I'm just not all that excited by 5e

[TenaciousJ]

That +11 at level 17 for non-rogues, non-bards with no access to Guidance or the like is rolled against DC 10, 15, or 20 most of the time.  Another player can use the Help action to grant advantage as well.  DC 25 is something considered very hard to do and DC 30 is a nearly impossible activity for a non-opposed skill.  Yes, you need a little luck with the d20, but percentages are pretty good for proficient skills without considering the class-specific ways to boost them more.

[LordBlades]

That +11 at level 17 for non-rogues, non-bards with no access to Guidance or the like is rolled against DC 10, 15, or 20 most of the time.  Another player can use the Help action to grant advantage as well.  DC 25 is something considered very hard to do and DC 30 is a nearly impossible activity for a non-opposed skill.  Yes, you need a little luck with the d20, but percentages are pretty good for proficient skills without considering the class-specific ways to boost them more.

My problem isn't with the total rolled vs. DCs, but rather that much more rests on the roll itself rather than the bonus.

The simple fact of giving yourself advantage is worth about a +3 (average roll raises from 10.5 to 13.82). how many levels worth of prof. bonus and stat increases is that ?

[Strill]

The simple fact of giving yourself advantage is worth about a +3 (average roll raises from 10.5 to 13.82). how many levels worth of prof. bonus and stat increases is that ?

What is an "average roll"? Do you mean the average of Advantage's benefit at all 20 possible chances to succeed? That's not a reasonable estimate. There's very little likelihood that you'll be forced to make a roll with either a 5% or 95% chance to succeed, and that increases the value of Advantage.

[LordBlades]

The simple fact of giving yourself advantage is worth about a +3 (average roll raises from 10.5 to 13.82). how many levels worth of prof. bonus and stat increases is that ?

What is an "average roll"? Do you mean the average of Advantage's benefit at all 20 possible chances to succeed? That's not a reasonable estimate. There's very little likelihood that you'll be forced to make a roll with either a 5% or 95% chance to succeed, and that increases the value of Advantage.

Simply the mathematical average. If you average N results of 2d20 take best, the average result will be 13.82.

To put it in game terms: if you face a roll and want an even chance of success without advantage, you need the DC to be no more than your bonus +10.5, while with advantage it's your bonus +13.82. Fir most practical purposes, having advantage affects your probability to succeed at a given task at least as much as having an extra +3 (which is about 10 levels worth of prof. bonus).

[TenaciousJ]

That +11 at level 17 for non-rogues, non-bards with no access to Guidance or the like is rolled against DC 10, 15, or 20 most of the time.  Another player can use the Help action to grant advantage as well.  DC 25 is something considered very hard to do and DC 30 is a nearly impossible activity for a non-opposed skill.  Yes, you need a little luck with the d20, but percentages are pretty good for proficient skills without considering the class-specific ways to boost them more.

My problem isn't with the total rolled vs. DCs, but rather that much more rests on the roll itself rather than the bonus.

The simple fact of giving yourself advantage is worth about a +3 (average roll raises from 10.5 to 13.82). how many levels worth of prof. bonus and stat increases is that ?

How much should the bonus vs. the roll be?  The roll allows a chance that someone untrained could still pull off a difficult task with luck.  The proficient PC with a +5 ability score mod has a chance of a failure on a hard task.  That seems to fit a hard task's description quite well.  If the bonus significantly outweighs the roll, then the untrained and/or those without right ability modifiers may as well not even try to use that skill, but those with the right proficiencies and abilities will auto-succeed the task.  5E explicitly states the point of any d20 roll is to determine the outcome of an uncertain situation.  If the gap between untrained and train gets too wide due to available bonuses, the scenario is no longer uncertain.  At that point, why have the roll at all for unopposed skill checks?  You just have skill DCs that effectively say, "You must be this tall to ride."  5E does not have special rules for 1s and 20s except when it comes to attack rolls and death saving throws.

[littha]

Have base classes with dead levels!

There are no dead levels. You're mistaken. If a class doesn't get an explicit class feature, it's because they get a new spell level.

Paladin level 4 only consists of 1 extra 1st level spell and the stat advancement that everyone gets. (which was boiled in to character progression in 3.5)

[LordBlades]

That +11 at level 17 for non-rogues, non-bards with no access to Guidance or the like is rolled against DC 10, 15, or 20 most of the time.  Another player can use the Help action to grant advantage as well.  DC 25 is something considered very hard to do and DC 30 is a nearly impossible activity for a non-opposed skill.  Yes, you need a little luck with the d20, but percentages are pretty good for proficient skills without considering the class-specific ways to boost them more.

My problem isn't with the total rolled vs. DCs, but rather that much more rests on the roll itself rather than the bonus.

The simple fact of giving yourself advantage is worth about a +3 (average roll raises from 10.5 to 13.82). how many levels worth of prof. bonus and stat increases is that ?

How much should the bonus vs. the roll be?  The roll allows a chance that someone untrained could still pull off a difficult task with luck.  The proficient PC with a +5 ability score mod has a chance of a failure on a hard task.  That seems to fit a hard task's description quite well.  If the bonus significantly outweighs the roll, then the untrained and/or those without right ability modifiers may as well not even try to use that skill, but those with the right proficiencies and abilities will auto-succeed the task.  5E explicitly states the point of any d20 roll is to determine the outcome of an uncertain situation.  If the gap between untrained and train gets too wide due to available bonuses, the scenario is no longer uncertain.  At that point, why have the roll at all for unopposed skill checks?  You just have skill DCs that effectively say, "You must be this tall to ride."  5E does not have special rules for 1s and 20s except when it comes to attack rolls and death saving throws.

The fact that some level 1 character that receives the Help action from somebody else (that can be a complete blundering incompetent at the task being attempted) suddenly becomes as competent as a 13th level character (his effective bonus for the check increases by 3 on average) just shows me that your character's training in skills is largely irrelevant.

Bounded accuracy just doesn't work for the skill system IMO. Outside rogues it just turns it into a dice fest with the added 'find a way to gain advantage on this roll' minigame tacked on. Largely it doesn't matter what your character can do, as a lucky roll can solve the problem anyway.

[TenaciousJ]

Sounds like 5E is not for you because the same arguments you're making apply to attack rolls and saving throws in this edition.  Higher level characters are not wildly more competent (independent of build choices) than low level characters outside of HP, potential damage output, and spellcasting effects.  If you want to be so good at skills you never or rarely fail, the Skilled feat, numerous ways to access Guidance, Rogue/Bard dips, etc. exist.

Unbounded skill check DCs give more options for setting difficulty, but run into the problem I previously stated that players won't even try to do things they aren't completely maxed out to do and/or skill checks are no longer a challenge at all for the right builds.

Thought experiment:  What would you think if skill check DCs were all lowered by 5 and resolved on a d10 but no other numbers changed?

[LordBlades]

Sounds like 5E is not for you because the same arguments you're making apply to attack rolls and saving throws in this edition.  Higher level characters are not wildly more competent (independent of build choices) than low level characters outside of HP, potential damage output, and spellcasting effects.  If you want to be so good at skills you never or rarely fail, the Skilled feat, numerous ways to access Guidance, Rogue/Bard dips, etc. exist.

It's not entirely true. Take attacks for example. While a level x fighter with a Belt of Str and a level x wizard with a Belt of Str have the exact attack bonus with a pointy stick, the fighter (and pretty much any other class that's expected to poke people with pointy sticks) has a bunch of stuff (extra attacks and whatever) that make it vastly more effective at poking people with sticks than a wizard. Which is all good, because wizards aren't supposed to be good at poking stuff with a stick, it says so on the tin.

Now, in regard to skills, it doesn't really say so on the tin that apart from Rogues and Bards nobody is supposed to be good at using skills, but that's the feeling I got out of the system.  I used to play a Noble Elf Wizard in a 5e campaign who had Persuasion and, by virtue of Position of Privilege was using it quite often. From level 1 to level 12 my bonus in Persuasion has increased from +3 (2 prof +1 Cha) to +5 (4 prof +1 Cha). Taking the Lucky feat at level 12 was a significantly larger boost to my ability to use Persuasion than the previous 12 levels of character development.

Unbounded skill check DCs give more options for setting difficulty, but run into the problem I previously stated that players won't even try to do things they aren't completely maxed out to do and/or skill checks are no longer a challenge at all for the right builds.

Meanwhile, bound DCs run into another problem: everybody (including henchmen and everybody else you can persuade/coerce to try) will attempt to do anything, regardless of their competence in the task at hand because if you throw enough dice at something, a natural 20 will come up sooner or later. Build yourself a focus group of 20 clueless averages (10 in all stats, no skill proficiencies) and they can statistically succeed at any Hard difficulty task you set them on.

Thought experiment:  What would you think if skill check DCs were all lowered by 5 and resolved on a d10 but no other numbers changed?

It would probably be more to my liking. You no longer have Joe Average solving difficult problems he's clueless, and the actual bonus your character provides will matter more.

EDIT: on the other hand, upon thinking about it more, that would also be a pretty big boost to rogues/bards, but I don't think it would be a problem.

[TenaciousJ]

Sounds like 5E is not for you because the same arguments you're making apply to attack rolls and saving throws in this edition.  Higher level characters are not wildly more competent (independent of build choices) than low level characters outside of HP, potential damage output, and spellcasting effects.  If you want to be so good at skills you never or rarely fail, the Skilled feat, numerous ways to access Guidance, Rogue/Bard dips, etc. exist.

It's not entirely true. Take attacks for example. While a level x fighter with a Belt of Str and a level x wizard with a Belt of Str have the exact attack bonus with a pointy stick, the fighter (and pretty much any other class that's expected to poke people with pointy sticks) has a bunch of stuff (extra attacks and whatever) that make it vastly more effective at poking people with sticks than a wizard. Which is all good, because wizards aren't supposed to be good at poking stuff with a stick, it says so on the tin.

I covered that in the damage potential part.  I like the implication that a large enough army now actually has a chance to take down a level 20 character to reign in their behavior a little bit.

Now, in regard to skills, it doesn't really say so on the tin that apart from Rogues and Bards nobody is supposed to be good at using skills, but that's the feeling I got out of the system.  I used to play a Noble Elf Wizard in a 5e campaign who had Persuasion and, by virtue of Position of Privilege was using it quite often. From level 1 to level 12 my bonus in Persuasion has increased from +3 (2 prof +1 Cha) to +5 (4 prof +1 Cha). Taking the Lucky feat at level 12 was a significantly larger boost to my ability to use Persuasion than the previous 12 levels of character development.

+5 meant you had a chance to solve a medium encounter 55% of the time before Lucky came into play.  I pointed out the ways to boost that with a level dip or a feat if it was important enough to maximize the skill.  10% is a small gain, but the system gave you that 10% with no additional investment, since wizard is not a class you invest in for skill buffs, generally.  A 1 level rogue dip could have boosted that by 20% at 12.  A 1 level cleric dip for Guidance could have boosted that 5-20%.

Unbounded skill check DCs give more options for setting difficulty, but run into the problem I previously stated that players won't even try to do things they aren't completely maxed out to do and/or skill checks are no longer a challenge at all for the right builds.

Meanwhile, bound DCs run into another problem: everybody (including henchmen and everybody else you can persuade/coerce to try) will attempt to do anything, regardless of their competence in the task at hand because if you throw enough dice at something, a natural 20 will come up sooner or later. Build yourself a focus group of 20 clueless averages (10 in all stats, no skill proficiencies) and they can statistically succeed at any Hard difficulty task you set them on.

That's not how statistics works!  Their chance of success is 1-(.95)^20.  That's roughly 64% chance of success.  A mob can be more convincing than 1 person even with weaker arguments, though that's probably Intimidation instead of Persuasion.

Thought experiment:  What would you think if skill check DCs were all lowered by 5 and resolved on a d10 but no other numbers changed?

It would probably be more to my liking. You no longer have Joe Average solving difficult problems he's clueless, and the actual bonus your character provides will matter more.

EDIT: on the other hand, upon thinking about it more, that would also be a pretty big boost to rogues/bards, but I don't think it would be a problem.

Rogues and bards already hack the skill system pretty hard, but the guy between "rogue/bard expertise" and "totally untrained" gains from that suggestion I made.  But then "average" turns into "actually below average and incompetent."  A 10 ability score is supposed to be average even though it's not really average to a PC's preferred abilities.  5E was built with a lot less automatic gains in class levels in exchange for finer tuned specific gains per class level.  The result seems to be that rogue and bard have their niche as skill masters protected but multiclassing is a more potent source of skill gains than it was before.  Proficiency in a skill is necessary but not sufficient to be awesome at a skill.

The fact that some level 1 character that receives the Help action from somebody else (that can be a complete blundering incompetent at the task being attempted) suddenly becomes as competent as a 13th level character (his effective bonus for the check increases by 3 on average) just shows me that your character's training in skills is largely irrelevant.

That +3 you're talking about from the Help action isn't quite accurate because the gain of advantage is the highest when you need a 10 and lower as you move toward needing a 1 or a 20.  It's helpful to look at the target number you want to achieve.  If the commoner needs a 20, advantage is only getting him about +1, or 4.75% extra chance of succeeding.  However if you need a 15 for a hard Diplomacy check due to your +5 bonus to Persuasion, advantage is taking you from 30% chance of success to 51% chance of success, or about +4.  Your minimal investment in Persuasion had an effect on how much the Help action benefitted you.

e: So yeah you need a good roll but your minimal investment of proficiency and +1 Charisma affected how much effective bonus you're getting from being able to roll twice.

[LordBlades]

+5 meant you had a chance to solve a medium encounter 55% of the time before Lucky came into play.  I pointed out the ways to boost that with a level dip or a feat if it was important enough to maximize the skill.  10% is a small gain, but the system gave you that 10% with no additional investment, since wizard is not a class you invest in for skill buffs, generally.  A 1 level rogue dip could have boosted that by 20% at 12.  A 1 level cleric dip for Guidance could have boosted that 5-20%.

Sp if my wizard wants to be better at skills (even wizardly skills), he needs to dabble in either thievery and underhand combat (since Sneak Attack and Thieves Cant come automatically the moment you multiclass to rogue) or become the priest of a god? To me that's an argument against the skill system, not in favor if it. Why is for example a Rogue more  knowledgeable in Arcana than an equally (or more) intelligent wizard for example? Also, why does a wizard know vastly more about Arcane stuff just after spending a level NOT focusing on arcane stuff (taking a one level dip to rogue) ?

That's not how statistics works!  Their chance of success is 1-(.95)^20.  That's roughly 64% chance of success.  A mob can be more convincing than 1 person even with weaker arguments, though that's probably Intimidation instead of Persuasion.

What about more esoteric skills? should 20 illiterate barbarians have a 64% chance to succeed on hard Arcana, Religion or History checks for example?

[awaken_D_M_golem]

I just found out, Santa Claus 
is gonna bring me the 5e MM.

I'm trying to Divination who thought
I was such a good boy this year.

I'd rather enjoy the misbehaving,
so that later getting Coal will keep
me warmer when its -20^o out.

I can't help it, I feel a little bit excited.

[TenaciousJ]

+5 meant you had a chance to solve a medium encounter 55% of the time before Lucky came into play.  I pointed out the ways to boost that with a level dip or a feat if it was important enough to maximize the skill.  10% is a small gain, but the system gave you that 10% with no additional investment, since wizard is not a class you invest in for skill buffs, generally.  A 1 level rogue dip could have boosted that by 20% at 12.  A 1 level cleric dip for Guidance could have boosted that 5-20%.

Sp if my wizard wants to be better at skills (even wizardly skills), he needs to dabble in either thievery and underhand combat (since Sneak Attack and Thieves Cant come automatically the moment you multiclass to rogue) or become the priest of a god? To me that's an argument against the skill system, not in favor if it. Why is for example a Rogue more  knowledgeable in Arcana than an equally (or more) intelligent wizard for example? Also, why does a wizard know vastly more about Arcane stuff just after spending a level NOT focusing on arcane stuff (taking a one level dip to rogue) ?

If you want to be thematic, bards have been portrayed as the knowledgeable casters since before 5E, and now that is their thing.  Bards get expertise with a 3 level dip, lore bards would get 3 more skill proficiencies, and you still get all the same spell slots and eventual 9th level spells.  If you want thematic for 1 level, Knowledge cleric gives expertise in two of the intelligence-based knowledge skills.  And then you get medium armor and shields, and access to Guidance.  Knowledge domain tends to be available to the gods of arcane stuff.

That's not how statistics works!  Their chance of success is 1-(.95)^20.  That's roughly 64% chance of success.  A mob can be more convincing than 1 person even with weaker arguments, though that's probably Intimidation instead of Persuasion.

What about more esoteric skills? should 20 illiterate barbarians have a 64% chance to succeed on hard Arcana, Religion or History checks for example?

Barbarians are not illiterate in 5E, and the totem barbarians even have the kind of fluff that could be stretched to explain how they know something about more magical things.  And that's assuming they have a 10 intelligence instead of dumping it to 8, in which case their chance of success is 0%.

[LordBlades]

That's not how statistics works!  Their chance of success is 1-(.95)^20.  That's roughly 64% chance of success.  A mob can be more convincing than 1 person even with weaker arguments, though that's probably Intimidation instead of Persuasion.

What about more esoteric skills? should 20 illiterate barbarians have a 64% chance to succeed on hard Arcana, Religion or History checks for example?

Barbarians are not illiterate in 5E, and the totem barbarians even have the kind of fluff that could be stretched to explain how they know something about more magical things.  And that's assuming they have a 10 intelligence instead of dumping it to 8, in which case their chance of success is 0%.

It was just an example. Base thing that irks me is the following: you take a bunch (20) of intelligence 10 guys that have no clue about a given field (the most clueless intelligence 10 guys you can find), and every time you present them with a difficult problem in the said field there is a better than even (64%) chance one of them will know the answer. Moreover, there is a higher chance this focus group of clueless individuals will produce a correct answer than most trained specialists in the field (a non rogue/bard caps at +11 to a skill at level 20, meaning he needs a 9 or more to succeed on DC 20, giving him a 60% chance of success). I find that silly.

[TenaciousJ]

You have an interesting definition of average.  Average intelligence is not completely clueless all the time.  That's why it's average.  The level 20 guy can work with just 1 other person who uses the help action and goes from 60% chance to succeed to 84% chance to succeed.  You don't generally play D&D alone. Groups have greater chance of success than individuals, and the advantage mechanic ensures that not everyone needs to be equally skilled to contribute.

By 5E's standards, you're not a "trained specialist" if you don't have expertise.  You're just trained.

[Centinull]

Long Story Short: It's a matter of personal preference

3E Skill Adds eventually make attribute Modifiers nearly irrelevant

5E Attribute Modifiers stay relevant

[LordBlades]

You have an interesting definition of average.  Average intelligence is not completely clueless all the time.  That's why it's average.  The level 20 guy can work with just 1 other person who uses the help action and goes from 60% chance to succeed to 84% chance to succeed.  You don't generally play D&D alone. Groups have greater chance of success than individuals, and the advantage mechanic ensures that not everyone needs to be equally skilled to contribute.

By 5E's standards, you're not a "trained specialist" if you don't have expertise.  You're just trained.

By average I meant average intelligence (int 10). Since there is only so much knowledge one can accumulate in a limited timeframe, there's bound to be subjects one has (almost) no clue about. By 5e rules, such a person is still able to succeed on hard knowledge checks on subjects he/she knows nothing about.

Long Story Short: It's a matter of personal preference

3E Skill Adds eventually make attribute Modifiers nearly irrelevant

5E Attribute Modifiers stay relevant

The irrelevance of attributes in 3.5 is debatable. Of course it all depends on the level of optimization (at the highest level there are things available to set any attribute to how much you want, but none afaik for max skill ranks) and skill (some, like Jump receive a riduculous amount of synergy) but for many of them, if you're going for full skill optimization, attributes will matter (not as much as in 5e, but still a lot).

[TenaciousJ]

By average I meant average intelligence (int 10). Since there is only so much knowledge one can accumulate in a limited timeframe, there's bound to be subjects one has (almost) no clue about. By 5e rules, such a person is still able to succeed on hard knowledge checks on subjects he/she knows nothing about.

I have been assuming average intelligence is 10 throughout the discussion.  My point of contention is that just because a person has generally limited understanding does not preclude them from having ANY understanding or from knowing random facts that they do not understand but are true regardless.  If you put enough of these people in the same room and ask a hard question, it's not completely unreasonable that one of them may be able to answer.

Anyway, that's a tangent.  I demonstrated that the +11 guy has a realistic way to succeed on a DC 20 skill check 84% of the time independent of class features, and I demonstrated the usefulness of the Help action is dependent upon one's existing level of training.  There's no auto-succeeding a hard check just by virtue of being high enough level, but a player has several options if hard (or higher) skill checks are that important to him (I didn't even mention that Guidance is available via a feat yet).

I suggest that your issue is not with the skill system itself, but how easy, medium, hard, etc. are vaguely defined in the system by design.  Ask yourself how hard it is to hunt for game or to build a house.  Commoners do those things.

[Centinull]

attributes will matter (not as much as in 5e, but still a lot).

+5 attribute mod out of a +11 is 45% of the total
+0 attribute mod out of a +6 is 0% of the total

+4 attribute mod out of a +40 is 10% of the total (+4 abb, +2 racial, +24 skills, + 10 magic).
+0 attribute mod out of a +36 is 0% of the total (+2 racial, +24 skills, + 10 magic).

They barely matter in 3E,  and not nearly as much as 5E.

With the same amount of time put  in, people who are naturally inclined towards a skill remain better at it than someone who wasn't naturally inclined to begin with.

 

[LordBlades]

attributes will matter (not as much as in 5e, but still a lot).

+5 attribute mod out of a +11 is 45% of the total
+0 attribute mod out of a +6 is 0% of the total

+4 attribute mod out of a +40 is 10% of the total (+4 abb, +2 racial, +24 skills, + 10 magic).
+0 attribute mod out of a +36 is 0% of the total (+2 racial, +24 skills, + 10 magic).

They barely matter in 3E,  and not nearly as much as 5E.

With the same amount of time put  in, people who are naturally inclined towards a skill remain better at it than someone who wasn't naturally inclined to begin with.

At level 20, a 3.5 e character can have (for mental stats, your base physical stats are largely irrelevant in high level 3.5 for any purpose, not just skills due to Polymorph and it's bigger brothers, PAO and Shapechange):

18 (base)+2 (racial) 3 (venerable age)+5(stat increases which do not compete with feats)+5 (wishes/tome)+6(item)=39, for a total bonus of +14. With +23 skills (max rank is level +3, not level +4) and +10 magic, that's 14 out of 47, so almost 30%. Less than the 45% you got in 5e, but far from insignificant.

EDIT: If your argument is that BASE stats don't matter in 3.5, then yes, you are largely correct,but that is simply due to the fact that 3.5 offers by design a multitude of ways to augment one's base stats, while 5e offers next to none. Different design principle (in 3.5 your stats are expected to be augmented by magic at mid-to-high level so the system is designed with that in mind).

EDIT2: your comparison is also a bit misleading IMO. 5e base stats contribute a lot to very little (since the bonus, even at +11 is still small compared to the swinginess of a d20) while 3.5 base stats contribute a little to a lot (in 3.5 the amount of skill points greatly overtakes the 1-20 result of a d20, resulting in high level people always succeeding on stuff that would trouble lesser mortals).