As with any version of D&D I've ever played there are always rules and systems that I really like and others which have made me question the designer's decision to include them. But there is one specific rule in 5th Edition that really baffles me. It's not that this is the one element in the game that particularly dislike most, but it's more like I just can't grasp the reason why they would do it.
Double proficiency bonus!
Why? Beacuse it seems like this one rules goes directly against the whole idea of Bounded Accuracy which they supposedly based the entire edition on.
I thought the whole idea of Bounded Accuracy was to avoid the enourmous gaps between character rolls we had in the past, especially at high levels. Such as we had, for example, in 3e where a high level Rogue would check a +20 Hide (or more) against maybe a +0 or +2 of the also-high-level character but who is untrained in Spot. And other big disparities of the sort.
And then, voilà, they introduce double-proficiency bonuses, which seems to me just scraps the concept of Bounded Accuracy and at high levels causes the exact same disparities in checks. We're looking at +17 checks (5 from an ability score and 12 from double proficiency) often against +0 of the untrained character or maybe a little more if he has some bonus from an ability.
I know this is all in the "white room theory-crafting" realm, as I haven't played 5e at high-levels yet, so if anyone has done so and somehow found that this is not true feel free to disagree and explain why so. But the idea of double-proficiency to me just feels like it goes directly against the idea of Bounded Accuracy. At this point I'm seriously inclined to change any double-proficiency to some other bonus, maybe Advantage, rerolling, or something of the sort.
+17 is still within range of a d20.
It's the very edge of the bounds.
Expertise is hard-wired into the "bounds" of 5E's "bounded accuracy".
Actually, because rogues can't roll less than a 10 on a trained skill, their minimum roll is 27. This is outside the bounds as even a monster with a 10 wisdom and training in perception can never spot the rogue (max roll of 26).
But, most high level monsters have blindsight or other supernatural senses that make stealth mostly useless anyway.
Being "possible" to check against the +17 roll doesn't mean it's reasonable to expect a check like that to succeed.
Someone with the average check of +0 has only 0.75% chance of succeeding against a +17 roll.
Even in past editions if you used the "20 always succeeds" rule everyone still had a minimal chance to win a roll even if against a +30 or whatever. But it isn't practical to expect that at all.
I thought the idea this time with Bounded Accuracy was precisely to avoid those huge gaps in numbers.
+11 against +0 is a more than reasonable margin to ensure the specialized character will win most of the time, but the average guy won't just "not even bother" and still have some reasonable-even-if-small chance to overcome it.
Now doubling that proficiency bonus? I'm not liking it. The very ceiling of +6 seems to have been well thought of to avoid such disparities... only to be scrapped later with doubling abilities.
As I said actual play in high-level might prove this not a big deal, but I'm doubting it. And since bounded accuracy is "the one thing" that 5e is built upon I'd rather stick to its supposed standards and change those double bonuses to some other sort of bonus.
Rastapopoulos wrote:
Saelorn wrote:
Saelorn wrote:
Saelorn wrote:
Gnarl wrote:
I found it better than Expertise in the playtest, which was a flat +5 at level 1 
Whether automatic successes and misses should be used or not is pointless to this discussion, really.
My point is Bounded Accuracy was supposedly made to lessen the gap between the specialized and the average character, so that even if the specialized is most likely to succeed most of the time, it is not an unsurmountable task for the average one to try and beat that guy sometimes.
Having 0.75% chance of doing that doesn't feel at all like it falls into this propposal. For practical purposes you will always fail. Even if you have a slighlty better modifier and not +0, you will for all practical purposes almost all fail except in very very VERY impossibly rare occasions, to a point that it's completely useless to count on that and if once in a character's lifetime this happens by chance it matters very little to the overall game.
As I understand it, Bounded Accuracy was made to lessen the numeric gap for the unspecialized characters to still have a reasonable chance of succeeding. Not for them to have an infinitesimal chance of ever hoping to succeed.
Novacat wrote:
Rastapopoulos wrote:
Rastapopoulos wrote:
I think that without it they noticed any high dex PC was more or less just as good as the rogue at doing rogue type things. The Rogue is the only one who can use expertise on thieves kits for example. I would be more inclined to stick ot on perception and stealth but YMMV.
I don't think that the intent of BA was to keep specialists and the non-proficienct within the same bounds. I think the goal was to keep the proficient and the non-proficient in the same bounds, and to keep the proficient and the specialized in the same bounds. I don't have a problem with a specialist trouncing someone who is entirely non-proficient.
Shiroiken wrote:
I think you are all forgetting that bounded accuracy was not about limiting the bonuses, but constraining the DC's.
Rastapopoulos wrote:
Zardnaar wrote:
Wuzzard wrote:
A don't think double profeciency is the problem. It's scaling profeciency itself that's the problem.
A level 1 character has +2 prof bonus and fairly low stats and a limited amount of abilities.
a level 10 character has +4 prof bonus, perhaps better stats or if nor very expanded amount of abilities
a level 20 character has +6 prof bonus, better stats and tons of nasty abilities.
Now, is it really nesceserry for the prof bonus to increase at all for the character to feel growth? Not in my opinion, and I think it's THE one thing that is stopping this great idea about a very narrow band of bonuses that was so fresh in the early playtests. It's the same with spells/saves., a treadmill that just increases for the sake of it. Oh well, 5e is still new and fresh, can't have all the stuff I wanted 
Rastapopoulos wrote:
Super stealth thieves is one of the reasons I rule that stealth in combat is quite difficult after the first "ambush" round (I impose disad on later hide checks, generally).
Psikerlord wrote:
Rastapopoulos wrote:
Bounded Accuracy isn't about the bonuses, it's about the DCs. The bonuses matter because of this, yes, but they're not the focus. The effect of BA is that the DCs don't adjust as the players level up. The effect of that is that all bonuses actually improve your chances of success. The rogue you descibe is the stealthiest a human can possibly be without the use of magic and has spent their lifetime working to achieve that status. It's OK that the mooks can't find them. They actually are that good.
Orzel wrote:
Sailing_Pirate_Ryan wrote:
ClockworkNecktie wrote:
Wuzzard wrote:
Maybe level 1 guards with 11 Wisdom who are not trained in perception, and do not have blindsight, shouldn't be encountering a level 20 Rogue...?
Is this a theoretical argument, or a practical one, exactly?
Edit: Oh, it's a philosophical one. Sorry, responded after only reading half-way. To that end: Near Impossible things are DC 30, so it will still take a level 20 rogue with 20 DEX and expertise a 13 or so to accomplish these tasks. But hey, they're level 20, let'er rip. You're legendary, so BECOME legendary!
The game is bounded. However, while it is roughly balanced overall, it is not balanced across pillars. Rogues are a little less powerful in combat. In turn, they are a little more powerful at exploration and social interaction. Yes, I realize that if you just analyze the largest standard deviation between any two rolls then exploration and social interaction can be made to seem to feature a large discrepancy due to the sheer size of the die modifier. That is because in-combat interaction tends to be one of the "more important pillars." The combat system is a little more complex than the skill check system. With a skill check you make a single roll with a binary success or fail. With combat, there is a complex network of rolls. It is much easier to feel discrepancies during the course of play when that network is brought into use. In order to balance the slight weakness rogues have in combat, a weakness that will be perceptible to players, the bonus to exploration/social interaction (via the skill system) needs to be potent enough that it too can be felt. A small bonus (+1 or +2 over other characters) is not really perceptible in that area of the game. As a result, the rogue gets a big bonus to a few primary skills.
TL;DR: The game is designed both on principles of bounded accuracy and spotlight design. The rogues slight weakness in one of the most important spotlights, combat, results in a stronger focus on skill use.
The first step is to realize what people say and the truth aren't always the same thing, and what people say that are trying to sell you something almost never is. Bounded accuracy is marketing, the game doesn't have it. Accept that and move on.
FFSAA wrote:
Why is it so many people still don't understand what bounded accuracy means?
Fighter bonuses to hit could be +500 and not only does it not break Bounded accuracy, it has nothing to do with it. Bounded accuracy applies to the defense or difficulty side of the equation, not proficiency. AS players progress they should feel more competant, thus an AC 18 difficulty (present at all CRs of monsters) becomes easier to hit as the players level. The difficulty itseld is not Bound to level or CR anymore. Expertise is, if anything, an extreme example of bounded accuracy, where competence grows about 1/3 faster than normal. Whether it is against a static difficulty number or an opposed roll (which is the same essential thing), players can expect to get better. Their compentency improves, the difficulties of the actions don't. Picking a medium difficulty lock at level 1 and level 20 is the SAME actual target number. A medium Lock isn't DC 10 at level 1 and 22 at level 20, that's bounded accuracy. Wether it's a static difficulty or an opposed-roll, the idea is the same.
The complaint expressed, and perhaps valid at heart, is that expertise grants too much compentancy. That competency is over-powered as compared to other classes and/or monsters. I think this could be true depending on the DM playstyle. If a DM allows all persuasion rolls to resolve any encounter, then this ability seems way too powerful. If the DM allows an intimidation check to scare away an ancient dragon, then yes, it's overpowered. We, my groups, might allow a charachter who passes an opposed persuaion check to get a dragon to stop and listen cautiosly for a few seconds before he roasts us, and simply woudln't allow a dragon to be intimidated by any player unless the circumstacnces or idea was truly fitting (however unlikely). In our games it is not overpowered, and nicely balanced with other class abilities.
Let me express a few more things that I think are important, at least about the way I DM, that roughly manages the capability of these kinds of abilities as well.
1. I try to balance the capability of a skill or proficiency check against the normal capability of other special chracter actions. It is one action to do a check. What can other classes do in one action that is comprable. A fighter can perhaps kill one enemy in one action, so a bard with intimidation can at most scare away one similiar combatant. A powerful combatant might need multiple checks. If the character wants to do something more comneserate with a daily or short rest abilty like a fear spell, then I might impose a greater risk to failing the check. "You want to scare away the whole gang by yourself, well then you have to get face to face with the boss, and move past at least 3 bandits in the process (oppurtunity attacks). If you get hit you;ll have disadv on your
2. Group checks. Powerful actions require support of the party. If the bard wants to persuade a group of bandits to leave the 'party' alone, then the bandits will size up every one, not just the bard. The Bard's value is expressed in his tremendous contribution to the group check.
3. Mixed results. If a player wants to do something well beyond what seem liek his normal capability, the the effects can be mixed.
a few ideas,
FFSAA wrote: