Flipping Proficiency Dice on its head

[Theodoxus]

So, the DMG has a variant rule for Proficiency, replacing the +2 to +6 range with d4 to d12 instead. This is to bring "more randomness to the game". But I've never played or run a game where anyone wanted MORE randomness. In fact, one of the biggest complaints I've heard online is in regard to how someone proficient with a skill, even at top tier, is only slightly better than random chance on a d20.

What I'd propose, specifically for skills, not for combat (although the general idea wouldn't be horrible, but it does wreck bounded accuracy)... is to turn the dice on their head.

An untrained skill check would still use a d20 roll. But trained skllls would swap out that "d20+2" for a "d12 +8". So, a 1st level character proficient in say, Perception, now rolls a d12, adds 8 and their Wisdom modifier, say, +2. Passive would be defined as half the die, plus bonuses, so 6+8+2, for 16.

+3 Proficiency becomes d10+10; +4 becomes d8+12; +5 becomes d6+14 and +6 becomes d4+16. You're still maxing out at 25 on the roll if you roll max with +5 in the pertinent attribute, but the chances of you succeeding ramp up far and away over the statistical average of an untrained person.

As with the DMG option, Expertise grants advantage on the roll.

[EggKookoo]

I like the concept. Might be too complex for D&D and d20, though.

There's something interesting about a system where, as you gain experience, it's not the ceiling that increases but the floor. A lowbie and a high level char can both succeed to the same degree, it's just that the high level char is less likely to fail.

[JNAProductions]

Eh... There's something to be said for reducing randomness, but I feel it's better achieved by using bell-curved rolls. 2d10 or 3d6, for example.

It's an interesting idea, but not one I think works well with 5E.

[Greywander]

Eh... There's something to be said for reducing randomness, but I feel it's better achieved by using bell-curved rolls. 2d10 or 3d6, for example.

It's an interesting idea, but not one I think works well with 5E.
Yeah, on its head 2d10 or 3d6 might seem like good alternatives to 1d20, but the thing is that 5e is designed with it being just as likely to roll a 1 or 20 as it is a 10. Going to a bell curve means average rolls will be much more common, while extreme rolls on either end become a lot less common. So now that monster with really high AC, instead of having a 5% chance of hitting them, you now have a less than 0.5% chance of hitting. That's a >90% drop in the number of hits.

If you switched to a bell curve, you'd want to tighten up bounded accuracy so that both the highs and the lows fall closer to the average. Since extreme results are much less likely to be rolled, you could tighten things up and still have it be less likely to make those rolls. When you have a bell curve, +1 becomes an even bigger deal than it is for a flat curve.

One idea that comes to mind is to change ability score modifiers to bonus dice rather than a flat bonus. So, say 3d6 is the base roll, but if you have an ability score of 12 then you'd roll an extra d6 and only take the highest three, whereas if you had an ability score of 8 or 9 you would roll an extra d6 and take the lowest three. So like advantage and disadvantage. This way, stats aren't just a static +1 that radically shifts your average roll. For derived stats, such as AC, you might also change it from a flat bonus to a rolled bonus (or penalty, for stats less than 10).

It's still probably a bad idea, or rather, it would just be a lot of work to change the rules so that the math works out properly, but there are some interesting ways this could be handled.

As for the concept in the OP, it's an interesting one. The main drawback that I see is that you aren't able to roll as high. 1d20 + 11 (prof. of +6 and ability mod of +5) can get you as high as 31, or as low as 12. 1d4 + 16 + 5 only gets as high as 25, but only as low as 22. It becomes impossible to fail a DC 20 check, or to pass a DC higher than 25.

Something that would work similarly while keeping the math mostly the same would be if rolling with proficiency were 1d12 + 10 + ability mod, with that 10 increasing up to 14 as your proficiency bonus increases. This raises the floor considerably while keeping the ceiling in the same place. It might be more balanced start with 1d12 + 8, increasing to 1d12 + 12, which ends up with a ceiling that is 2 points lower but a floor that is still much higher.

I'm comparing the probability curves for 1d20 + 6 vs. 1d20 + 1d12, and I honestly can't figure out if proficiency dice or more or less swingy. Average rolls are still 5% each, so they're equally likely either way. Extreme results get interesting, though. Technically, extreme results are less likely to be rolled when using proficiency dice, however, the extremes are also much more extreme. With 1d20 + 6 you can't roll lower than 7 or higher than 26, whereas with 1d20 + 1d12 you can roll as low as 2 and as high as 32. Your odds of getting a specific extreme result is much lower, but your odds of getting a result 7 or less or 26 or more are actually higher overall. The problem is that it is, indeed, creating a bell curve where extreme results are less likely, but at the same time it's also widening the number of possible results, which ends up with extreme results actually being more common.

[Chronos]

(although the general idea wouldn't be horrible, but it does wreck bounded accuracy)
That's a feature, not a bug. Bounded accuracy was a terrible idea to begin with, and is the worst thing about 5e as a system.

[stoutstien]

That's a feature, not a bug. Bounded accuracy was a terrible idea to begin with, and is the worst thing about 5e as a system.

May I ask why you think this? the execution could have been better but the concept about it accurate is solid in my opinion. the idea proficiency dice something that I played with but never implemented as of yet. I was just going to replace every point of proficiency as a 1D4 I just haven't gone through and applied it to the NPC side.

[NaughtyTiger]

That's a feature, not a bug. Bounded accuracy was a terrible idea to begin with, and is the worst thing about 5e as a system.

I don't remember this issue. what don't you like about bounded accuracy?

[Belzique]

So now that monster with really high AC, instead of having a 5% chance of hitting them, you now have a less than 0.5% chance of hitting. That's a >90% drop in the number of hits.

It doesn't have to be all or nothing. I intend to use 3d6 for checks only the next time I DM while attack rolls and saving throws remain 1d20. That way the bounded accuracy of combat remains unaffected.

[Segev]

I have found that there is some misunderstanding of what, exactly, bounded accuracy is. I was, myself, disappointed when I learned that it's not actually a mechanical constraint in the system. It is, instead, a design constraint that is very easily broken if you don't know what you're doing. The idea isn't that the system inherently bounds accuracy, but that they were very, very careful to keep the AC/DC structure from going too terribly high (AC 16 is pretty high, contrary to my gut feelings from 3e; AC 19 or 20 is really high), and conversely keeping bonuses to less than +10. If you have a +6, you're pretty high. A +8 is VERY high.

By this rigorous attention to bonuses and target numbers, they "bound accuracy" so that it's never impossible for even somebody with level 1 proficiency (+2) and a 10 in a stat (-0) to hit on more than a natural 20, but also that even a high-level character is likely to miss a DC 13 (Mage Armor) with their likely +10 (+4 stat, +6 proficiency) on other than a natural 1.

In 3e, the hedge that a 20 always hits and a 1 always misses was often the only saving grace of a low level vs a high level. Bonuses vs. target numbers could easily have more than a 20 shift between them. 5e, it's very hard to get to the point that the die isn't a major component of the result.

In addition, ACs aren't actually scaling with level nearly as much as proficiency is, and even proficiency scales only moderately.

But...again. These are not hard-coded too deeply in the system. It's mostly in careful attention to what the highest allowable bonuses and highest alloweable DCs are. The biggest breakers of this are in Ability checks with skill proficiencies, thanks to Expertise. And there, it's "acceptable" because it's meant to make skill characters able to "just succeed" more often. But even then, until you hit a +4 proficiency bonus, Expertise is not scaling outside of bounded accuracy's normal guidelines.

All bounded accuracy is is a design consideration to ensure that the randomness of the dice will always matter significantly in determining success or failure. It is also, from the DM's side of the proverbial screen, a reason why monsters have such apparently low to-hit values: they're part of what makes 5e so very survivable for PCs.

[Zhorn]

It sounds interesting, but I'm not too sure about actually playing by it. Does this eliminate concepts like 'nat 1 is always a fail'? Also, items and features that are built around getting reliable 10's on a d20 roll (example: Reliable Talent) will either need to be reworked, or excluded from play.

[Arkhios]

That's a feature, not a bug. Bounded accuracy was a terrible idea to begin with, and is the worst thing about 5e as a system.

So, when PC's gain more levels, you'd prefer to progressively make more and more of monsters useless, and progressively reduce the options you could use to create encounters, and as such, require either new monster manuals one after another until they number at Monster Manual 10 or something, or to homebrew lots and lots of new creatures once the players have progressed beyond the threshold that the
official books provide. Are you sure you're ready, willing, and capable of paying 50 bucks per new book every time you need more material, or to be able to satisfy the need all on your own?

You see, bounded accuracy has the benefit of keeping existing options reusable. Not even the smallest CR creature is meaningless. Go creative on how the encounters are built. Pitting PC's against hordes of goblins on flat plains is quickly handled by high level characters and quite boring. But put that horde into a canyon and give them some tactical advantage and suddenly a single fireball does very little while the goblins pepper even high level characters with arrows.

[Pex]

So, when PC's gain more levels, you'd prefer to progressively make more and more of monsters useless, and progressively reduce the options you could use to create encounters, and as such, require either new monster manuals one after another until they number at Monster Manual 10 or something, or to homebrew lots and lots of new creatures once the players have progressed beyond the threshold that the
official books provide. Are you sure you're ready, willing, and capable of paying 50 bucks per new book every time you need more material, or to be able to satisfy the need all on your own?

You see, bounded accuracy has the benefit of keeping existing options reusable. Not even the smallest CR creature is meaningless. Go creative on how the encounters are built. Pitting PC's against hordes of goblins on flat plains is quickly handled by high level characters and quite boring. But put that horde into a canyon and give them some tactical advantage and suddenly a single fireball does very little while the goblins pepper even high level characters with arrows.

Some people actually do want that. They do want goblins to become irrelevant to high level characters. They don't want to fight goblins at level 1 and 5 and 7 and 9 and 11 and 14 and 16 and 20. They don't need a continuous series of new Monster Manuals. One Monster Manual can have enough high level threatening monsters on its own. A new Manual is only exciting because it's new. Even people who like fighting goblins at level 1 and 5 and 7, etc., can be excited about a new Monster Manual to fight new monsters at all levels. New Monster Manuals are irrelevant on the topic of Bounded Accuracy.

[Knaight]

So, when PC's gain more levels, you'd prefer to progressively make more and more of monsters useless, and progressively reduce the options you could use to create encounters, and as such, require either new monster manuals one after another until they number at Monster Manual 10 or something, or to homebrew lots and lots of new creatures once the players have progressed beyond the threshold that the
official books provide. Are you sure you're ready, willing, and capable of paying 50 bucks per new book every time you need more material, or to be able to satisfy the need all on your own?

Bounded Accuracy is new to 5e. Meanwhile 4 previous editions* managed to have reasonably complete monster manuals for all 10, 20, 30 or 36 levels of play. More got added, sure, but they weren't by any means required. Meanwhile, outside the D&D bubble plenty of games get by just fine without 300 pages of stats for the opposition, especially when the opposition doesn't have to be cycled through. This hypothetical just isn't born out in practice.

Similarly, bounded accuracy is just one way to pair fairly extensive character power differences with maintaining usability of a large set of opposition across the power differential. There are plenty of others, and while it's an elegant system given the design parameters of D&D (attacks vs. a static AC, ascending HP and damage by level) it's far from the only available system there.

It's also worth observing that the thread started aimed at skills, and not combat - that's not a coincidence. Bounded accuracy works great in the combat system, which is full of things to mitigate the swinginess (the nature of extended tasks, extra attack, damage spells that do half on a miss, etc.). Skills don't have that, and it's skills where things tend to feel off. An extra attack analog would really have helped here, as would altering Proficiency a bit for skill checks.

*Technically more like 9, but that's getting into the weeds.

[Theodoxus]

Yeah, I probably shouldn't even have mentioned combat at all in my opening... that was my bad - I was just trying to stave off the conversation that ended up happening anyway.

@Greywander, thanks for the review. I tend to get excited about something and overlook the little things. You're correct, by top tier, skill checks end up being nigh meaningless under this..

For full disclosure, I used the 2d10 method for an entire campaign, liking the bell curve aspect, but it really just made me have to flatten out DCs more than anything. I just started reading through Dungeon World, and their static DC of 10+ on 2d6+mod is intriguing as well... But I fear I'm moving further away from D&D in general looking at other systems. My players have decent elasticity when it comes to experimenting with some things, but I had to start a new 5E Core rules game because they whined about not playing "D&D" anymore :smallwink:

[Yunru]

Just swap all d20 rolls for a different die size (I've found a d16+2 has the best impact, any smaller and bounded accuracy starts to break too much, any larger and it doesn't address the issue).
That way proficiency matters more, and luck matters less.

[Segev]

Skills are a different beast than combat, yes. That’s why expertise exists for them, but not for attack rolls or saves.

If you want more frequent high success, either just set the DCs lower, or make a feat that gives Expertise on one or two skills.

[KorvinStarmast]

r the statistical average of an untrained person.
That isn't what this game is about.

Just a note on the DMG method: works fine. We had a DM run it for about six months. It didn't have a big influence on game play, though now and again a spectacular success happened. One example was when my sailor-background-proficient-in-Athletics Sorcerer (pirate, eh?) shoved a hostile guard over a railing ... by succeeding on an contested athletics check by one thanks to that d4 I rolled for proficiency. It could have gone badly. The uncertainty of the outcome is built into this game's DNA from as far back as you can go.

The d20 system is already, and by intention, swingy.
This "crunch numbers and be in control of all outcomes" approach is very much a computer game min max mind set. It also is related to different dice systems in different games. If you want that kind of game experience, I suggest you find a different game system.

If you read the intro to the game in the PHB, Mike Mearls spells it out: the dice are fickle, and can be cruel.

That's a feature, not a bug.

Eh... There's something to be said for reducing randomness, but I feel it's better achieved by using bell-curved rolls. 2d10 or 3d6, for example.

It's an interesting idea, but not one I think works well with 5E.
Yes, I had a few years ago discussed using a 2d10 method instead and was shown how that skews the results considerably.

[Belzique]

Finally got a chance to sit down and think about your original post OP.


But trained skills would swap out that "d20+2" for a "d12 +8"... +3 Proficiency becomes d10+10; +4 becomes d8+12; +5 becomes d6+14 and +6 becomes d4+16. You're still maxing out at 25 on the roll if you roll max with +5 in the pertinent attribute, but the chances of you succeeding ramp up far and away over the statistical average of an untrained person.

It's an interesting idea but it raises the average of the rolls.
1d12 + 8 = 14.5 vs 12.5 of 1d20 + 2
1d10 + 10 = 15.5 vs 13.5 of 1d20 + 3
1d8 + 12 = 16.5 vs 14.5 of 1d20 + 4
etc
This also makes the floor very high. Imagine a 0 CR commoner NPC that works at the docks so you decide to give him proficiency in athletics (or whatever, just an example). While before he did 10.5 on average, having a 40% chance of getting a roll of 8 or lower, he now cannot get a roll lower than 9.

Like I said earlier in the thread I'm interested in making 3d6 the standard for my next campaign. It'll have the same average while only increasing the floor by 2, which I believe might make sense because it makes only creatures inherently bad at things get a 1. The max roll of 18 will also help with things such as manacle DC, which is a DC 20 check to escape from according to the PHB, and I have a hard time believing a normal commoner can escape those 5% of the time.

[Pex]

Skills are a different beast than combat, yes. That’s why expertise exists for them, but not for attack rolls or saves.

If you want more frequent high success, either just set the DCs lower, or make a feat that gives Expertise on one or two skills.

Or have example DC tables to help DMs figure out task difficulty and see for themselves all sorts of things that are autosuccess or DC 5 or whatever to mitigate swinginess.

(quickly runs away) :smallbiggrin:

[Segev]

Or have example DC tables to help DMs figure out task difficulty and see for themselves all sorts of things that are autosuccess or DC 5 or whatever to mitigate swinginess.

(quickly runs away) :smallbiggrin:

Now, now, let's not ask for the apparently-impossible. Closing the loop might make it a fully functional subsystem!

[EggKookoo]

Or have example DC tables to help DMs figure out task difficulty and see for themselves all sorts of things that are autosuccess or DC 5 or whatever to mitigate swinginess.

(quickly runs away) :smallbiggrin:

https://media1.giphy.com/media/NAnrX13MH49cA/source.gif

[KorvinStarmast]

Or have example DC tables to help DMs figure out task difficulty and see for themselves all sorts of things that are autosuccess or DC 5 or whatever to mitigate swinginess.

(quickly runs away) :smallbiggrin:psst. There are some examples in Xanathar's Guide
:smallbiggrin:

[Unavenger]

psst. There are some examples in Xanathar's Guide
:smallbiggrin:

I mean, you really shouldn't need an expansion pack before you have a complete game. It's like shipping not-day-1 DLC which you need to finish the main storyline.

[Yunru]

A commoner has a 50% chance to pass a DC 11 skill check. A trained commoner has a 50% chance to pass a DC 13 skill check.

Such basic maths shouldn't need to be in a book.

[Unavenger]

A commoner has a 50% chance to pass a DC 11 skill check. A trained commoner has a 50% chance to pass a DC 13 skill check.

Such basic maths shouldn't need to be in a book.

That would be great, if I knew what kinds of arcane knowledge a fantasy commoner had a 50/50 chance to know.

[Yunru]

That would be great, if I knew what kinds of arcane knowledge a fantasy commoner had a 50/50 chance to know.You're the DM, you know exactly what kinds of arcane knowledge a fantasy commoner has a 50/50 chance to know. No book can tell you that, it's your world. (Unless you run a module in which case they should tell you in said module.)

[KorvinStarmast]

I mean, you really shouldn't need an expansion pack before you have a complete game. It's like shipping not-day-1 DLC which you need to finish the main storyline. Without derailing this thread into the usual harangue that people get into about the skill system, I take the position that what Pex was asking for isn't necessary. But that's one person's opinion, there are others.

You do touch upon an important point. They publish this game to make a profit. :smallbiggrin:

You're the DM, you know exactly what kinds of arcane knowledge a fantasy commoner has a 50/50 chance to know. No book can tell you that, it's your world. (Unless you run a module in which case they should tell you in said module.) One of our posters who has abandoned this forum did some digging into the first few published modules and what he found was kind of interesting.
Most DC's were 10, 12, 15, 17, 20 or 22.
See a pattern there? :smallcool:

[Segev]

One of our posters who has abandoned this forum did some digging into the first few published modules and what he found was kind of interesting.
Most DC's were 10, 12, 15, 17, 20 or 22.
See a pattern there? :smallcool:

Boy, wouldn't it be nice if the pattern were spelled out as a guideline in the DMG or PHB?

But yeah, this isn't about the flaw in the skill system. This is about whether raising the floor by changing the die type and making the bonus bigger is a good idea.

I think...ultimately, no. Not for 5e D&D. Advantage/Disadvantage is the mechanic chosen, here, for not modifying the range while modifying the probability. Proficiency (and stat bonuses, and expertise, etc.) are actually meant to let you push higher. The stingy distribution of bonuses is how they achieve their bounded accuracy, keeping everything pretty well with 30 as the highest possible roll. (I'm sure you could design a theoretical build that can hit higher than 30 on a roll, but it'd be really hard.)

[EggKookoo]

I'm still intrigued by a mechanic that is mostly about raising the floor as you gain experience, rather than raise the ceiling. Are there any systems out there that work that way?

[Unavenger]

I'm still intrigued by a mechanic that is mostly about raising the floor as you gain experience, rather than raise the ceiling. Are there any systems out there that work that way?

I believe... Window, is it called? Works this way - it's a roll-under system, and your die gets smaller the better you are at something. So if you have 1d4 in something, you always ace a DC 4 check, and if you have d8 in something, you're 50/50 to do that same check. It's functionally identical to 1d4+4 vs 1d8 to make a DC 5 check in a roll-over system, which is the kind of thing ultimately being proposed here.

[Segev]

I'm still intrigued by a mechanic that is mostly about raising the floor as you gain experience, rather than raise the ceiling. Are there any systems out there that work that way?

Kind-of. Paladium's skills are all percentile based, and you just have a flat percentage chance to succeed on them (which improves with level).

Technically, raising the bonus DOES raise the floor as well as the ceiling. What this proposed system does, by reducing the variance and raising the floor but NOT the ceiling, is shrink the region of values where it's possible that the roll matters. With a bonus, the way 5e currently does it, there are still 20 DC values where the result of the die roll matters. They're just higher values than they were before the bonus was increased. Things which were previously impossible become possible, at the same time that things which were previously possible to fail become impossible to fail.

With this, things go from possible to fail to impossible to fail, but nothing new becomes possible.

But, more interestingly, the better your proficiency, the HIGHER the variance on whether you succeed or fail when the DC is in the range of values that success or failure are not automatic.

If the DC is 15, and you're rolling a d8+12, you succeed on a 3+ on the die. That's 6/8 chance to succeed, or 75%. To match that chance to succeed on a d20+bonus roll, you'd need to succeed on a 6+, meaning you'd need a +9 bonus.

Now, let's look at a DC 20 with these same rolls. d8+12 has a 1/8, or 12.5%, chance to succeed. d20+9 succeeds on an 11+, or a 50% chance to succeed.

On the other hand, a DC 14 has the d8+12 succeeding 87.5% of the time (i.e. on a 2+), while the d20+9 only succeeds on a 5+, for an 80% chance. Going down by 1 has changed the odds by 12.5% for the d8+12, and by only 5% for the d20+9.

Going down by 2 makes the d8+12 a sure thing, but only an 85% chance for the d20+9.

Meanwhile, going to DC 21 makes it impossible for the d8+12, while the d20+9 succeeds still on a 12+, for 45% chance! And a DC 25 is still doable for the d20+9 (25% chance), but remains impossible for the d8+12.

[stoutstien]

I'm still intrigued by a mechanic that is mostly about raising the floor as you gain experience, rather than raise the ceiling. Are there any systems out there that work that way?

Have stat modifiers added to rolls but proficiency bonuses reduce the DC? That way you have a very small but meaningful increase to minimum rolls up to auto pass DCs of 5 and proficiency reduces the chance of failure without compounding DC growths

[Amechra]

Have stat modifiers added to rolls but proficiency bonuses reduce the DC? That way you have a very small but meaningful increase to minimum rolls up to auto pass DCs of 5 and proficiency reduces the chance of failure without compounding DC growths

That's legitimately the same thing as we have now - roll are effectively d20 + Stat + Proficiency - DC vs. 0. You just changed it to d20 + Stat - (DC - Proficiency) vs. 0.

Something like "the minimum number you can roll on your d20 is equal to your Proficiency bonus" might work - Expertise would still add to the roll itself, and Reliable Talent would just add your Proficiency to the minimum roll of all of your ability checks.

[EggKookoo]

Something like "the minimum number you can roll on your d20 is equal to your Proficiency bonus" might work - Expertise would still add to the roll itself, and Reliable Talent would just add your Proficiency to the minimum roll of all of your ability checks.

I was thinking something like that too. Plus your ability mods. So, +2 ability, +2 prof = treat any roll less than 4 as a 4. You don't modify the roll, and DCs remain the same.

Talk about bounded!

[stoutstien]

That's legitimately the same thing as we have now - roll are effectively d20 + Stat + Proficiency - DC vs. 0. You just changed it to d20 + Stat - (DC - Proficiency) vs. 0.

Something like "the minimum number you can roll on your d20 is equal to your Proficiency bonus" might work - Expertise would still add to the roll itself, and Reliable Talent would just add your Proficiency to the minimum roll of all of your ability checks.

Not quite. Because the Prof lowers the DC vs adding it to the roll it keeps DC relevant without bloating totals. What is important is the range of results.
The minimum/maximum rolls stay within a curve compared to as presented where proficiency shifts the results 10-40% higher for both values.
It also prevents from proficiency overshadowing stat modifiers because one moves the floor and the other moves the ceiling.

[JNAProductions]

Not quite. Because the Prof lowers the DC vs adding it to the roll it keeps DC relevant without bloating totals. What is important is the range of results.
The minimum/maximum rolls stay within a curve compared to as presented where proficiency shifts the results 10-40% higher for both values.
It also prevents from proficiency overshadowing stat modifiers because one moves the floor and the other moves the ceiling.

No, it's the same thing from a mathy point of view.

What's the difference between 1d20+5 versus DC 15 or 1d20+10 versus DC 20?

[Pex]

While the goal may be worthy, adding more dice rolls and changing the die based on circumstances adds too much complexity. 5E is not completely simple but outcome resolution of doing things is purposely relatively so. 5E is not a simulation.

What needs to be learned is acceptance of the idea of things not needing a roll. The 20 ST barbarian will never lose an arm wrestling contest against the 8 ST warlock. Pick a variance. Someone with minimum 16 ST gets to roll against the barbarian. Less than that barbarian wins. For any two arm wrestlers, as long as their ST scores are within 4 of each other - roll. Otherwise, higher wins. I just made that up. It's what you need to do as a 5E DM.

[stoutstien]

No, it's the same thing from a mathy point of view.

What's the difference between 1d20+5 versus DC 15 or 1d20+10 versus DC 20?

Because the max roll is now fixed. By having a firm roll max of 25 with max stats (barring 20 + modifier with the Barbarian capstone and so on) the DCs never need to go past 25. From the players view nothing changes which was intentional so you are correct that the outcomes wouldn't change other than values.

I just wanted to add the additional effect of making stats and proficiency to have different impacts on the rolls.

It was just a spitball idea to add a ceiling.

[JNAProductions]

Because the max roll is now fixed. By having a firm roll max of 25 with max stats (barring 20 + modifier with the Barbarian capstone and so on) the DCs never need to go past 25. From the players view nothing changes which was intentional so you are correct that the outcomes wouldn't change other than values.

I just wanted to add the additional effect of making stats and proficiency to have different impacts on the rolls.

It was just a spitball idea to add a ceiling.

But it doesn't actually change anything. The base DC remains the same, modified now by Proficiency bonus.

X+Y+Z=DC is the same as X+Y=DC-Z.

[Theodoxus]

If the DC is 15, and you're rolling a d8+12, you succeed on a 3+ on the die. That's 6/8 chance to succeed, or 75%. To match that chance to succeed on a d20+bonus roll, you'd need to succeed on a 6+, meaning you'd need a +9 bonus.

Now, let's look at a DC 20 with these same rolls. d8+12 has a 1/8, or 12.5%, chance to succeed. d20+9 succeeds on an 11+, or a 50% chance to succeed.

On the other hand, a DC 14 has the d8+12 succeeding 87.5% of the time (i.e. on a 2+), while the d20+9 only succeeds on a 5+, for an 80% chance. Going down by 1 has changed the odds by 12.5% for the d8+12, and by only 5% for the d20+9.

Going down by 2 makes the d8+12 a sure thing, but only an 85% chance for the d20+9.

Meanwhile, going to DC 21 makes it impossible for the d8+12, while the d20+9 succeeds still on a 12+, for 45% chance! And a DC 25 is still doable for the d20+9 (25% chance), but remains impossible for the d8+12.

Well, this isn't a fair comparison. The d8+12 is your static proficiency bonus. You'd still add your attribute modifier, just like you're doing for the d20 roll example. Without magic, the highest DC you could obtain is 25, so you'd have a chance (provided a 20 in the relevant stat...) But otherwise the overall point is true.


While the goal may be worthy, adding more dice rolls and changing the die based on circumstances adds too much complexity. 5E is not completely simple but outcome resolution of doing things is purposely relatively so. 5E is not a simulation.

What needs to be learned is acceptance of the idea of things not needing a roll. The 20 ST barbarian will never lose an arm wrestling contest against the 8 ST warlock. Pick a variance. Someone with minimum 16 ST gets to roll against the barbarian. Less than that barbarian wins. For any two arm wrestlers, as long as their ST scores are within 4 of each other - roll. Otherwise, higher wins. I just made that up. It's what you need to do as a 5E DM.

And this is true too. One thing I keep meaning to ask my players is to stop trying to roll dice all the time. Especially as some kind of weird competition. I'll have some NPC monologue and one will inevitably shout "I roll Insight!" and then all the rest are there rolling too... I'm thinking of granting XP bonuses to RPing said checks, and only for the one rolling and potentially anyone assisting. Something like "we put our heads together and compare notes on whether we think Mr. Shopkeeper is being honest with us." XP really is the only carrot I have that my players respect...

[EggKookoo]

And this is true too. One thing I keep meaning to ask my players is to stop trying to roll dice all the time. Especially as some kind of weird competition. I'll have some NPC monologue and one will inevitably shout "I roll Insight!" and then all the rest are there rolling too...

I've come to rely on mockery.

"I make a roll!"

"Ok, blueberry? Cinnamon? How does it taste?"

Or "I roll for perception."

"Ok, everyone looks at you funny while you roll around on the floor."

I try to drill into my players' heads that they take actions, not make rolls. I ask for rolls only if needed.

[stoutstien]

But it doesn't actually change anything. The base DC remains the same, modified now by Proficiency bonus.

X+Y+Z=DC is the same as X+Y=DC-Z.

For the DM it does because now they have a clear cut off for DCs. Having proficiency does increase the total into the 30s like it can with expertise which causes the case of of setting player specific DCs which feels off IMO.

[Yunru]

For the DM it does because now they have a clear cut off for DCs. Having proficiency does increase the total into the 30s like it can with expertise which causes the case of of setting player specific DCs which feels off IMO.

Except it shouldn't.
There's no point in having varying abilities if you're just going to adjust the difficulty for them.

A rogue with expertise and maxed out ability succeeds on a DC 25 65% of the time? Good for them, that's what they invested in.

Inventing a higher difficulty just makes that investment pointless.

[Amechra]

For the DM it does because now they have a clear cut off for DCs. Having proficiency does increase the total into the 30s like it can with expertise which causes the case of of setting player specific DCs which feels off IMO.

It also makes it a pain to manage, because instead of the player keeping track of one number (stat + proficiency) and the DM keeping track of another (the DC for [blank]), you have the player keeping track of one number and the DM keeping track of at least one number per PC. Congrats, you just increased the DM's workload.

If your rolls and difficulties are all out in the open, it's still worse, because subtraction is harder than addition.

[Theodoxus]

Except it shouldn't.
There's no point in having varying abilities if you're just going to adjust the difficulty for them.

A rogue with expertise and maxed out ability succeeds on a DC 25 65% of the time? Good for them, that's what they invested in.

Inventing a higher difficulty just makes that investment pointless.

This was my (and apparently a good number of players, based on old threads) problem with Pathfinder DCs. At 1st level, it was a DC 10 to climb a tree. At 15th level it was a DC 30 to climb the same tree. Because instead of getting better at climbing trees, the system just tailored the DCs to the same chance of success.

It was why Bounded Accuracy was hailed so highly at first - finally, DCs that don't scale with level but are in theory static regardless... Of course, as we see here, such a thing brings about it's own issues...

[Yunru]

It was why Bounded Accuracy was hailed so highly at first - finally, DCs that don't scale with level but are in theory static regardless... Of course, as we see here, such a thing brings about it's own issues...

IMO the OP's approach is the correct one, but also not.
Reducing the die size and adding a static modifier fixes the randomness issue, however it should be uniform and unchanging.

For instance, at my table all d20 rolls have been replaced by d16+2. (Original it was d10+5, but we found that reduced randomness too much.)

[stoutstien]

It also makes it a pain to manage, because instead of the player keeping track of one number (stat + proficiency) and the DM keeping track of another (the DC for [blank]), you have the player keeping track of one number and the DM keeping track of at least one number per PC. Congrats, you just increased the DM's workload.

If your rolls and difficulties are all out in the open, it's still worse, because subtraction is harder than addition.

Like a said it was just spitballing. First ideas are rarely perfect. It's a tough job to get proficiency bonus to work with bounded accuracy. Stacking bonus and static values and all that.