What drives a poor reputation for the Rogue class?

[Just to Browse]

I think we may actually have an opportunity for the discussion to leave quantum build territory! At least with regards to Jellypooga's post here, their build was quite specific, so I intend to post some comparisons, assuming the thread does not combust before I get around to it

lol fair enough, I'm looking forward to it

I mean, true, but it's another way of saying "the DM will cater the challenge to the party and make up for the fact the rogue has less to offer and needs special considerations." Which is exactly the point made (20 pages ago? I lost count).

Might have showed up beforehand as well, but at the very least this is roughly my argument on p21 and Xervous's fantastic writeup on p27. Highlighting the rogue requires a more intentional approach to scenario design than internet theorycrafting would have one believe.

[JellyPooga]

I mean, true, but it's another way of saying "the DM will cater the challenge to the party and make up for the fact the rogue has less to offer and needs special considerations." Which is exactly the point made (20 pages ago? I lost count).



I think there's a few things here that could be useful (stuff that should've been part of the base game, if WotC had made a proper skill system).

1) Not everyone can roll for everything
If the bonuses are generally going to be low (bounded), then the dice roll becomes that much more important. To prevent the feel-bad chaos of a moron being more knowledgeable than the sage, or the weakling out-lifting the strongman, there needs to be guardrails on what can be rolled for. The DC might be 13 but that doesn't mean an entirely untrained person with an 8 in the associated trait has a 35% chance of succeeding - it means it's quite a hard task and if you don't have proficiency, you don't get to roll. Essentially, separate DC (a game construct) from the in-world difficulty of the task. I personally like DC 10 being the cap on untrained checks.

2) Shift DC's downwards
Most skill checks should be DC 10 or less, reflective of the fact that a trained person (proficiency) with reasonable talent (~+2 in the associated ability) will have a +5 check. They should succeeded at most tasks they are trained in most of the time. Many tasks? Don't even roll. They're DC 5. A character with a +5 or +6, they mathematically can't fail that.

3) Create unique uses of skills for characters with large modifiers
As a continuation of (1), a character with a +10 modifier should be able to attempt to do things a character with merely a +7 modifier cannot. This is both to have skills scale a little more, and also make more explicit how extraordinary someone with a +10 or +12 is. They aren't only 35% more likely to succeed; they can attempt things other people would consider impossible. I would probably gate these high level, extraordinary tasks behind expertise: with proficiency, you can attempt tasks up to DC 20. Expertise, and 21+ is open to you. Let's say swimming up a waterfall is DC 24. A character with 18 str but no training might have a 5% chance of success going purely by the dice roll - but they're not allowed to try. The 10th level fighter/rogue w/ 20 str, prof, expertise, and a stone of good luck though, can attempt to swim up a waterfall and they have a 55% chance to do it.

I think there's some false assumptions here. Namely, that everything is rolled for. Your 1st point is basically RAW anyway; only the GM calls for a roll. Ever. And they should only call for a roll when it's appropriate i.e. when the outcome is uncertain. A trained professional that's not under pressure, can and will RAW succeed at any task within their training.

Don't forget, an average professional has an ability score of 10 and only a +2 modifier, but without pressure, they effectively roll a 20 on the dice. That means any DC up to and including DC:22 is what any commoner in a trade can achieve on a day-to-day basis. This includes everything from a librarian finding a book in an organised library to a locksmith picking a lock; that stuff just happens when there's no pressure. Your second point presupposes that librarian and locksmith are rolling to do their day job.

That brings us to your 3rd point. If an average professional with a +2 can perform tasks up to DC:22 without fail, you can probably assume a PC adventurer with even +5 is damned well capable of a heck of a lot more, let alone +10 or higher. That's a baked in assumption of the RAW that you're supposing is just GM-fiat or whatever. Any idiot (quite literally) can walk up and touch a peak-Intelligence Archmages Major Illusion to disbelieve it and you only need an Int of 8 to look at one and realise it's fake if you have 10 rounds (1 minute). That's how shoddy illusions are. Assuming there's no other pressures, the same can be said of making almost any Saving Throw; DC's simply don't go above 20 whether it's for a Save or an ability check, outside of extreme cases, because you only roll when it matters.

With all that in mind, do you see where the system is already telling you what can or can't be done and what assumptions can be made, not only by random GM fiat, but by the guidance inherent within the DC's given?

[Blatant Beast]

Given that the arguments have finally entered Quantum Build territory, I think the only unknown us what causes this thread to get locked. Will it be because it hits 50 pages, or because people get into a fight and the thread is left on indefinite review?

Everyone has been mostly polite up to this point.

Speaking personally, those kind of super context-dependent skill checks and challenges is one of the things I hate the most about DM'ing. It's awful. I spend so much time on it, and all for something that's going to go by in a few rolls or a few minutes of play. It's such an ridiculously poor use of my prep time, and every time I do it, I wish there was a more generalized system in place.

10, 15, 20 is your friend, (or 5, 15, 20 if you want Easy Tasks to be easier). Then sprinkle in Advantage/Disadvantage for situational adjustments. "The Generalized System" is actually in place. An issue I think is the "generalized system" is too general for some tastes.

Now with that stated, there should absolutely be some published guidance on what skills can accomplish depending upon the type of game theme: a Jump to the moon style game vs a Wuxia game vs a gritty realism game, for example.

A couple of thoughts though... the "cleverness is a quality of the player" argument applies to everything, including your run-of-the-mill normie optimization as well.

Absolutely! That is the point indeed. As Skrum and Pex have shared, 5e has an abundance of rules that makes adjudicating a Fireball spell much easier for a DM than trying to use a skill in a way that might set a precedent.

Setting aside that aspect, I still find the Rogue Chassis to be less advantageous addition in terms of Party Survivability, due to my experience that "Save your bacon" abilities such as Bardic Inspiration or Flash of Genius are quite good at keeping people in the fight. Not everyone will rank such abilities as highly as I do, of course.

[Skrum]

Don't forget, an average professional has an ability score of 10 and only a +2 modifier, but without pressure, they effectively roll a 20 on the dice. That means any DC up to and including DC:22 is what any commoner in a trade can achieve on a day-to-day basis.

Ok is this actually true? Do you have a pg number? Legitimately curious. In the process of making my homebrew skill system, I've read the section on ability checks, and I don't recall anything like this - but that entire section is way too vague for my tastes (hence this entire conversation )

That brings us to your 3rd point. If an average professional with a +2 can perform tasks up to DC:22 without fail, you can probably assume a PC adventurer with even +5 is damned well capable of a heck of a lot more, let alone +10 or higher. That's a baked in assumption of the RAW that you're supposing is just GM-fiat or whatever. Any idiot (quite literally) can walk up and touch a peak-Intelligence Archmages Major Illusion to disbelieve it and you only need an Int of 8 to look at one and realise it's fake if you have 10 rounds (1 minute). That's how shoddy illusions are. Assuming there's no other pressures, the same can be said of making almost any Saving Throw; DC's simply don't go above 20 whether it's for a Save or an ability check, outside of extreme cases, because you only roll when it matters.

What skills can do when there's no pressure is a lot less relevant than what skills can do when there is pressure. Handwaving various things, especially day-to-day ones, is hard-baked into RPG's. Even BG3's software isn't rolling for whatever NPC's are doing in the background, they're just animated to be doing whatever it is they're doing.

But under pressure (making rolls)....
1d20+5 has a 73.5% chance of being equal to or greater than 1d20. That means a 1st-4th level character with 16 in the relevant stat and proficiency with the relevant skill will only outright beat a untrained character with 10 in the relevant stat ~3 out of 4 times. That's like...what real-world task looks ANYTHING like that. Or put another way, that same trained character has a 55% chance to do a DC 15 task, while the no-talent novice has a 30% chance of success. Like what the heck is that.

This dynamic is why I hate making DC's for things, and why it feels so unsatisfying to me. Should a hard task (tracking a creature through a snow storm, convincing a knight to "loan" you their sword, jumping 15 ft farther than you've ever jumped before) be DC 25, thus metaphysically impossible for an untrained person? Well that also means the expert of experts, a 7th level rogue with proficiency, expertise, and +4 in the relevant ability score has only a 30% chance of success. That's not something they can count on using, ever. It's a moonshot, when the class' entire schtick is being super-skilled. Or maybe it's DC 15 - now the rogue is very likely to succeed, but any clown who cares to try has a decent off-chance of success at what should be a kind of crazy thing.

And if the answer is "make difference DC's for different creatures," well mathematically you're just doing an ad hoc +5 or +7 or whatever bonus to the character that "should" be making the roll. If that's where they should be, give them that bonus explicitly! Or honestly, why are we rolling at all?? Just narrate the characters doing cool things, and save the rolls for combat.

With all that in mind, do you see where the system is already telling you what can or can't be done and what assumptions can be made, not only by random GM fiat, but by the guidance inherent within the DC's given?

Oh I can see exactly why people ascribe all kinds of qualities to 5e's skills - it has almost no qualities at all, thus it can be anything to anyone. Not my cup of tea.

Setting aside that aspect, I still find the Rogue Chassis to be less advantageous addition in terms of Party Survivability, due to my experience that "Save your bacon" abilities such as Bardic Inspiration or Flash of Genius are quite good at keeping people in the fight. Not everyone will rank such abilities as highly as I do, of course.

Completely agree; I made a similar point way back on page 1 about the rogue's presence (or lack thereof). They don't tank, they don't have any abilities that support or boost allies, or debuff enemies; the rogue ideally wants to be "not there" when it's not their turn. This puts a lot of extra pressure on the other members of the party.

[Pex]

I'm not talking about mimicking specific spells, I'm talking about high DC skill uses that create effect commensurate in scope to 3rd+ level spells. This would would be enormously helpful in martial classes scaling in similar ways to casters.

Off the top of my head, I'd love to see skill checks allowing stuff like
- using stealth to turn invisible
- using athletics (grappling) to counter teleporting
- using acrobatics for teleport-like infallible escapes
- using perception to see invisible creatures and objects
- using medicine to bring someone back from the dead
- using intimidate to create fear effects
- using persuasion to charm
- using performance for mass hypnosis, suggestion, fear, inspiration, etc
- using insight to read minds or see the future

We agree on principle, but I do think it's ok for spellcasters to have their own unique niche. As an extreme example I don't need a fighter to cut a hole in reality to travel into another plane. Let spellcasters own far traveling. Medicine to do CPR to mimic Revivify, sure, but not bring back someone dead for three days like Raise Dead. I suppose it's a matter of degree. These ideas are worth considering, but the devil is in the details of how and maybe not everything. You are certainly not responsible to provide details on how you would implement these ideas or else forget about it. It's enough just to talk about their worthiness.

Cheers

[LudicSavant]

Ok is this actually true? Do you have a pg number? Legitimately curious. In the process of making my homebrew skill system, I've read the section on ability checks, and I don't recall anything like this - but that entire section is way too vague for my tastes (hence this entire conversation )

Average commoners have 10s across the board, but average professionals (as opposed to unskilled labor) tend to have higher stats, especially in their primary ability score (reference: MM appendix B).

As for the other thing, 5e's version of the take 20 rule can be found on pg237 of the DMG:

Sometimes a character fails an ability check and wants to try again. In some cases, a character is free to do so; the only real cost is the time it takes. With enough attempts and enough time, a character should eventually succeed at the task. To speed things up, assume that a character spending ten times the normal amount of time needed to complete a task automatically succeeds at that task. However, no amount of repeating the check allows a character to turn an impossible task into a successful one.

e.g. You can do a task at 1/10th speed in order to auto-succeed, so long as that task is possible.

[Pex]

Being stuck with bounded accuracy you can only spell out so much before either: arming players to run off with a baseline N% success rate entitling then to at will performance of tasks in low risk situations, or writing out that skills are always only as strong or as weak as the GM wants for the scene.

That's just the math. You can have other metrics. Skill Use Awesome Thing may be DC 15 but you can't use the skill for that Awesome Thing until level X. Anyone can do Skill Trick, but only Proficieny can do Cool Skill Trick and Expertise is needed for Cool Skill Trick++. To be base about it just as there are Spell Levels there can be Skill Use levels, with some Skill Uses anyone can do at character level X, others based on having proficiency and expertise at character level X, and a few only at class of level X. Devil in the details. Not saying it has to be implemented this way. If someone doesn't like this, I don't care. I'm not married to it. It's just an idea of one way to do it.

[Xervous]

That's just the math. You can have other metrics. Skill Use Awesome Thing may be DC 15 but you can't use the skill for that Awesome Thing until level X. Anyone can do Skill Trick, but only Proficieny can do Cool Skill Trick and Expertise is needed for Cool Skill Trick++. To be base about it just as there are Spell Levels there can be Skill Use levels, with some Skill Uses anyone can do at character level X, others based on having proficiency and expertise at character level X, and a few only at class of level X. Devil in the details. Not saying it has to be implemented this way. If someone doesn't like this, I don't care. I'm not married to it. It's just an idea of one way to do it.

It’s one thing to add examples to place the functionality of skills in a general ballpark, it’s another thing to write up a system. It was my understanding at the time that the topic in consideration was clarifying and giving a bit of standardization to the existing system.

Handing out different tiers of skill permissions is indeed the cleanest way to expand on a bounded accuracy system. Such an expansion would require a lot of detail and specifics in order to avoid causing additional confusion, and I suspect this would be too explicit a definition for those opposed to DC charts.

All that being said, I would welcome such a system if it let me know what guarantees are in store for the highly skilled rogue I am considering for a campaign without having to play 20 questions with the GM for each of a variety of topics.

[Crusher]

This dynamic is why I hate making DC's for things, and why it feels so unsatisfying to me.

100% agree with this and your logic. Even just keeping it within the party, in theory it should be surprising and be an opportunity for fun roleplay if the illiterate INT-dump stat Barbarian happens to know the magic clue and the 20 INT Arcana Proficiency Wizard does not, because the Barbarian rolled a 20 and the Wizard rolled a 3. Because in theory it should be rare and noteworthy. But at low levels it happens *all the time* (if you have multiple 8 INT unskilled martials allowed to roll, the proficient 16 INT Wizard will probably only beat them collectively around half the time) which just makes it frustrating.

Only letting people roll who have actual proficiency in a skill (or otherwise have a clear reason to know the information/be skilled at the task/etc.) helps a lot with this, though it still doesn't feel like a great solution. I get what they're trying to do, but it doesn't feel like the gap between being skilled vs unskilled at a task at low levels is a big enough boost.

[Theodoxus]

It’s one thing to add examples to place the functionality of skills in a general ballpark, it’s another thing to write up a system. It was my understanding at the time that the topic in consideration was clarifying and giving a bit of standardization to the existing system.

Handing out different tiers of skill permissions is indeed the cleanest way to expand on a bounded accuracy system. Such an expansion would require a lot of detail and specifics in order to avoid causing additional confusion, and I suspect this would be too explicit a definition for those opposed to DC charts.

All that being said, I would welcome such a system if it let me know what guarantees are in store for the highly skilled rogue I am considering for a campaign without having to play 20 questions with the GM for each of a variety of topics.

Skrum started a thread on just this topic. And I've been ruminating on such a beast after reading the last 3 pages on this thread.

I think 5E really did a disservice with moving away from skill levels. I do think 3.PF was a bit too graduated; one doesn't need 4+1 per level to differentiate between knowledgeable and just lucky.

I do like the idea of tiers for skill proficiency (beyond just proficiency bonus, which just allows the dumb brute at any level to have as much luck in knowing something as everyone else). I also agree that the genre of the game should also fit the skill system used (and I really liked Kan0's idea of allowing expertise to 'bump up' the skill genre - though I'm not totally married to it, if it ends up being problematic.

That said, stealing from both Skrum's post, and my own system I've been tinkering with for a few years, I'd go with:

Untrained: no proficiency. Can auto-succeed on a DC0 check, can attempt higher DCs, but can't succeed higher than a DC15 result no matter the roll.
Novice: standard 5E skill level at 1st level. +2 PB bonus. Auto-succeeds at DC5 tasks, has no cap on what they can try for.
Expert: Equivalent to Expertise. Auto-succeeds at DC10 tasks.
Master: No equivalent in 5E. Auto-succeeds at DC15 tasks.
Adept: Auto-succeeds at DC20 tasks.
Grandmaster: Auto-succeeds at DC25 tasks.

The question becomes how does one elevate beyond Novice (or Expert, if the only way to be an Expert is through the Expertise mechanic). My system uses successive successes, taking time to slowly gain mastery over a specific skill. I could see Boons, Divine Gifts, Time, Experience Points, new Feats, etc. as possible ways as well.

The other way I was thinking might work (though I guess it ultimately falls on the DM to adjudicate, so less good for those folks that don't want more work) is to not have a DC at all, but let the player roll and then decide what the roll represents. So, if you want to climb a tree, and roll a modified 13, the DM then states something like "you easily climb the tree and can get up to 30', where the limbs become too weak to support your weight. How high do you want to climb?"
It's particularly useful for knowledge checks; you don't need to know a DC to find a specific name in a tome or whatever; they roll, and you let them know how much information they get.

[Blatant Beast]

This dynamic is why I hate making DC's for things, and why it feels so unsatisfying to me. Should a hard task (tracking a creature through a snow storm, convincing a knight to "loan" you their sword, jumping 15 ft farther than you've ever jumped before) be DC 25, thus metaphysically impossible for an untrained person? Well that also means the expert of experts, a 7th level rogue with proficiency, expertise, and +4 in the relevant ability score has only a 30% chance of success. That's not something they can count on using, ever. It's a moonshot, when the class' entire schtick is being super-skilled. Or maybe it's DC 15 - now the rogue is very likely to succeed, but any clown who cares to try has a decent off-chance of success at what should be a kind of crazy thing.

The simplest way to address this issue is to assume that PCs have some degree of Destiny in their fate, and have the chance to influence events, in a manner most people will not.

The average green grocer would never have a chance, (and thus no dice are rolled), to convince a strange Knight to give them their sword. A PC does have a chance, (albeit, potentially only a very small chance). Giving the PC opportunity to roll dice, does not necessarily set a precedent for NPCs.

3e created a rules ecosystem.
5e creates a rules resolution rule, the D20 Test, and does not care too much about creating a rules ecosystem.

Each approach has it's merits and demerits. In the vein of the film Dr. Stangelove, the title of D&D's currently running film is: 5e or: How I learned to stop worrying and love the isolated, non-ecosystem based Ability Check. :)

[Psyren]

Everyone has been mostly polite up to this point.

Yeah I don't get where the sudden spurt of fear/threadcrapping is coming from. Everything looks civil to me, even in the areas where we have impassable misalignment or mistaken/irreconcilable assumptions.

10, 15, 20 is your friend, (or 5, 15, 20 if you want Easy Tasks to be easier). Then sprinkle in Advantage/Disadvantage for situational adjustments. "The Generalized System" is actually in place. An issue I think is the "generalized system" is too general for some tastes.

Now with that stated, there should absolutely be some published guidance on what skills can accomplish depending upon the type of game theme: a Jump to the moon style game vs a Wuxia game vs a gritty realism game, for example.

My thing about Wuxia is that it really should be its own supplement, similar to ELH and Mythic Adventures before it. I think it's fine for the DMG to nod in that direction and say "this exists" but the focus of the base game should really be on Heroic Fantasy, and maybe Gritty Sword & Sorcery as a common alternative, but the more campaign styles you flesh out in core the more cluttered core becomes. Things like Horror and Wuxia and Mass Combat/War and No-combat/Political Intrigue/Slice of Life are all valid ways to play D&D, but don't need to be core.

Absolutely! That is the point indeed. As Skrum and Pex have shared, 5e has an abundance of rules that makes adjudicating a Fireball spell much easier for a DM than trying to use a skill in a way that might set a precedent.

I don't see what's so challenging about being up front with your players, that a given skill ruling or DC doesn't set a precedent. All of them are specific to a given situation, scenario or encounter; be clear about setting that expectation going into the campaign. You yourself may choose to reuse a DC or creative ability check use because of a similarity the new situation has to a previous one, but it should always be your choice as the DM to do so. And rationalizing differences isn't hard either, because you control the world.

("You said it was Easy for us to sneak past the guard outpost last time!" "Yeah, but those guards were lazy/drunk/loudly gambling/fighting and thus weren't paying attention to what might be happening outside. These guards are closer to the Duke's manor and take their jobs a bit more seriously, so it's a tougher challenge this time. You can also take a quick look around for alternatives if you want.")

Setting aside that aspect, I still find the Rogue Chassis to be less advantageous addition in terms of Party Survivability, due to my experience that "Save your bacon" abilities such as Bardic Inspiration or Flash of Genius are quite good at keeping people in the fight. Not everyone will rank such abilities as highly as I do, of course.

The rogue is indeed not a team player in the same sense as a Bard or even a Ranger. But I think that's okay, because that's part of the class fantasy. Going all the way back to Tolkien, Bilbo wasn't brought on so that he could play a lute and buff the Dwarves from the back row, he was brought on so he could go off by himself and do things the rest of them weren't doing. That's still being a team player, just indirectly.

[Blatant Beast]

My thing about Wuxia is that it really should be its own supplement, similar to ELH and Mythic Adventures before it.

The rules module component for different styles of play that was advertised in the D&D Next playtest packets never really materialized. Honestly, I am not sure that Wuxia style is all that different from Heroic style at this point in the popular zeitgeist.

I don't see what's so challenging about being up front with your players, that a given skill ruling or DC doesn't set a precedent. All of them are specific to a given situation, scenario or encounter; be clear about setting that expectation going into the campaign.

I think a major aspect is you are not trying to apply a rules ecosystem. You accept, and convey that that 5e has a Rules Resolution rule, and each roll is mostly independent of each other.

I do not think that is how Skrum nor Pex approach the rules, for example. Of course, I can not speak for them, but I think the heart of the differences in approach lies in the ecosystem vs independent rolls mindset.

The rogue is indeed not a team player in the same sense as a Bard or even a Ranger. But I think that's okay, because that's part of the class fantasy. Going all the way back to Tolkien, Bilbo wasn't brought on so that he could play a lute and buff the Dwarves from the back row, he was brought on so he could go off by himself and do things the rest of them weren't doing. That's still being a team player, just indirectly.

Thematically, that is fine, and to restate again for utmost clarity: nobody in the thread is saying Rogues are unplayable nor an unfun class. There is, however, a substantial difference between literature and RPG gameplay.

Bilbo Baggins, in an RPG sense, would have been played by a person that was either late to or missed a lot of game sessions:

• Bilbo's player was over an hour late to one of the early sessions, but luckily the player arrived in time to confuse the trolls and save the day.
• The Goblins kidnap the party in the cave in the Misty Mountains, but Bilbo's player is missing, and Bilbo receives a solo adventure.
• The structure of battling Smaug in the Hobbit bears no resemblance to an actual D&D game, at least not a good one. ("The fierce Dragon Smaug flies past you, and dies to an NPC. Sorry folks, no XP for You!!").
• In the Battle of Five Armies, Bilbo yet again does next to nothing.

Such a setup, is fine for a novel, but is atrocious for an actual RPG game.

In a RPG designed for today's audience, all classes should have built in, non feat based ways of directly helping their friends, is my opinion. Perhaps the skill expert could help others with a Guidance like ability or the Rogue can swap their Ability Check with their friends roll, etc, etc. The design space seems open enough to accommodate abilities like this.. If Psi Knights can protect others from damage, seems fine to have a Rogue be able to give pointers to someone, (and thus a bonus), on an ability check.

[Pex]

I mean, true, but it's another way of saying "the DM will cater the challenge to the party and make up for the fact the rogue has less to offer and needs special considerations." Which is exactly the point made (20 pages ago? I lost count).



I think there's a few things here that could be useful (stuff that should've been part of the base game, if WotC had made a proper skill system).

1) Not everyone can roll for everything
If the bonuses are generally going to be low (bounded), then the dice roll becomes that much more important. To prevent the feel-bad chaos of a moron being more knowledgeable than the sage, or the weakling out-lifting the strongman, there needs to be guardrails on what can be rolled for. The DC might be 13 but that doesn't mean an entirely untrained person with an 8 in the associated trait has a 35% chance of succeeding - it means it's quite a hard task and if you don't have proficiency, you don't get to roll. Essentially, separate DC (a game construct) from the in-world difficulty of the task. I personally like DC 10 being the cap on untrained checks.

2) Shift DC's downwards
Most skill checks should be DC 10 or less, reflective of the fact that a trained person (proficiency) with reasonable talent (~+2 in the associated ability) will have a +5 check. They should succeeded at most tasks they are trained in most of the time. Many tasks? Don't even roll. They're DC 5. A character with a +5 or +6, they mathematically can't fail that.

3) Create unique uses of skills for characters with large modifiers
As a continuation of (1), a character with a +10 modifier should be able to attempt to do things a character with merely a +7 modifier cannot. This is both to have skills scale a little more, and also make more explicit how extraordinary someone with a +10 or +12 is. They aren't only 35% more likely to succeed; they can attempt things other people would consider impossible. I would probably gate these high level, extraordinary tasks behind expertise: with proficiency, you can attempt tasks up to DC 20. Expertise, and 21+ is open to you. Let's say swimming up a waterfall is DC 24. A character with 18 str but no training might have a 5% chance of success going purely by the dice roll - but they're not allowed to try. The 10th level fighter/rogue w/ 20 str, prof, expertise, and a stone of good luck though, can attempt to swim up a waterfall and they have a 55% chance to do it.

I'm guilty of this too. If we limit fantastical things to Expertise then we are saying only rogues and bards need apply. Swimming up a waterfall is cool, but it feels bad a fighter must multiclass to do it. This clearly becomes a case of multiclassing makes you more powerful than staying single class. In a sense it already does by the game as is now, but that's by accident (in theory). It shouldn't be on purpose. The solution then is to give everyone Expertise slots. Rogues and Bards can have more as their thing, but everyone gets some to allocate Expertise. If we are to gate some skill uses to Proficiency then characters need to be given more than they are now of 4, two from background, two from class. A few races and classes give one or two more but not all. If all do such that you get access to Proficiencies from different sources but in total you get more than 4 that's fine. Where they come from is less important than how many you get in total. It's fine for some characters to have more than others because of class, but you still need more than 4 total as a minimum you get now. How many and how to get them is up to the hypothetical game designers.

[Skrum]

I think a major aspect is you are not trying to apply a rules ecosystem. You accept, and convey that that 5e has a Rules Resolution rule, and each roll is mostly independent of each other.

I do not think that is how Skrum nor Pex approach the rules, for example. Of course, I can not speak for them, but I think the heart of the differences in approach lies in the ecosystem vs independent rolls mindset.

An astute observation - I for one pine for the days of 3e's Theory of Everything approach to characters, NPCs, monsters, etc. As a DM and as a player I want a structure to work within and guide how I think about what's happening, and most importantly, what can and will happen next. I do not vibe with a rules-lite, make it up as you go style system.

And the thing is, 5e does not have that kind of rules-lite approach to combat. In fact, bounded accuracy implies to me that each and every number is super-duper important and shouldn't be "messed with" or changed because why have bounded accuracy if it can just be broken all the time. But skills are the total opposite - what do the numbers mean, who knows, make it up!

[Psyren]

The rules module component for different styles of play that was advertised in the D&D Next playtest packets never really materialized. Honestly, I am not sure that Wuxia style is all that different from Heroic style at this point in the popular zeitgeist.

I see it as massively different. For me, Heroic Fantasy embodies two tropes in particular:

Zero to Hero: You most often start your career as a character with humble beginnings, like a farmer, recruit/militiaman, apprentice, church acolyte or pickpocket. As DMG 38 describes it: "These characters typically come from ordinary backgrounds, but something impels them into an adventuring life."

The World is Feudal: That means even moderately powerful magic is relatively rare and most large-scale threats are handled via standing armies or elite mercenaries. The PCs and other impactful characters in the setting are allowed to deviate from that - generally with the assistance of rare magic items and/or unique training - but absent such heroes, royalty and other organizations rely on their militias, soldiers and knights to deal with many threats. DMG 38 again: "Technology and society are based on medieval norms, though the culture isn't necessarily European."

Mythic Fantasy and Wuxia meanwhile deviate from this pretty sharply, with premises like "the PCs are demigods drawing on their divine heritage" and "acrobatic leaps are equal to teleportation." Again, there's nothing wrong with the latter if you consider them fun, but the DMG didn't put those examples under Heroic Fantasy for a reason.

I think a major aspect is you are not trying to apply a rules ecosystem. You accept, and convey that that 5e has a Rules Resolution rule, and each roll is mostly independent of each other.

I do not think that is how Skrum nor Pex approach the rules, for example. Of course, I can not speak for them, but I think the heart of the differences in approach lies in the ecosystem vs independent rolls mindset.

I think that's right, and I also think that designer intent is closer to my approach than it is to theirs. Certainly treating rolls as individual/independent situations seems more in line with the way they've described the system and how it errs more on the side of DM than player agency.

Thematically, that is fine, and to restate again for utmost clarity, nobody in the thread is saying Rogues are unplayable nor an unfun class. There is also a difference between literature and RPG gameplay.

Bilbo Baggins, in an RPG sense, would have been played by a person that was either late to or missed a lot of game sessions:

• Bilbo's player was over an hour late to one of the early sessions, but luckily the player arrived in time to confuse the trolls and save the day.
• The Goblins kidnap the party in the cave in the Misty Mountains, but Bilbo's player is missing, and Bilbo receives a solo adventure.
• The structure of battling Smaug in the Hobbit bears no resemblance to an actual D&D game, at least not a good one. ("The fierce Dragon Smaug flies past you, and dies to an NPC. Sorry folks, no XP for You!!").
• In the Battle of Five Armies, Bilbo yet again does next to nothing.

Such a setup, is fine for a novel, but is atrocious for an actual RPG game.

In a RPG designed for today's audience, all classes should have built in, non feat based ways of helping their friends, is my opinion. Perhaps the skill expert could help others with a Guidance like ability or the Rogue can swap their Ability Check with their friends roll, etc, etc. The design space seems open enough to accommodate abilities like this.. If Psi Knights can protect others from damage, seems fine to have a Rogue be able to give pointers to someone, (and thus a bonus), to an ability check.

1) This is why I like Cunning Strike, because debuffing an enemy is mathematically similar to buffing an ally (or multiple allies) in most cases, and fits perfectly with the rogue's class fantasy of being an opportunist who fights dirty.

2) I'm not saying Bilbo is the ideal or quintessential D&D rogue, he does indeed go off on his own a lot more than a typical D&D player would, but the principle is the same - they didn't bring him along so that he would be doing all the same things they were, but rather specifically because he wouldn't be. There are other examples of this archetype too, like Batman in most Justice League stories, Gonff from Redwall etc.

[Pex]

Skrum started a thread on just this topic. And I've been ruminating on such a beast after reading the last 3 pages on this thread.

I think 5E really did a disservice with moving away from skill levels. I do think 3.PF was a bit too graduated; one doesn't need 4+1 per level to differentiate between knowledgeable and just lucky.

I do like the idea of tiers for skill proficiency (beyond just proficiency bonus, which just allows the dumb brute at any level to have as much luck in knowing something as everyone else). I also agree that the genre of the game should also fit the skill system used (and I really liked Kan0's idea of allowing expertise to 'bump up' the skill genre - though I'm not totally married to it, if it ends up being problematic.

That said, stealing from both Skrum's post, and my own system I've been tinkering with for a few years, I'd go with:

Untrained: no proficiency. Can auto-succeed on a DC0 check, can attempt higher DCs, but can't succeed higher than a DC15 result no matter the roll.
Novice: standard 5E skill level at 1st level. +2 PB bonus. Auto-succeeds at DC5 tasks, has no cap on what they can try for.
Expert: Equivalent to Expertise. Auto-succeeds at DC10 tasks.
Master: No equivalent in 5E. Auto-succeeds at DC15 tasks.
Adept: Auto-succeeds at DC20 tasks.
Grandmaster: Auto-succeeds at DC25 tasks.

The question becomes how does one elevate beyond Novice (or Expert, if the only way to be an Expert is through the Expertise mechanic). My system uses successive successes, taking time to slowly gain mastery over a specific skill. I could see Boons, Divine Gifts, Time, Experience Points, new Feats, etc. as possible ways as well.

The other way I was thinking might work (though I guess it ultimately falls on the DM to adjudicate, so less good for those folks that don't want more work) is to not have a DC at all, but let the player roll and then decide what the roll represents. So, if you want to climb a tree, and roll a modified 13, the DM then states something like "you easily climb the tree and can get up to 30', where the limbs become too weak to support your weight. How high do you want to climb?"
It's particularly useful for knowledge checks; you don't need to know a DC to find a specific name in a tome or whatever; they roll, and you let them know how much information they get.

The weakness of this is what makes a task DC 10, DC 15, DC 20? If the game designers put ink on paper the game rules say This Task is DC 20. This Other Task is DC 15, then fine. They can even offer add or subtract 5 for gritty/wuxia games. Since it's impossible to list everything imaginable of all things skills related, the DM will have a ballpark of where to place difficulty DCs for such things not printed ink on paper. However, if you still leave it to DM decides/Make it up, then we're no better off because what one DM says is DC 15 another will say DC 20 and we're back to my character can only do things based on who is DM that day. Neither is playing the game wrong. They just disagree on the difficulty of the task, and the rogue has a poor reputation because the DM won't let a player do cool stuff.

I knew I would get to my infamous rant eventually.

[Skrum]

I'm guilty of this too. If we limit fantastical things to Expertise then we are saying only rogues and bards need apply. Swimming up a waterfall is cool, but it feels bad a fighter must multiclass to do it. This clearly becomes a case of multiclassing makes you more powerful than staying single class. In a sense it already does by the game as is now, but that's by accident (in theory). It shouldn't be on purpose. The solution then is to give everyone Expertise slots. Rogues and Bards can have more as their thing, but everyone gets some to allocate Expertise. If we are to gate some skill uses to Proficiency then characters need to be given more than they are now of 4, two from background, two from class. A few races and classes give one or two more but not all. If all do such that you get access to Proficiencies from different sources but in total you get more than 4 that's fine. Where they come from is less important than how many you get in total. It's fine for some characters to have more than others because of class, but you still need more than 4 total as a minimum you get now. How many and how to get them is up to the hypothetical game designers.

I'm in the process of redoing the skill system (though to a less drastic degree than I posted about here), and as part of that I've made Skill Tricks; little maneuvers associated with each skill that someone with expertise in that skill can use. Since they're relatively small benefits, having rogues and bards get access to them natively and other classes can pick something they like via Skill Expert would work fine.

But you are correct; if I were to go all the way to what I described above, giving more classes native access to expertise would be necessary. Off the top of my head, something like

- Normalize getting 2 skill prof from backgrounds (I think this is happening in One)
- Rogues are the king of skills, gaining 4 skill proficiencies at 1st level and an additional proficiency every 4 levels after
- Fighters, barbs, monks, rangers, rangers, paladins, and artificers get 3 skill prof at 1st and an additional prof at 10th
- Sorcerers and wizards get 2 skill prof + arcana, clerics get 2 + religion, and druids get 2 + nature. No additional prof as they level
- Bards get....2 prof + performance. No additional prof as they level
- Move rogue expertise from level 1 to level 3, and give them an additional 2 at 12th. In total, they'd get 2 at 3rd, 2 at 6th, and 2 at 12th
- Fighters et al get an single expertise every 5 levels
- The full casters gain expertise in their associated skill (arcana, religion, nature, performance) at level 5

This could be a whole other thread, but bards should not get expertise the way they do. It's a little weird they get it AND jack of all trades. Seriously, what the heck is that about? Up to me, I'd take away expertise and lean into Jack of all trades a bit - probably have that scale up over the levels to prof in all skills (thus giving bards access to prof-tier skill checks, maybe by level 10). And then maybe in tier 3 they can get an ability to make an expertise-level skill check 1/LR or something.

[Blatant Beast]

And the thing is, 5e does not have that kind of rules-lite approach to combat. In fact, bounded accuracy implies to me that each and every number is super-duper important and shouldn't be "messed with" or changed because why have bounded accuracy if it can just be broken all the time. But skills are the total opposite - what do the numbers mean, who knows, make it up!

5e's Tactical complexity is more about the spacing, initiative, the variance between spell attacks and weapon attacks in terms of adding ability score modifiers to damage, and things like the line of sight rules that come into play once combat starts. Resolving how to hit and roll for damage is not very different from the Ability Check, mechanic.

Bounded Accuracy is just a design philosophy, not an actual rules presence. In general, a leveled PC will be able to handle appropriate challenges without requiring the PC to have a specific degree of wealth or items.

4e, by contrast, at initial release did assume that at certain level breaks PCs would have certain types of exotic armor. In practice, as a DM, you either had to work in an opportunity to give the players the expected amount of loot, or put in stores and give them the cash to buy it. If you as the DM did neither of these, then the players were as a matter of baseline performance, unable to perform up to the level the system expected.

I've allowed opportunity for PCs to discover +5 weapons as 3rd level PCs in 5e games. The system does not break. Of course a PC with a +5 weapon is going to be more accurate and deal more damage as a consequence, but this is very easy to account for, as a DM. 5e is a robust system and takes modification fairly well.

I see it as massively different. For me, Heroic Fantasy embodies two tropes in particular:

Zero to Hero: You most often start your career as a character with humble beginnings, like a farmer, recruit/militiaman, apprentice, church acolyte or pickpocket. As DMG 38 describes it: "These characters typically come from ordinary backgrounds, but something impels them into an adventuring life."

Wuxia stories has this as well, the poor peasant that goes to the monastery and excels at the sweet, sweet gong-fu. Starting out as a Demi-god at level one, strikes me as Super Hero, not Wuxia, but I see where you are coming from.

I think that's right, and I also think that designer intent is closer to my approach than it is to theirs. Certainly treating rolls as individual/independent situations seems more in line with the way they've described the system and how it errs more on the side of DM than player agency.

I would agree the game has much less aggravation when one learns to love the non ecosystem based, independent d20 test. That said, clearly, there are a sizable number of people that like ecosystem based RPGs.

1) This is why I like Cunning Strike, because debuffing an enemy is mathematically similar to buffing an ally (or multiple allies) in most cases, and fits perfectly with the rogue's class fantasy of being an opportunist who fights dirty.

It is a start. The Rogue class is still tied a bit too much thematically with the old AD&D Thief, for my taste, sometimes. Why does a debuff have to be an attack, for example, is a design question I would ask myself, if I was on the WotC design staff.

However, if you still leave it to DM decides/Make it up, then we're no better off because what one DM says is DC 15 another will say DC 20 and we're back to my character can only do things based on who is DM that day.

That is just the name of the game. Two judges, will handle their courtrooms differently, and sometimes arrive a different conclusions, the same is true with DMs. If every game ran the exact same way, every time it was played, then replay value goes out the window.

I've ran and played Village of Hommlet/Temple of Elemental at least a score of times. That would not be possible if the module was always the same, and always handled the same, despite a change in participants....it would just be too boring.

Static Tumble DC's in 3e did not turn moving through another creatures occupied space into some sort of exciting moment, in which the whole table goes silent while watching the roll with baited breath, to see the crucial outcome.

Instead static Tumble DCs, turned into a Character Optimization challenge, that actively killed any potential excitement from a Tumble check, because people only took enough points to achieve the auto-success threshold.

I'm sorry Pex, but what you want in this regard, makes for a dull game, in my opinion.

[Skrum]

Instead static Tumble DCs, turned into a Character Optimization challenge, that actively killed any potential excitement from a Tumble check, because people only took enough points to achieve the auto-success threshold.

I'm sorry Pex, but what you want in this regard, makes for a dull game, in my opinion.

I mean, that thread I started is about that - what if the entire concept of "skill checks" was redone and largely thrown out. What if being "adept" at [tumble] meant you could move passed one enemy each turn without drawing OA's from that enemy, no roll necessary.

Basically, rather than skills being a check with DC's and random outcomes, they were more like a skill tree, and the further down the tree one got the more abilities are gained.

There'd still be space for skill rolls (and tension, if indeed tension comes from uncertain outcomes) in the form of opposing checks - grappling, for instance - but for the most part skills would look like "I have expert-level athletics, I climb the wall with my full land speed."

[Kane0]

I do like the idea of tiers for skill proficiency (beyond just proficiency bonus, which just allows the dumb brute at any level to have as much luck in knowing something as everyone else). I also agree that the genre of the game should also fit the skill system used (and I really liked Kan0's idea of allowing expertise to 'bump up' the skill genre - though I'm not totally married to it, if it ends up being problematic.

That said, stealing from both Skrum's post, and my own system I've been tinkering with for a few years, I'd go with:

Untrained: no proficiency. Can auto-succeed on a DC0 check, can attempt higher DCs, but can't succeed higher than a DC15 result no matter the roll.
Novice: standard 5E skill level at 1st level. +2 PB bonus. Auto-succeeds at DC5 tasks, has no cap on what they can try for.
Expert: Equivalent to Expertise. Auto-succeeds at DC10 tasks.
Master: No equivalent in 5E. Auto-succeeds at DC15 tasks.
Adept: Auto-succeeds at DC20 tasks.
Grandmaster: Auto-succeeds at DC25 tasks.

Ooh, got a thread? IIRC PF2 has a similar tiered skill setup, but I think that largely governs the bonus you apply and feats you can qualify for more than unlocking additional capabilities on its own.

Perhaps as you gain steps on the skill ladder you unlock the ability to attempt higher DCs rather than auto-succeed on the lower ones? In addition to the higher modifier of course. I think i'd go with four stages (Talented/Trained, Expert, Adept, Master makes for an acronym) to correspond with the 4 tiers of play (and a bunch of other stuff that comes in fours like subclass breakpoints, cantrip progression, etc)

[Psyren]

It is a start. The Rogue class is still tied a bit too much thematically with the old AD&D Thief, for my taste, sometimes. Why does a debuff have to be an attack, for example, is a design question I would ask myself, if I was on the WotC design staff.

I mean, it doesn't - you can improvise all kinds of actions/conditions like sneaking up and tying an orc's bootlaces together or throwing sand in their eyes etc. But the fact that you can impose debuffs AND do damage is a situation I vastly prefer.

That is just the name of the game. Two judges, will handle their courtrooms differently, and sometimes arrive a different conclusions, the same is true with DMs. If every game ran the exact same way, every time it was played, then replay value goes out the window.

I've ran and played Village of Hommlet/Temple of Elemental at least a score of times. That would not be possible if the module was always the same, and always handled the same, despite a change in participants....it would just be too boring.

Static Tumble DC's in 3e did not turn moving through another creatures occupied space into some sort of exciting moment, in which the whole table goes silent while watching the roll with baited breath, to see the crucial outcome.

Instead static Tumble DCs, turned into a Character Optimization challenge, that actively killed any potential excitement from a Tumble check, because people only took enough points to achieve the auto-success threshold.

I'm sorry Pex, but what you want in this regard, makes for a dull game, in my opinion.

Agreed - and even if two DMs run different DCs that doesn't mean the outcomes have to be that different.

If DM A says DC 15 to jump the chasm and you get a 17 to clear it, and DM B says DC 20 to jump the chasm and you roll a 17 and he says partial success, you made it but skinned your knees on the other side (take some damage) or you made it but dropped one of your daggers, (pick one of the orcs' pockets on the other side) - either way you've cleared the chasm.

[Pex]

That is just the name of the game. Two judges, will handle their courtrooms differently, and sometimes arrive a different conclusions, the same is true with DMs. If every game ran the exact same way, every time it was played, then replay value goes out the window.

I've ran and played Village of Hommlet/Temple of Elemental at least a score of times. That would not be possible if the module was always the same, and always handled the same, despite a change in participants....it would just be too boring.

Static Tumble DC's in 3e did not turn moving through another creatures occupied space into some sort of exciting moment, in which the whole table goes silent while watching the roll with baited breath, to see the crucial outcome.

Instead static Tumble DCs, turned into a Character Optimization challenge, that actively killed any potential excitement from a Tumble check, because people only took enough points to achieve the auto-success threshold.

I'm sorry Pex, but what you want in this regard, makes for a dull game, in my opinion.

Here we go again.

Yet the game functions fine for every campaign every DM every game every long sword does 1d8 damage, every platemail is AC 18, and all Fireballs cast at 3rd level spell slot does 8d6 fire damage in a 20 ft radius. I've run Dragonheist twice. They were different because the players were different making different decisions. My villain for both campaigns was they were all in on it, but it was the bartender of the Yawning Portal who was ultimately responsible for the Nimblewright using the Fireball necklace because he was a Harper and wanted the Stone of Golorr for himself to destroy it because it was an evil artifact. He didn't care about the vault. The first group trusted the Cassalanters too much telling them everything and lost a lot of time and info for not going into the Gralhund Villa. In the second group one player chose to befriend Davill and joined the Zhentarim. Davill became a father figure and Yaggra became his girlfriend accidentally because she misinterpreted his intentions wanting her to teach him Thieves' Cant. Same module, same statistics, same general events, same NPCs, same villain, same DM, yet totally different games. It is the players (DM included) who make each game different. That doesn't go away just because a hypothetical rule says a tree is DC 10 to climb.

[Psyren]

Here we go again.

Yet the game functions fine for every campaign every DM every game every long sword does 1d8 damage, every platemail is AC 18, and all Fireballs cast at 3rd level spell slot does 8d6 fire damage in a 20 ft radius.

(Here we go again indeed...)

If you truly see no difference between those examples and how ability checks are resolved, then you're never going to understand why the designers did what they did, and you're probably better off with a third-party skill system if not a different game.

[Skrum]

(Here we go again indeed...)

If you truly see no difference between those examples and how ability checks are resolved, then you're never going to understand why the designers did what they did, and you're probably better off with a third-party skill system if not a different game.

See this I don't understand - the claim on the table seems to be
1) ability checks are free-form and essentially rule-less by design
2) the choice to make ability checks that way is intrinsic to 5e
3) changing that or giving any more structure to ability checks will ruin 5e
4) ergo one is better off playing a different game

Despite, as I've pointed out a ton of times, the relative tight ship that the combat rules are (and character creation for that matter). Turns are neatly defined, attack rolls, spells, movement on a grid, hit points, AC, everything; like it's right there. Plenty of room for creativity, tons of tactical play, but rules are neatly defined and everyone knows (well, anyone who wants to know) exactly how what their options are, how things interact, the whole nine yards.

But bringing even a fraction of this rigor to ability checks, and the game just loses something essential about it. Totally collapses like a bad souffle. IMO, if ability checks really are supposed to be the "make it up, live your best life" part of the game, why do they even have numbers? Just have proficiencies and the Rule of Cool.

Like I know we decided to agree to disagree about ability check rigor upthread, but I'm just mystified about this position. I really don't get it lol.

[Psyren]

Like I know we decided to agree to disagree about ability check rigor upthread, but I'm just mystified about this position. I really don't get it lol.

Plenty of your positions mystify me too if that helps

3) changing that or giving any more structure to ability checks will ruin 5e

They give a ton of structure to ability checks actually, it's called modules. There's printed DCs all over the place in those. Your DM (or you if that's you) could probably reverse-engineer some kind of DC table by diving through all of them.

[JNAProductions]

Plenty of your positions mystify me too if that helps

They give a ton of structure to ability checks actually, it's called modules. There's printed DCs all over the place in those. Your DM (or you if that's you) could probably reverse-engineer some kind of DC table by diving through all of them.

So you're not even opposed to having DC examples, you just don't want them organized?

[Psyren]

So you're not even opposed to having DC examples, you just don't want them organized?

I want them tailored for specific scenarios/encounters, which is what modules are. If you or someone else then wants to use those to extrapolate some kind of general case, have at it.

[Skrum]

I want them tailored for specific scenarios/encounters.

I mean, this would still exist. In spades. It's impossible to have a literal exhaustive list of ability checks. But
1) some amount of guidance, like tasks that are approx. this hard should be DC X, tasks that are approx. that hard should be DC Y
2) some context and guidance for the skills themselves. What can they do. Would shouldn't they do. That sort of thing (can a skill check beat a spell?)
3) and personally, I think some kind of "breaker" system, even a simple one, would be nice. Given that 5e has done away with massive modifiers, it leaves a lot to the luck of the die. I find this pretty unsatisfying, and would like there to be more limits on what can be accomplished with dumb luck.

[Psyren]

I mean, this would still exist. In spades.

I know, it exists now Hence my suggestion.