way you that you find fun and don't let anyone tell you that way is wrong unless nobody wants to play with you.
Then maybe think about what is wrong with your way of playing.
EDIT: Not to zombify a thread, but I feel like the issue wasn't completely addressed, I just created an account here, and decided to offer this response.
✌️😎