5e Math, Take 2!

File: Probabilities.png (69KB, 548x1365px) Image search: [Google]

69KB, 548x1365px

>>52905037
[Part 2]
Anyways, I ran some numbers based on what I read in the thread to see what they were talking about (will post as pictures in the next couple posts). I ran the numbers on >ii, as that was the part that seemed most interesting.

The best example given was when a trained PC attempts a DC 15 task and fails, a party member without skill will almost certainly succeed, carrying the group, but making him look like an idiot.

So I thought about how to examine this, and the way I see it, there are basically 3 types of checks:
>A. Only one attempt - either you can't try again, or it gets harder with each failure, or failure results in serious consequences.
This one isn't "giving us trouble".
>B. Everyone has to pass - Whether its jumping across a pit, or outrunning a bear, or the party is separated and being interrogated, or you're trying to all stealth into a castle - if one person fails, the whole group suffers consequences, and the whole group has to try independently.
This one seems like it would "give us different trouble", based on the math I ran. Not what I'm going to talk about first though.
>C. If one person passes, the group is good. This is generally your knowledge checks. With some DMs it might also be social checks and investigation checks (With others the social and investigation checks might be type A)
This is the one where the trained guy looks like a chump next to his allies. This is the one that was called out as not fun in that thread.

File: Trained Probabilities.png (7KB, 271x274px) Image search: [Google]

7KB, 271x274px

>>52905037
[Part 3]
>C.
So I ran the numbers (Anydice, GSheets) for DCs 5-24, for some plausible parties. For all DCs between DC5 and DC17 (except DC 10 and DC 11), the weaker (at skill bonuses) party rolling is like having a +11, and the stronger is like having a +12.

The AutoSuccess rule was posted (which is apparently a kludgey approximation of a "Autosucceed on <= AS+Prof+5 (No Expertise or other Modifiers)" rule), and it mostly resolves the issue when you hit either Ability Score 20 or Level 10 and have Proficiency, but until you reach that level you're going to frequently look like a nincompoop next to a bunch of generic adventurers without training in the thing.

File: Expert-Group Matching.png (15KB, 341x462px) Image search: [Google]

15KB, 341x462px

>>52905037
[Part 4]
It's a tricky situation.

I don't have the answer to make trained characters not feel like chumps in C. The frequency this would happen at level 17 isn't so high, especially if the autosuccess rules in the DMG are used (then you only need to worry about it happening at DC 16-24) but before then, it's rampant.

It also emphasizes two things:
>I. Unless it's for DC18+ checks, there's no reason to take Knowledge skills. Investigation and Social Skills are the same, unless there are escalating penalties for each failure.
>II. You should never use "the whole group needs to individually succeed" checks, it's basically a guaranteed failure.

Here's what I've got for a conversation starter/suggestions to consider:
>a. Knowledge checks are trained only?
>b. In C type scenarios do not allow all party members to roll. Allow only a single roll for the whole party - maybe grant advantage if the party has someone else with at least a +3?
>c. in B type scenarios, again, single roll, and give disadvantage if a party member has a penalty?

Thoughts?

(AnyDice)
http://anydice.com/program/b6d9
http://anydice.com/program/b75a
http://anydice.com/program/b75b

[Done]

>>52905037
TL;DR

Easy answer is to just double or triple or otherwise more heavily weigh proficiency bonuses for skills a character is trained in.

>>52905118
You'd need your level 1 proficient characters to net a +11 for it to balance out with type C checks.

And that doesn't address the problem of failures for type B checks.

It also means either really high endgame numbers, or it means giving the everyone the endgame numbers at level 1 and dropping the per-level proficiency scaling. Either option has cascading side effects to address.

Because of those things, I'm not convinced this is a problem that's best addressed by increasing the size of proficiency.

>>52905118
>>52905118
But I mean, if you wanted to both set up your chargen so you start with a 20 in a stat of your choice, and then say Proficiency = +6 full stop, I suppose that would do it.

It also gets rid of the "gradually increased specialization" thing that D&D normally has.

Still doesn't address the Type B issue though. Group stealth checks are still a shit.

Stealth checks aside, Type B situations are generally handled best by having the first character who succeeds do something to help the others succeed.

>Jump across chasm
>Can now catch other players who are attempting to jump or secure a rope to something on your side
>Check gets easier for everyone

>>52905541
Sure, that's better than trying to have everyone jump. But it's not always an option. That staircase scene in LOTR comes to mind.

>>52905371
So, I asked this question last thread, but everybody was too busy namecalling to answer:

If you dropped the number scaling in your primary stat and proficiency, and started both of them at the max value, what would be the consequences?

>Youd need to raise AC values.
>The players would have much less need for ASIs, and would take more feats.
>The gap between good and bad saves would always be at endgame levels, increasing the need to target the right saves of an enemy.
>Trained Characters will be much better at skillchecks and you can have tougher challenges.
>No number scaling: Characters stay in the same groove the whole game, easier to balance around.
>No number scaling: Characters may get bored at the lack of number advancement, if they're big into number growth.

Anything I'm missing?

>>52905864
Enemies will stay usable the entire campaign, as saves and AC and attack bonuses no longer change.

All you have to worry about is HP and Damage.

>>52905864
You want ASIs for damage *and* accuracy. 5E's damage numbers are stupidly fucking low to the point where any +1 helps and AC isn't ignorable so the +1 to hit is also important.

>>52906144
>All you have to worry about is HP and Damage.
>All you have to worry about
>All

Oh, is that it?

>>52906227
As opposed to normally, where you also have to worry about AC and Attack roll scaling?

That makes things easier. I don't know how you could think otherwise.

>>52906223
Sure he would. But if your stat is already at 20, you've got your ASIs worked in. My point was that there's fewer parts scaling with level, and that even a level 1 enemy would have a decent chance to hit a level 20 enemy with such small scaling. They wouldn't hit very hard, but they would hit.

>>52905864
You'd likely also want to rework ASI progression if you're going to work several of them into up front attribute generation.

>>52905864
>
If you dropped the number scaling in your primary stat and proficiency, and started both of them at the max value, what would be the consequences?

I can think of two.

1) A lot of arguments over whether a paladin's primary stat is Strength or Charisma, and similar arguments for Bladelocks, Monks, Rangers, Eldritch Knights, Monks, Spellthiefs, MONKS, M O N K S, Barbarians, and FUCKING MONKS.

2) A really fucking boring character progression.

>>52906401
Oh, and

3) A really incongruous experience. So my snot-nosed barbarian kid fresh from the hill clans and just starting his path of adventure, is already as strong as Conan the Barbarian? My wizard who only yesterday graduated from the Unseen University already has intellect on-par with that of Merlin?

>>52906452
Yes, they're just lacking in experience, hence having less abilities, weaker spells, and lower proficiency bonuses

>>52906452
>using characters that are in no way related to dnd as a comparison to the mechanical side of dnd characters

>>52905037
Why don't you just change Proficiency to "add your level to skill checks you're proficient in"?

And adjust DCs as necessary.

>>52906476
But my 1st-level barbarian might still beat Conan around, say, The Hour of the Dragon (he probably wasn't 20th by then), in an arm-wrestling match (I presume the guy still wants the Barbarian to keep his capstone ability). And my snot-nosed wizard from the Unseen University is just as good at trivia night as Merlin.

Y'all need to stop looking at the raw numbers and start remembering what those numbers are supposed to represent.

(The guy I was responding to, you'll note, stated that he wanted to drop the number scaling from both primary stat AND proficiency, starting both at their max values. Meaning your primary stat starts at 20 and your proficiency starts at +6).

>>52906401
>Arguments over what starts at 20.
Why? You give them higher powered arrays to pick from, and let them put things where they want. What's the argument?

>Boring character progression
The interesting part is the new class features, not number scaling.

>>52906524
Did you read the OP, at all? How would that address "all kinds of group checks are shit"?

>>52906541
>The interesting part is the new class features, not number scaling.

How do you address current features and ASIs (Fighter and Rogue especially) where number scaling IS the new class feature.

>>52906536
Did you catch the reason why it was suggested?

A group check with +0/+1/+1/+2 is as effective as a roll with +11 for like all DCs up to DC17.

+5 attribute and +6 proficiency = +11

>>52906559
Feats?

>>52906559
For arguments sake, what if it was
Ability @20(+5)
Prof @+6
And then tack on an extra +(1/3.5)*lv to add in number scaling, without the group roll being way better than skill.

>>52906541
>Did you read the OP, at all? How would that address "all kinds of group checks are shit"?
Not sure what the fuck you're talking about. I just thought this would be an instant fix for bounded accuracy.

Personally I don't think it's even a problem, but OP looks like he's investing way too much time into this.

>>52906521
Oh, shut the fuck up, you know the point I'm trying to make and trying to distract from it won't change that it's a fundamentally good point that will have to be addressed if you want to implement the proposed change into your games.

>>52906524
>Why don't you just change Proficiency to "add your level to skill checks you're proficient in"?

Because that leads to 3.5 and Pathfinder. Well, slightly less since you cap out at 20 rather than 23 in a skill, but fundamentally the same thing.

It creates impossibly swingy problems. Like, say, a 10th-level seafairing adventure involving pirates attacking a ship.

- Oh no! A storm strikes. Make Acrobatics checks to stand on the rolling deck while fighting pirates.
- - A character with 20 Dexterity trained in Acrobatics has a modifier of +15. He can pass DC 26 checks 50% of the time.
- - A character with 20 Dexterity untrained in Acrobatics has a modifier of +5. He can pass DC 16 checks 50% of the time. He can't ever hit a DC 26.
- - A character 10 Dexterity untrained in this skill has a modifier of +0. He can pass DC 11 checks 50% of the time, DC 16 checks 25% of the time, and can't ever pass DC 26 checks.

So what do you set the DC at? A DC 16 check literally can't be failed by someone trained in Acrobatics at this level under your system, but it's a coin-flip for someone untrained but still dexterous, and a dangerous gamble for someone without any notable dexterity. It's too easy for one player and too dangerous for the others.

But if you set the DC at something high enough to actually challenge the +15 Acrobat player, you make things far too difficult for everyone else.

(cont'd)

>>52906622
The op says the "problem" with bounded accuracy is not the number scaling, so much as it's the shitty effect of group roll mechanics.

So shitty, that if you were to try to fix it by +#s instead of changing the mechanics for group rolls, it would mean starting with +11 in the stuff you're good at.

There are charts and numbers in the first 4 posts and everything.

>>52906650
What the fuck are you even on about? How are group checks in any way fucked?

>>52906643
THAT'S THE FUCKING POINT.

>>52906643
The main reason for bonded accuracy is that limiting how high skills checks can possibly get makes designing adventures much easier. This is particularly true for Wizard of the Coast themselves, who publish adventures like Curse of Strahd or Storm Kings' Thunder but don't have the benefit of actually knowing the players who will be playing it.

By capping the skill modifier for most characters at +11 and by having that modifier progress at a steady rate that's easy to anticipate, it becomes much easier for Wizards of the Coast to design an adventure for characters of a given level and have any given character be capable of actually completing the adventure, rather than being stymied by a lock that's too difficult to pick; or by making them roll for skill checks that are laughably easy. Instead, they can present DCs that are challenging for trained characters and difficult but not impossible for untrained ones.

Expertise, meanwhile, allows characters in classes with a focus on skills to notably shine by being much better at a given task than the rules normally assume. While everyone else is fighting pirates on a rolling deck and desperately to stay on their feet, the character with Expertise in Acrobatics is zipping along without a care. Expertise is a case of deliberately breaking bounded accuracy, basically.

>>52906643
> But if you set the DC at something high enough to actually challenge the +15 Acrobat player, you make things far too difficult for everyone else.

It seems like the primary complaint isn't finding a DC that challenges characters of different aptitudes, but that trained characters with maxed stats could ever fail "simple" categories of task.

In the example given DC 16 is appropriate. The graceful acrobat basically ignores the storm, because that's what he's specifically good at. The nimble guy is pretty good, but he's still got a chance at failure. The averagish guy is going to fail a lot.. because he sucks.

This is the point that 3.5/pathfinder missed. Once you reach a certain level of aptitude in your specialty, obstacles from that category SHOULD cease to be a challenge.

>>52906661
Are you blind, illiterate, or just retarded?

>>52905044 >>52905054
Look at the pretty pictures. And the relevant snippets.

>>52905054
>>For all DCs between DC5 and DC17 (except DC 10 and DC 11), the weaker (at skill bonuses) party rolling is like having a +11, and the stronger is like having a +12.

>>52905063
>>I. Unless it's for DC18+ checks, there's no reason to take Knowledge skills. Investigation and Social Skills are the same, unless there are escalating penalties for each failure.
>>II. You should never use "the whole group needs to individually succeed" checks, it's basically a guaranteed failure.

It's spelled out in math, in black and white.

>>52906688
It's a stupid fucking point from a game design perspective, as outlined here >>52906705

>>52906729
No it's not, there is a very good reason nearly every single RPG does not do anything like that stupid fucking bullshit.

>>52906737
Most RPGs don't have a level progression. They just have a cap on attributes and skills, and let you buy up to it.

And often, attribute+skill = dice size, if it's a one die system.

>>52906710
>but that trained characters with maxed stats could ever fail "simple" categories of task.

Everyone has an off day. No NBA player has a perfect free-throw record, for example, as was pointed out upthread. This is particularly the case when that NBA player is trying to perform a free-throw - a simple task in and of itself - when a goblin is gnawing on his legs, and the thing he's throwing isn't a basketball, but a bomb, and his target isn't a hoop, it's a dragon's mouth.

>This is the point that 3.5/pathfinder missed.

It didn't miss it at all. 3.PF has a tendency to meticulously detail what the DC is for any given task and set it in stone for all time. The problem comes form the "swingyness" that results.

Like I said here, >>52906705, it makes adventure design a nightmare, particularly if you don't actually know the specific people who will be in your player's party, because you don't have players, because you're Wizards of the Coast and you instead have millions of consumers, each of whom have between them millions of characters.

In 3.5, Wizards of the Coast would know that a 10th level Rogue could be reasonably expected to have a +18 to Open Lock. What Wizards didn't know was a) whether or not a given party has a rogue, and b) whether or not that rogue has taken Open Lock.

A coin-flip challenge for the rogue would be a DC 29. For anyone untrained in Open Lock, however, this would be impossible to hit. So what were they supposed to do? Design all their adventures so that only parties with rogues that had Open Lock maxed out could progress? Or design adventures where the DC to open a lock was never beyond the reach of a party without Open Lock? But if they were doing that, why bother having locks at all when a rogue who IS trained in Open Lock will just walk right through them?

>>52906737
>there is a very good reason nearly every single RPG does not do anything like that stupid fucking bullshit.

Regale me as to the reason why most RPGs would prefer, with their published adventures, to include parts where it is possible that a party will not physically be capable of progressing, due to the players not having foreknowledge of what skills would be requires to overcome a given challenge and so therefore are being stymied by a obstacle designed to challenge someone who has been dedicated to advancing that particular skill since character creation but which is as a consequence impossible for characters who have NOT been so dedicated to that skill to pass?

>>52906878
Because other RPGs, including other editions of D&D, are built with the assumption that you'll cover your bases or you'll contract out NPC help?

Thread is tl;dr

What's wrong with using saga edition math of:
half level + stat + 5 if proficient + 5 if expert?

>>52906932
Because certain people get triggered at numbers that are larger than what you can count on your hands and toes and tell you to go back to Pathfinder.

>>52906836
NBA players would all be pretty low level so that example is terrible.

Everyone who wondered how "group checks" should be handled can consider what I've been using the whole time as the DM:
>If the group gets into a situation where all of them need to make a check (such as sneaking past some enemies etc.), only the player who has the worst stats for the situation (in this case, the worst stealth) rolls. If they succeed, I as a DM believe that if the worst succeeded, everyone else did so too.

>>52906932
Group stealth checks (Type B checks) still shit that way, for one.

For two, those bonuses are too low on the non expert side. Change the +5 to a +6, or have attributes at +6, and proficiency at +5 and it works out alright.

>>52906995
No point putting ranks in stealth with those rules. Unless the whole party is built to be good at it, everyone is shit at it.

>>52906932
And >>52906999 is just based on the math from OP.

Half level may be too fast scaling for 5e though. If the bonuses are big enough, sticking with somewhere around 1/4 lv is probably fine.

>>52906995
Why not average the bonuses and roll a single check?

>>52907041
I suppose that's one way to do it.

Or have the guy with the best bonus roll, with some kind of modifier. Presumably everyone else is following his lead.

>>52906999
>Group stealth checks (Type B checks) still shit that way, for one.

But the way stealth checks work for 5e is that if 2 out of your group succeed, all of your group succeeds, no?

I guess "jumping a pit" would be a better example of Type B, but even an untrained, non-good stat person would have an about ~ 25% chance for those under Saga math (assuming level appropriate challenge).

>>52906836
>specialist tears through challenges meant for non-specialists with no effort
>this is somehow a problem
Why.

>>52907125
I may be misremembering stealth, don't have my book open.

Jumping a pit would probably be a better Type B example though.

>>52906878

Because sometimes that shit happens in the stories I read and I want my game to look like the stories I read.

Do you have any further questions?

>>52907170
Do you enjoy having the wrong sort of fun?

>>52907143
Gotta keep those pesky players down, where they belong.

>>52906836
>So what were they supposed to do?
Design multiple avenues for progression. You have a skill-monkey with Lock Picking? Great, you open the door. You don't? Well, you can either track down the key (maybe involving some knowledge checks or social checks) or you can take an alternate route that's more dangerous.

Or, since we're talking about D20, a Wizard just casts Knock and chortles disdainfully.

>>52906836
>it makes adventure design a nightmare
No, not really, you build around the assumption that the players covered the Thief/Fighter/Wizard/Cleric bases in some form and let them get fucked/let the DM modify shit on the fly if they refuse to do so AND won't use hired hands for whatever they're missing. Designing for 3E is a shitshow that rarely works out well because many parts of 3E don't work as advertised, NOT because gork the barbarian can't nail a Knowledge check outside of his realm of expertise or pick a lock.

>>52906995
Better solution is to tailor the end-result to the number of successes. For example, those characters that pass are able to reach advantageous positions and/or have a surprise round while the characters that fail are caught out in the open.

>>52905037
Make simple tasks autosucceed for sufficiently skilled characters

Don't let ameteurs roll for suffiently high level challenges.

>>52906727
Knowledge skills aren't one and pass. They're group checks and there's potential for misinformation on low rolls.

Social checks are usually roll once. Investigation can usually be solved by spending extra time, with the downside of whatever that time costs you.

You don't need to change the system when you aren't using the system right in the first place. Rather, there's multiple valid ways to run it, and it's only wrong if you don't like the outputa method gives. So use a different method.

>>52907155
Jump this pit or you die is one of those "pick this lock to continue" kind of barriers. If it's necessary, then those who can't jump have to go around a slower way or spend resources to get across.

File: Variant, Automatic Success.png (524KB, 465x699px) Image search: [Google]

524KB, 465x699px

I'll leave this here again.

Wow this devolved into autistic screeching and name-calling really goddamn fast.

The point of the skill system as is is partly to have all PCs in the group the ability to contribute, even however little they can, It's a safety net for a fucked up roll really.

Making something like this is essentially telling the others to just sit it out, they're not needed, don't you want all of the party to function as a team?

5e doesn't need all of this maths at all, that's not even the point of the system.

>>52908408
It's also really easy to design around.

Those with acrobatics could walk a rope connecting the rooftops (possibly after someone jumped over with a rope).

Those with neither DEX or STR will have spells to help them out.

Those with neither DEX, STR or spells better be lucky as shit to survive as adventurers.

Wow, turns out the buttblasted autists are from the proponents of this method, even when faced with well reasoned and well constructed discussion questioning its worth and and the fact that Wizard's method works for what it is. Very good to know.

>>52908564
>Wizard's method works
>fact
Wew.

Why not use different dice e.g 2d10?
It would make it harder to roll high and make point boni much more important to reach high numbers

>>52905037
>Most of the thread was a clusterfuck of namecalling,
Welcome to neo-/tg/, enjoy your eternal summer.

>>52908804
Even the examples given in the other thread didn't display quite the parity between the skilled and untrained, coupled with the well formed and articulated arguments in this thread.
Calm your autism.

>>52908469
And that point is shit. There's a reason only a tiny handful of games try anything like it.

>>52908455
>...can aggravate the problem you're trying to solve
Wow it's like they knew that the bonuses were too small and went ahead and did it anyways. Fuck WotC.

>>52906932
Becuase it results in math that is too swingy, becuase in order to challenge someone trained in a skill you have to set the DC so high that a person untrained in the skill is unlikely or outright incapable of hitting.

This is fine if you want each class to have exclusively its own niche with no overlap. It's problematic if everyone on the team needs to, say, hop on a bantha and ride across the desert, and only one of them has any proficiency in ride.

Ironically the Saga system fails at its intended purpose, because the entire reason why they came up with that formula and ditched the skill rank system was becuase they wanted to make sure that if a GM wanted his players to hop on banthas and ride across the desert, they would all be capable of doing it without issue.

Or in other words SAGA was trying to do exactly what 5e has managed: ensure that only the most difficult skill tasks are beyond the reach of most players.

>>52907143
>this is somehow a problem

It's not a problem. The thing is that 5e's system ALREADY DOES THAT, without making the skill check DC so high that it is utterly impossible for non-specialists to pass - merely unlikely.

>>52907170
>Do you have any further questions?

No, but a comment: this is why you don't work at Wizards of the Coast or some other gaming company: because you design for YOUR players, not ALL players.

I'm actually not aware of another system besides 3.PF/4e D&D and derivatives that makes the local equivalent of skill checks utterly impossible for a character unless it requires specialized knowledge. Like, in White Wolf's World of Darkness, for example, any player could pass any talent or skill check becuase the Difficulty is never higher than 10, their dice pool is never smaller than 1, and only 1 success is ever needed to succeed.

I have a WIP 5e hack that addresses this problem. There's a lot of shit in there but typical DCs and scaling proficiency bonuses go a long way to fixing the math.

http://homebrewery.naturalcrit.com/share/HymIj_ZO6g

TL;DR: Bounded Accuracy designs its bonuses in part because of the way Armor Class generally falls among creature types. When you structure DC's the same way the game tends to structure AC, shit works.

>>52907369
There are 18 skills in 5e D&D. The Fighter/Rogue/Cleric/Wizard combination have between them, including skills from background, 18 skills. This means that the only way to have every single base covered is if the players deliberately build their characters to have zero overlap. More realistically there will be some overlap, however, because most players design the character they want to run, not the character the party needs.

>NOT because gork the barbarian can't nail a Knowledge check outside of his realm of expertise or pick a lock.

Usually it's more of an issue that Zot the Wizard is going to have to be left behind to die to the goblins chasing you because he can't make an Athletics check to jump a gap.

Or the group needs to travel overland from point A to point B really fast, but the only way to do so is on horseback. Once again Zot screws everything up for the party because he sucks at ride, can't make the DCs, and so if the party is attacked between A and B by outriders then he's going to be a giant load unless he gets of his horse, but doing that leaves him essentially immobile when compared to everyone else.

Or, more relevantly for my most recent play experience, the Thief is the only one with Stealth so half the game involved the Thief being 60+ feet away from everyone else and everyone having to rely on the Thief's lackluster Perception check results and/or the fact that if the Thief is spotted, she's gotta endure an entire round of combat by herself before help can arrive.

Or so on.

>>52911654
>Wow it's like they knew that the bonuses were too small

They anticipated that some autists would be stuck in the 3.5 mentality that it should be impossible for rank amateurs to ever even once succeed, and that specialists would never, ever fail.

There is nothing wrong with the math, the problem is the people whining that Gork the Barbarian got a lucky roll and beat Zot at trivia night once.

In most of the groups I play with we use guidance cantrip and the help action to make all out of combat ability checks with proficiency, advantage and +1d4.

Furthermore, many spells exist to help like Pass w/o Trace and Enhance Ability. Not to mention magic items.

If anything, these magical solutions call for ability checks to be more difficult, not easier.

>>52911866
Shadowrun. Try making a check a specialist failed as an untrained fuck, I dare you.

>>52907408
The party of amateurs dramatically outdoes the proficient guy from DCs 8-21 until level 17. Are you seriously suggesting that you only allow rolls above dc7 if trained?

Your suggestion is bad.

>>52908455
Only helps at all with type C scenarios after around level 10, and doesn't help type B scenarios at all.

It was brought up right in the OP.

What does reposting it accomplish?

>>52908478
Last thread people claimed the problem didn't exist because no math was provided.

This thread there is mathematical proof of when there is an issue right in the OP.

On group checks. Both when each person is allowed to take a crack at something and one success carries the group(Type C), and when the worst roll holds the group back(Type B).

The OP suggested changing the mechanic for group checks, and showed mathematically that under the way most people do group checks, for the professional to be *as* good in type C checks as the party of amateurs, it needs to be for between DC18 and DC21+ checks (depending on the size of his bonus), or he needs a +11 or better.

And the autosucces rule only helps with DC<=15 checks, and only once you've got your 20 Str.

>>52908469
Read the OP posts. They lay out the issue plainly. So far the only proposed solutions to the issue involve:
>Turn all type C checks into type A by introducing serious consequences for failure.
>Design around type B so in your campaign there are no type B checks
the group never faces consequences they have to handle for one guy failing a check?
>Change the group rolling mechanic such that you never have the party rolling for group checks individually, come up with a new mechanic for group checks.
>characters with a proficiency bonus (one way or another) start with a +11 to their rolls.
Doesn't address type B at all.
>Start with ability at 20 or at level 10, use the autosucces rules, and don't allow nonprofit isn't characters to roll on DC 16+ checks.
Only addresses Type C.

>>52908564
>>52908469
If you've got nothing but name-calling to contribute, you're wasting your time.

>>52908835
Does someone wanna run anydice and modify OPs programs so we can see the odds for these and see how "use different dice" plays out?

I suspect it wouldn't *solve* either problem, but it might work as part of a solution.

>>52911991
I can see at a glance that your fix doesn't fix it. Have you run the numbers in anydice?

>>52912027
There's no reason for Zot the wizard to even take knowledge skills, unless it's for DC21+ checks. Anything easier, and his drinking buddies will succeed far more often than he will anyways, until he hits level 17 (or gets a total +11 or higher through expertise).

And that +11 only matches their output for rolls DC17-.

>>52914756
>there is an issue

There isn't, though. The intent of the system is that for all but the most difficult of tasks, anyone has a chance of accomplishing it, but some people are more likely to accomplish it than others if they've dedicated time and effort to training in the skill.

You are complaining about the necessary consequence of this system, viz., that Grok the Barbarian can occasionally at low levels beat Zot the Wizard at trivia night (and like problems) provided the trivia questions are all DC 20 and below. The thing is that any attempt to "fix" this relative non-issue will necessarily detract from the original point of the 5e skill system, which, to repeat myself, is to ensure that everyone has a chance at all but the most difficult of tasks.

And you're also forgetting that it's not a real issue because it doesn't break the game. From a purely mechanical standpoint Zot is still more likely than Grok to answer any given trivia question correctly (increasingly so as their levels advance), but the game does not collapse in on itself if Grok occasionally beats Zot. It's not a bug, it's not a problem, it's just a thing that you do not personally like.

Oh, also, a sub-intent of the system is to make Wizards of the Coast's jobs easier by allowing them to know, +/- a few points, what the likely skill modifier for someone of a given level is going to be, and thus be able to design adventures based around that assumption.

>>52914957
>Anything easier, and his drinking buddies will succeed far more often than he will anyways

You're gonna have to walk me through how someone with a +5 modifier to a skill is less likely to hit a DC 15 than someone with a +0 modifier to the same skill, because I am just not seeing how that is mathematically possible.

>>52915129
Zot is much less likely to succeed at trivia than Grok barbarian, Steve fighter, Billy rogue, and Sally cleric combined.

Trying to be any good at such tasks <=DC21 requires a +11 or better. Otherwise the peanut gallery has much Better odds than you do anyways, just let them do it.

>>52915184
Did nobody read the OP? There's visual charts and a table.

Click the images at the top of the thread.

Open up the anydice programs if you wanna see the math.

>>52915184
Tl:Dr;
>Because the amateurs get 4 rolls.
Though OPs hypothetical peanut gallery isn't all +0, it's +0/+1/+1/+2.

>>52915257
>than Grok barbarian, Steve fighter, Billy rogue, and Sally cleric combined.

So you're telling me that because one guy cannot reliably beat four guys in Trivial Pursuit who are all working together against him, the system is fundamentally flawed?

THEN PAINT ME BLACK AND CALL ME A MONKEY, because I thought that was a pretty accurate reflection of real life. Indeed more than anyone a Dungeon Master should know this, because 90% of DM woes are caused by his players, who outnumber him, being able to think up things that he, as one guy, could not.

>Trying to be any good at such tasks

Zot IS good at such tasks. It's just that his brain, fabulously learnéd as it is, is not as good as four brains combined.

Whoop dee fucking doo.

>>52914873

Not sure what would be considered a 'fix' in that case. The major assumption under this errata is that not only do the DC's go down a few points, any reasonably experienced character is going to get 2 ranks of proficiency in a skill they want to be considered experts in.

Everybody can take 10 under non-stressful conditions, and experts take 10 at all times. The proposed DC table just adds 10 to the average bonus of characters at creation, and upgrades in tier as their proficiency and ability scores rise.

Amateurs auto-succeed on easy to moderate as they level, while experts auto-succeed hard tasks from the get go and by the end of their careers can contest nearly impossible tasks at a little less than 50%, assuming they don't have advantage, which is made easier by the third rank of proficiency which they easily have by that point.

That said I didn't address the part 2 questions.

http://blogofholding.com/?p=181

>>52915402
So why should Zot bother rolling then?

Why should he even take the skill? Just throw another +2 into the peanut gallery and up it to 5 unskilled checks. the group clearly doesn't need knowledge skills, someone will always know the answer without a specialist, and with a specialist, it's still rarely going to be the specialist who solves it.

>>52915436
There's no take ten in 5e.

The closest to take 10 is the autosucces rule, which averages out to a take 5 rule.

>>52915491

Well, I didn't say that the problem didn't need fixing in 5e. I said that take 10 IS my fix. I buffed that auto-success rule.

I don't think it's perfect though which is why this thread is a good thing.

>>52915593
Gotcha.

I personally think the solution lies in somehow changing the mechanics for group checks.

Take 10 means you start to match the peanut gallery in effectiveness around +6 though, which I admit is better.

I'll have to run the numbers on that one.

>>52915469
>So why should Zot bother rolling then?

1) Because Zot is more likely to succeed on any one given roll, and often you only get the one roll. When the Sphinx asks a riddle of the party and allows them only one answer, what makes the most sense is for the party to use the Aid Another action on Zot's check, giving him advantage.

2) Because the system is designed to enable Zot's companions so that if they individually need to pass a given check, they can, even if it's difficult. Zot's reward for investing in Knowledge is the ability to hit DCs higher than 20, and in doing so enable his party as a whole to progress.

>the group clearly doesn't need knowledge skills

No one in the group, even >>52915323's group of +0/+1/+1/+2, is capable of hitting a DC 25, so in fact they clearly do.

>>52915631
How many DC25 knowledge checks does the party really need to pass?

And in the sphinx scenario, theyd all roll knowledge before anyone talks.

The party is better off if the wizard takes whatever he can that *isn't* knowledge or insight.

>>52915631
For knowledge and insight, you almost always get one roll, per character.

>>52915701
If DC25 comes up, the party fails. The GM isn't going to have the game stall when that happens. So the one time a year that actually comes up you take the badstuff and handle it.

>>52915701
>How many DC25 knowledge checks does the party really need to pass?

I dunno. But they're really not going to like it when it happens.

Besides which, it's not just about Zot's knowledge roll, it's about the system as a whole. A DC 25 Knowledge check might not be that common. But a DC 25 lock? A DC 25 Acrobatics check when on a rolling ship in a tornado? A DC 25 Persuasion check?

Have you never heard the maxim that it's better to have a thing and not need it, then need it and not have it?

>>52915712
Taking a somewhat different tack on this - the party is REALLLY not gonna like it when they decide that because the 4 on them can untrained on average hit any DC under 20 better than 1 person specialized in a thing, they don't need someone trained in Perception.

And that's when a series of thieves steals all their money purses when they walk through town, because the thieves are actually trained in Stealth and Sleight of Hand and can beat their shitty Perceptions with ease.

>>52915869
Exactly.

Zot (and his party) will be glad he took proficiency in a skill with Type A or Type B checks instead, and that Zots proficiencies somehow include stealth, perception, and athletics, instead of arcana or geography.

Perception is type A as often as it is type C. You often have to roll it without the team to back you up.

>>52916067
This seems rather arbitrary. Any skill can be A, B, or C, and you will never find a system that satisfies all three without necessarily sacrificing the ability of untrained users to ever pass any but the most trivial of checks in that skill.

>>52916237
It's not arbitrary it's pretty simple.

It's type A if you can only take one shot at it.
It's type B if the worst party member holds everyone back.
It's type C if everyone can attempt it and the best check is the one that matters.

Apparently this article (link) talks about the same situation in the context of 4e.
>>52915436

Knowledge and Insight are basically always type C.

>>52916389
And, combined with 5es low bonuses, type C skills aren't really worth having until you've got a +11 - which is level 17 if you don't have expertise.

>>52916389
>Knowledge and Insight are basically always type C.

No, they're usually Type A, "only one attempt."

If you roll to see if you know something, and fail the roll, then you don't know it. Thinking harder about it will not allow you to know something you never did in the first place unless you're Psionic, I guess.

>>52916504
>And, combined with 5es low bonuses, type C skills aren't really worth having

Again, this falls apart the instant a DC 25 Knowledge skill check is required but the entire party, thinking themselves smart, has no higher than a +2.

Although it's a redundant point because 5e doesn't have a Knowledge skill. Each knowledge-type skill - Nature, Arcana, etc. - can be used for other purposes than simply knowing stuff.

The fact that Grok the Barbarian once beat Zot at Trivial Pursuit does not change that the party will need Zot to make DC 25 Arcana checks to decipher ancient runes or know the location of the Vale of Lost Women.

>>52916569
They're type C.

You don't know it after one attempt, but each other party member gets a try.

One attempt means if one party member fails, nobody else can try it.

>>52911866
>Becuase it results in math that is too swingy, becuase in order to challenge someone trained in a skill you have to set the DC so high that a person untrained in the skill is unlikely or outright incapable of hitting.

You don't need to challenge an expert and a scrub at the same time. If you don't understand that something hard for an expert is usually something that's outright impossible for a novice/untrained person, we are using different definitions for things. The 0 STR wizard, who doesn't even lift bro, SHOULDN'T be able to lift a gate that takes a 10 or better from Krognarr, the mighty.

The difference between an absolute scrub and someone with training and maxed stat is however only 10. If you put a difficulty to 20 (which is pretty hard), the absolute novice can still succeed on a 20, while the pro will succeed 55% of the time (risky!). Someone with either training or stats will have a chance of 30, which is, again, pretty good for someone who isn't actually good at this shit.

>>52914756
>If you've got nothing but name-calling to contribute, you're wasting your time.
This has been undeniably true of your side though, which is fucking bizarre because your side claimed the moral high ground by claiming it was all name calling and no provision of facts.

You don't get to claim the moral high ground again, you snug muppet, that's disingenuous.

Facts have been provided, arguments have been well reasoned and provided. Yet your side is still full of screeching autists and hecklers from the peanut gallery.
I only state what's true, that the other anon made the excellent point of only 3.5 era autists and autists in general are still stuck in the only one must absolutely be the best at a certain skill and the team oriented nature of the game is irrelevant.

5e is great in that it also allows the other team players to participate, there's been too many times in 3e, 3.5e and 4e where some characters don't even bother rolling. You're essentially advocating for this to happen again.

>>52916749
>there's been too many times in 3e, 3.5e and 4e where some characters don't even bother rolling. You're essentially advocating for this to happen again.
Seconded.

File: Untitled.png (70KB, 1142x403px) Image search: [Google]

70KB, 1142x403px

>>52916624
I'm really curious as to whether or not OP's math changes when it takes into account the Aid Another rule, then. Zot at level 10 can be reasonably assumed to have a +9 to Arcana...

Oh, would you look at that. His party gets better results if Zot is Aided in his Arcana checks than if his party competes with him.

>>52916749
>All that projecting
I'm advocating for fixing the mechanics for group checks, so 4 scrubs is no longer *more* effective than training in 95% of checks.

That, by no means is the same as calling for "unskilled characters can't possibly succeed".

The only name-calling I've seen is the "everything works perfect" side calling people autists.

>>52916931
Oh, for the record, you can duplicate this in Anydice by:

output [highest 1 of 2d20+5] named "Zot"
output 1d20-1 named "Grok"
output 1d20 named "Sally"
output 1d20+1 named "Steve"
output 1d20+2 named "Billy"
output [highest 1 of 2d20+2] named "Morons Zot is Surrounded By"

Zot is presumed to have a +9 Intelligence at level 10 due to +4 Proficiency and 20 (+5) Intelligence.

File: Untitled.png (81KB, 1161x406px) Image search: [Google]

81KB, 1161x406px

>>52916969
Whoops, Jesus, I screwed up. I gave Zot only a +5 modifier, or only slightly better than just his proficinecy bonus.

THIS is what the graph looks like if Zot gets his full +9 and is Aided by his fellow party members, rather than the four of them competing with him to no point or purpose.

>>52916951
You must live through life with blinkers on and fingers in your ears. The only name calling in this thread, if we want to continue to get petty like that, has only been from your side. Read the thread, and don't be disingenuous and act all smug.

And it's not projecting when your method as stated isn't really needed unless you want the play from previous editions to rear its autistic head again.

>>52917040
Actually in the interest of fairness, that's not true. I think OP is a whiny moron and have said as much. I've also thrown out "autistic" at least once.

>>52916951
You're still essentially saying the others shouldn't bother rolling because their chance for success is even lower than before comparatively speaking, I thought we had moved past that.

Tl;DR?
I admit to being lost at this point.

>>52917096
I actually don't think it's name calling if it's true, if you are that dude, you made excellent arguments. But you can't convince those whose minds have already made up and are fixed in their smug reasoning.

>>52917224
OP is fundamentally wrong for a number of mathematical and mechanical reasons, and has fundamentally missed the intent of the 5e skill system in favor of trying to construct a system where everyone in a given party sticks to their one thing and never deviates.

>>52917424
Oh, situation normal, same as last thead?

>>52917497
Yes, everything is normal here, now, we're fine, thank you.

How are you?

>>52917424
>>52917497
And if you make any arguments or points to the contrary, you get accused of being a contrarian, someone who just doesn't understand!, a name caller and an illiterate because you didn't read the op.

>>52917518
Tired.
>>52917562
I argued up to the last post last thread, but it didn't seem to work. Thanks for reading the book and for not letting numbers intimidate you.

>>52917587
>Tired.

Damn, I'm disappointed in you. You were supposed to say something to the effect of "we're sending a squad up", at which point I could regale you about the fact that we have a reactor leak up here. Large leak, very dangerous.

>>52917518
Who are you and what is your registration number?
>>52917587
Skimming by, is the math actually wrong, like it was for 4e and 3e, leading to a worse game, or is it someone doesn't like what the math leads to, despite it being sound for it's intended purpose?
>and before I am corrected, yes, 4e had it's math basically fixed towards the end of it's life, but it was fucked for the first crucial years, and it did a lot of damage

>>52916969
Your anydice for the schlubs Zot is surrounded by is wrong.

>>52905063
This is how you see the odds of at least one success from multiple party members.

You have to check each DC individually.

>>52917643
>mfw I started a joke, fell asleep, and forgot to reply the right line.
I think I'm collapsing on the keyboard again.

>>52917040
One guy got called a retard because he couldn't read.

You have any other examples?

>"My method"
What method is that? I'm simply evaluating things as they are and brainstorming ways to fix it. I haven't chosen a method.

>>52917131
What are you talking about? I'm saying type C skills are a bad investment under the current group roll mechanics.

>>52916969
Your group success anydice is wrong, I'll look at testing it myself when I get home.

>>52905037
I didn't read all the autistics screeching I suspect in this thread, but to add something if that wasn't already said:

I agree that the bounded accuracy system is not really ideal, because the only way to be better than everyone else is expertise, and the advancement is very linear and depends on the level.
You can't have a level 1 master craftsman, or even a wizard that know more than a rogue or a bard with expertise in arcana. This bother me, but I never tried to fix it (I just play other games)

If you want to fix it, maybe you can simply switch 1d20 to 3d6 for skills. It gives you a bell curve, make small bonus much more important, and really make a gap between someone untrained and someone trained.

>>52917650
Uh...

Pic related.

LUKE WE'RE GONNA HAVE COMPANY!

>>52917655
>Your anydice for the schlubs Zot is surrounded by is wrong.

Hmm, so it is, and for that matter so is Zot's seeing as on second glance the graph is going all the way up to results of 49 (or 2d20+9, rather than the highest 1 of 2 d20s with the highest getting +9. I'm not terribly familiar with Anydice. How do I get it to give me what I'm looking for?

>>52917224
OP mentioned there are 3 types of checks.

>1 roll or bust
>Each party member could succeed.
>All party members need to succeed.

He then started mentioning approaches to address it, and a bunch of other people started talking about irrelevant side points and name-calling.

>>52917029
Brought up the Aid another rules as a potential sidenote, but his math is wrong.

Someone mentioned bringing back take10 as a potential fix.

Someone else brought up the autosucces rules, despite the op pointing out their irrelevance.

File: tenor.gif (1001KB, 500x213px) Image search: [Google]

1001KB, 500x213px

>>52917879
Damn I am just sucking today. This is what happens when I only get 5 hours of sleep.

>>52917424
OP here.

1. What's wrong with my math?

2. Prove I'm "trying to construct a system wherein everyone sticks to their given thing and never deviates" because afaict that's a fabricated strawman on your part.

>>52917880
I'd like an example of b) and c), please.

>>52917880
>He then started mentioning approaches to address it, and a bunch of other people started talking about irrelevant side points and name-calling.
Those approaches aren't really all that needed as addressed in this thread.
Just because it doesn't align with your superior reasoning doesn't make them irrelevant or side points.

>Name-calling.
Fucking hell, this is the same as "I'm going to dob on you".

>>52917562
Yeah, I called you an illiterate if you're the guy who didn't read the OP and posted a bunch of stuff that had nothing to do with the topic or was directly addressed in the op and you had nothing to add.

I think I did that twice.

If you've got an actual counterpoint, I'm happy to hear it. If someone's math is wrong, point it out.

>>52917807
Address them all in a post, or several if you have to, of each skills method, a fix and it's current resolution in the system as is.

>>52917879
Look at the anydice in the op for the schlubs.

For Zot himself, iirc you want 1@2d20+x.

I'd like to see how aid another compares to each party member making independent rolls though.

>>52917927
>>52915436
Has an article where someone talks about this same issue in 4e with examples.

But tldr: B is stuff like group stealth checks and C is stuff like knowledge and Insight checks.

>>52917963
A lot of people have given you counterpoints, if you'd only bother to read them. Just because they aren't telling you you're so right, you're truly our savior, oh maths guru doesn't make them wrong or irrelevant in the least.

>>52905037
I think part of the real problem is that in later levels the game design assumes assistance by the team, spells, help action, etc, and thus the numbers are fine.

>>52917954
>Not needed
Okay, I'll bite:

Why is it desirable that type c skills are a bad investment, and that 4 schlubs rolling succeeds consistency and training does not until end game?

If you're saying that doesn't happen, please back that position up with evidence.

Is it the Aid Another anon's example? We haven't seen the math on that yet.

>>52917954
>How dare you suggest that calling someone a screeching autist is not a valid counterpoint!
But it's not a valid counterpoint.

>>52918085
>and that 4 schlubs rolling succeeds consistency and training does not until end game?
What the fuck are you blathering about there?

>>52917905
1. You didn't take Aid Another into account.

2. If you are trying to construct a system wherein untrained folk are incapable of meeting high DCs whereas trained folk breeze through most, then you are de facto trying to construct a system where super-specialization is rewarded more than a broad skill base; or in other words the exact problem of 3.PF D&D and its derivatives - the problem that SAGA Edition was trying to solve (and failed) and 5e is trying to solve (and succeeded).

>>52918146
Those weren't the actual counterpoints if you bothered to read the actually well reasoned arguments, but those were also statements in truth. But do whatever you have to do.

>>52918000
They're in the thread, scattered, but theyre here.

There's one such summary of approaches here.
>>52914756

The current examples of how the math works are plotted in the OPs posts.

>>52905037
>>52905044
>>52905054
>>52905063

If that doesn't answer your question, mention what we missed and I'll see what I can do when I get back from groceries.

>>52918150
Read the op I'm just talking about what's shown there with graphs and tables.

>>52918171
1. Fair. I'd like to see the math on that.

2. That's not my goal. That's never been my goal. I simply want skill ranks to matter more in type C group skill checks, and to have Type B group checks be easier for the group to go through.

>>52918210
Actually construct a coherent sentence in English please.

>>52918283
OP is short for Opening Post. In this case, it's a 4 part post. Welcome to 4chan.

Read the Opening Posts.
They contain the answer to ">>52918150". They do so with charts, tables, and and simple paragraphs, in plain English.

There you go.

>>52918409
No, you fucking autist, construct this sentence "and that 4 schlubs rolling succeeds consistency and training does not until end game?" in coherent English.

>>52918408
Your results show the odds of rolling *exactly* each number.

For our purposes we want to see the odds of showing *at least* each number.

Here's Zot's Odds of each DC, with advantage, with a each bonus from +3 to +17
http://anydice.com/program/b782

Look at the tables for "At Least"
Each number shows his odds of beating that DC.

Each table shows the odds for a character with a given bonus.

Zot's Party not included.

>>52918456
Four untrained individuals attempting type C checks as a group, will consistently succeed at them.
One trained individual attempting those same tasks on his own, is much less likely to succeed than the four untrained individuals.

Except in the following edge cases:
>His bonus reaches a +11
>The DC is 21+
>The Autosuccess variant rule is being used, the DC is 15 or less, and he has either a 20 in the attribute or has proficiency and is higher than level 10.

This does not factor in the Aid another rules, as people have not yet finished running the numbers on that situation.

What about that was unclear? He even told you he was just talking about the OPs charted results.

>>52918576
Hang on, I'm editing and getting my roommate, who is much better at math than me, to help me out. hence why I deleted the previous post.

You know, looking at lots of the DCs and ACs across 5e while assuming a character is proficient in a certain task, things settle in at around a 50% success rate for most of the level appropriate challenges.

Why not just house rule rolling 2d10 instead of a d20? Then a somewhat tricky task (dc10) is an auto success for a high level trained character, but a crapshoot for the novice. Additionally, DCs higher than 10 benefit far more from the flat bonuses. Just need to adjust DCs to be between 10-20, and the occasional 25+to throw off expertise (or for truly difficult/impossible checks).

>>52918657
The DC being 21 or higher is not an edge case, however. Particularly not if the character must contest his rolls.

>>52918085
>>52918200
>>52918259
Frankly I'm not seeing a problem if all members in the party can somehow contribute.

>>52918688
How many contested Knowledge checks do you see?

Insight, I'll grant you, sometimes that can get up there.

>>52918576
For doing this I prefer to stick to a single level, in this case, 10th; and a single modifier; in this case, +9. Mostly becuase it's easier to read.

What that graph tells me is that Zot with advantage has a 100% chance of hitting a 10, a 93.75% chance of hitting a 15, and a 75% chance of hitting a 20.

Doing "output [highest 1 of 4d20]" in Anydice, which should be Zot's party each making individual checks and then keeping the best result, tells me that they have a 95.90% chance of hitting a 10, a 75.99% chance of hitting a 15, and a 18.55% chance of hitting a 20.

Checking out Zot had 5th level with a +7 to Arcana checks (+3 proficiency, +4 Intelligence), Zot has a 99% chance of hitting a 10, an 87.75% chance of hitting a 15, and a 64.00% of hitting a 20.

So it would seem that Zot with advantage - from Aid Another - is more likely to succeed than the 4 morons who surround him each rolling individually and then taking the best result.

>>52918775
>How many contested Knowledge checks do you see?

In 5e? None. There is no Knowledge skill. The whole thing exists merely as an example ultimately dating back to me calling people whiny bastards because they didn't like that Gork the Barbarian has a chance of occasionally beating Zot the Wizard at Trivial Pursuit.

>>52918760
The "problem" is that some people don't like constantly failing checks they took the effort to specialize in, to then consistently watch *one* of the party members succeed and make them feel like a chump.

These people are talking about individual party members attempting, not using the Aid Another rules. The math provided in this thread doesn't consider Aid Another (people are running those numbers now).

But the argument being made is that it's not worth taking knowledge skills like Arcana at all, and insight is only worth having if you regularly face DC 21+ checks. Supposedly the wizard (for example) would be better off taking any other skill, due to the math of how such group checks work out.

>>52918960
There are several Knowledge skills. You primarily roll them to know stuff. Dropping "Knowledge" from the name doesn't change anything.

Here, I'll give you a list:
Arcana, History, Nature, Religion

Investigation generally works the same way.

Insight as well, but it tends to have opposed DCs.

Perception works that way sometimes, people often take watch. But if everyone is present, Perception works like Insight.

They're all "Everyone make me a roll for Information" skills.

Oh it's this thread again.

I stand by my fix that an X-point penalty for non-proficiency is good enough. The exact number is up to personal preference (I like 4), but the important bit is broadening the gap between skilled and unskilled users and then halting runaway bonuses so bounds stay bounded.

>>52918971
>The "problem" is that some people don't like constantly failing checks they took the effort to specialize in

But they don't constantly fail checks. A 5th-level wizard with a +7 in Arcana has a 65% chance of getting a 15 or better on any given Arcana check. Succeeding notably more than half the time is not what I'd call "consistent" failure.

>>52918971
They should really learn to roll better.
But that's also chance, with the swinginess of the d20 there's always the high likelihood for that to happen.

And the argument is a bit retarded, with DCA Paultin's low arcana number hasn't been really helpful, even though he's a bard and could get half proficiency whereas Strix, being trained in arcana has consistently gotten good results.
This is a game where the roll of 1 trumps all other rolls.

>>52919044
That really, truly is wholly irrelevant to the actual matter being discussed.

>>52919055
There are no critical failures for skill checks.

>>52919076
No shit. That wasn't the main point but clutch onto something however you can.

File: 1425529038111.gif (2MB, 600x338px) Image search: [Google]

2MB, 600x338px

>>52919104
SEE >>52918822

It be the Dubs of Truth. It's better to have a specialist with advantage then it is to have 4 schlubs with no modifier at all.

Apparently Wizards of the Coast knows what it's doing sometimes. Who'da thunk it?

File: 1447980120094.jpg (26KB, 640x480px) Image search: [Google]

26KB, 640x480px

>>52919051
>slightly better than a coin flip to succeed at middling checks when highly specialized

>>52918822
I tend to want to be able to see the big picture, but that means you need tables/charts/graphs.

Here are Zot's odds.

With advantage from aid another, it's certainly closer to the rest of the party rolling individually. I'm gonna grab the numbers and stick them into a table side by side so I can see them.

If "Aid Another" is close (or better), I could maybe see telling my players in my next campaign: You only get to roll once for the whole group, use aid another rules for the rest of you to help.

File: Untitled.png (42KB, 1143x405px) Image search: [Google]

42KB, 1143x405px

>>52919138
And here's a nice graph.

>>52919155
Your fucking 5th level, what do you want?

File: Probability-Adv.png (52KB, 548x455px) Image search: [Google]

52KB, 548x455px

Here's a chart with the odds when you give the trained guy advantage.

I grabbed the relevant numbers from anydice, built a new table to make them easy to compare, and then graphed them.

>>52919067
>irrelevant
>The skills where you have *Everyone* roll to know stuff are irrelevant to the discussion about what happens when you have *Everyone* roll to know stuff.
Wat??

>>52919138
Yeah I know bro, that's inconceivable isn't it?

Actually Paultin's low arcana has been unhelpful and consistently less helpful compared to Strix being trained in arcana, who'd've thunk it in an actual game?

>>52919226
...who the Hell are Paultin and Strix?

>>52919248
PCs played in Chris Perkin's livestreamed game, Dice Camera Action.
Paultin is a bard, Strix is a sorcerer.

We have actual numbers from an actual game compared to numbers being run in a vaccuum.

>>52919284
>Anecdotes are better than knowing the actual odds.
What is this, Paizo?

>>52915323
>Though OPs hypothetical peanut gallery isn't all +0, it's +0/+1/+1/+2.
The real peanut gallery is -1/-1/-1/-1 because everyone dumps int.

>>52919587
Rolls that have actually happened, and so can conclusively provide supporting evidence, compared to what may happen or may not happen. Good argument.

>>52916389
>Knowledge and Insight are basically always type C.
They're type D. Standard group check. The group passes if 50% of the group pass the check.

>>52919685
Anecdotes are not data.

>>52919587
The point is helpful and an extremely relevant one because it's actually is happening in a live game, there's none of this graphing because that's also liable to deviate.
Sorry if it proves something you'll rather not want to see.

>>52919755
>being this much of an autist.
It's not an an anecdote if you can watch all sessions of play.

>>52919765
I dont care if I see it, it's just not especially relevant. It shows you like 0.01% of the big picture.

>there's none of this graphing because that's also liable to deviate.
The graphing is only *not* more valuable if someone did the math wrong or forgot to factor something in, in which case you factor it in/fix your math and away you go.

The odds mentioned here also don't take into include a chance a player having shitty dice that unevenly favor a particular result.

>>52919791
>>52919812
The plural of anecdote is not data.

When you've got 10,000 anecdotes measuring a specific thing, you'll have a point of comparison that holds some amount of actual value as data.

>>52919823
If you feel threatened by being proven wrong and in an actually conclusive manner without your precious numbers being run in a vaccum, just say so.

You're essentially saying lalala, doesn't matter because it's still irrelevant because I say so.

>The odds mentioned here also don't take into include a chance a player having shitty dice that unevenly favor a particular result.
Jesus fucking Christ.

>>52919856
Your point of comparison is litetally irrelevant because that's not how irl games work.

Believe what you want to believe I suppose, that's really what you want, right?

>>52919903
Before, it was dubious data.
Now you are basically claiming that every die rolls the same numbers in the same situation.
This just became really dumb.

>>52919934
Ok, I'm this Anon.
>>52919939
Explain your point without quoting, please.

>>52919903
>proven wrong and in an actually conclusive manner without your precious numbers being run in a vaccum
We're discussing probabilities.

Unless you've got monte-carlo quality simulations you're not proving anything in any way other than advanced math, and any anecdotes anyone provides are 100% irrelevant beyond possibly inspiring you to run the actual math on something.

>>52919939
It's using a irl game example, sorry if that offends your precious sensibilities.
I'm claiming your position and op's desire to rewrite the skill mechanism, as others have put forward, is an ultimately flawed exercise, which my example illustrates, sorry if that also offends your delicate sensibilities.

One may not prove the majority but it still does prove something, especially if all you have are numbers in a vacuum.

>>52919934
>Your point of comparison is litetally irrelevant because that's not how irl games work.
I can't make any sense of this statement. What exactly do you mean, and please justify your position. I have a guess at what you're trying to say, but it wouldn't make any sense so I'm just going to assume I must not understand your actual point.

>>52920011
>"An IRL Game Example"
Says nothing about the underlying odds, unless it has at least thousands of friends to back it up.

Your example is simply an example of one of many possible outcomes. It illustrates nothing.

Putting a sentence around a number does not make the number more correct or relevant.

>>52919939
>>52919983
For one it does blow out OP's claim last thread that an untrained PC and a skilled PC have skill rolls that are too close for comfort, and that success rates are too low for amateurs and dabblers and success rates too low for proficiency.
It's not true.

>>52916389
Except B checks don't exist. The DM's guide states that when group checks matter like stealth, the party passes as long as 50% or more of the checks succeed.

It's in the player's guide on page 175. Have you as a DM not been using this? Is your DM not using this?

I appreciate the math work but you seem to be referring to several kinds of skill checks that the books actively discourage DMs from calling for.

>>52917880
OP is getting a lot of flak because OP is pushing a lot of flawed premises, like the idea that there are 3 categories of skill checks and that some skills are primarily in some category type or another.>>52920101
>I appreciate the math work but you seem to be referring to several kinds of skill checks that the books actively discourage DMs from calling for.
This, really.

>>52920011
Man I, too, think OP is making a dumb point, but you are doing something that does nothing but undermine yourself.
I can show you a log of rolls of my game showing how I need a +11 to hit a 20, but Pally does crits every combat.
>>52920091
Ok, I agree with your argument, but its pretty shoddy logic.
I just don't see how one roll of the dice is meant to prove anything.
>>52920101
Neither does C in that way.
A single character can keep trying, and I would let the Int char do Arcana every time.

>>52920065
I'd imagine Wizards has more knowledge over game design than either you or me.

But the example isn't illustrative of just one case, it's happening throughout the play sessions.

>Putting a sentence around a number does not make the number more correct or relevant.
So what you and the op are doing?

>>52920101
And let me clarify my point. I'm not saying B and C checks should never exist or that DMs shouldn't use them.

I'm saying that the framework of the rules gives a better alternative that rewards the party both for having a specialist and for having reasonably competent non-specialists.

It also provides a nice excuse for party splitting. George the platemail cleric can't sneak? Looks like he's in the rear guard. Let Paul and Sally go up front, one of them is gonna pass their stealth.

>>52920151
It's not one roll of the dice, I don't know how you're confusing that. Maybe it's the tard piping up with the anecdote comment, there's been multiple rolls of the skill to compare.

>>52920219
Oh, yes, and I like the actual system, but a single campaign is just not enough dice to sustain this claims. TMy mystic expertises make me happy, feels like an acomplishment to find in the fuckhueg library the lore of an ancient temple and their rites.

>>52920151
>Paultin's low arcana number hasn't been really helpful, even though he's a bard and could get half proficiency whereas Strix, being trained in arcana has consistently gotten good results.

>y Paultin's low arcana has been unhelpful and consistently less helpful compared to Strix being trained in arcana.

Please note the use of consistently.

>>52920250
But it's a start at least, we also have evidence of how probability stacks up in actual play.

>TMy mystic expertises make me happy, feels like an acomplishment to find in the fuckhueg library the lore of an ancient temple and their rites.
Good stuff, bro.

>>52920091
(OP)
But that's not my claim at all. That was some guy in the previous thread about 5e's math's claim.

My claim was specifically a complaint about how a group of chumps attempting a Type C check has better odds than a proficient guy attempting that check, for the vast majority of DCs, and until lategame.

Your handful of dicerolls in a vacuum say nothing about that whatsoever.

>>52920149
How do you prevent those types of checks from coming up then?

>>52920272
What's a low arcana number?
In this case anyway?
Is or not trained, have a relevant Int mod, Why was he/she even trying if they had an arcana-proficient member in the party?
Also is DCA the Wizard's page show?

>>52920320
Type c checks are really rare and the point of type c checks is to pass around the spotlight, not as a barrier to pass or fail at.

>>52920174
Showing the math for actual probabilities.

>>52916389
Knowledge and insight are definitely not always type C. It sounds like you have a terrible DM.

The way it should work. DM says, "okay guys, you know enough that sometimes merchants like this will try to trick you, what are your insight bonuses?"

Every character rolls insight. Tell each character what they think based on their rolls, but don't tell them what they got. Let them suss out which of their opinions make sense. Let them ask follow-up questions to cross check.

Think about it. If you and 5 of your friends were in the woods and 3 of you thought a plant was safe and 2 of you thought it was poisonous, and one of the people who thought it was poisonous was a botanist who gave a good example why? What if the botanist thought it wasn't poisonous, but wasn't entire sure, and was in the minority?

Making group decisions is WAY more fun when people know who knows what but not who is right.

>>52920320
Did you actually read or is mouthing off part of your personality?
The dicerolls didn't happen in a vacuum but rather across multiple play sessions against actual checks by the DM, Chris Perkins.

Your numbers do occur in a vaccum however.

>>52920347
You read the fucking DMG before you try to redesign the game from the floor up? There's far too much for me to fit into one post.

>>52920369
Irrelevant in an actual game.

>>52920375
This. Knowledge is a group check or a roleplaying opportunity. It is not a "one check per person only one success needed" situation.

>>52920362
Super old but back when they were 4th level, Paultin's arcana was +2, and Strix's arcana was +4.

>why was he/she even trying if they had an arcana-proficient member in the party?
Why not? Because they were split up? Because all members should have a go as a team?

Yes to DCA being on the Wizard's page.

>>52920364
>Type c checks are really rare

They're the majority of Arcana, Nature, History, Religion, and Investigation checks I've seen ever come up in 5e.

How are you avoiding running those as Type C checks?

>>52920376
>Did you actually read or is mouthing off part of your personality?
Both

Your handful of dicerolls in game sessions *are* in a vaccuum. They say nothing about the thousands of other possibility combinations, and are not even enough to be indicative of a pattern.

>>52920415
The odds show you patterns you can expect to come up over a large number of sessions. Entirely relevant. You may as well be saying the math doesn't matter, and just ignore the rules in favor of flipping a coin once in a while when you feel like it.

Seems to me like the biggest issue is with type B scenarios. Why not just allow the party to split and let the people who can solve the problem solve it and those who can't stand guard? In a group of 4 players, it's not unreasonable for 2 to pass a thing and have to assist in a solution for the two that will fail. But maybe I'm missing the point, I'm kind of reading this like "everybody should win".

How long before we get for 5e what Pathfinder was to 3e?

>>52920435
Exactly. I mean, it could be, for say, a puzzle based on a history check. Like, once someone figures out the answer to the puzzle, it's obvious to everyone else. But when it comes to knowing if a hobbled goblin's fart will substitute for a hobgoblin fart in a ritual, the group is just gonna have to rely on their best judgment.

Now, does this solve all the math problems 5e has? Not at all, which is why I appreciate OP's work, it's given me something to look at when setting DCs.

>>52920503

What is pathfinder to 3.5?

>>52920517
A kinda but not really improvement that shamelessly rips off most of it.

>>52920472
>dicerolls
Nope. But good thing Wizards seems to have already accounted for the thousands of other possibilities, especially when they were running thousands of playtests. Nothing about that is indicative of a pattern.

I'm iffy to the relevancy of numbers in a vaccum, ie not from actual play, as I do theorycrafting.

>>52920517
Pathfinder is a dakimakura with an image of an anime character that is a living heaping pile of shit. The pillow also happens to have some real live shit smeared on it, but in places where there is no image of shit on the pillow.

3.5 is just a show about some shit.

>>52920472
>How are you avoiding running those as Type C checks?
By not being a shit DM.

>You know something useful
>You know something not useful
>You know you don't know something
>You know something not helpful but at least it moves the scene forward
>You know something else useful
There's a lot of different results to give from a knowledge check.

>>52920503
2 years or 6e.
What happens later.

>>52920581
This guy has read the motherfucking book.

https://www.youtube.com/watch?v=GlKL_EpnSp8

>>52905864
Delete ASIs, increase proficiency bonus.

>>52920581
>By not being a shit DM.
Oh shit, dude, they're going to screech at you for name calling again. I hope you're ready for this.

>>52920581
>You know you don't know something

"There are things we know that we know, and things we know that we don't know. But there are also things we don't know that we know, and things that we don't know we don't know."
- Donald Rumsfeld, philosopher.

>>52920517
Pathfinder is an improved version of 3.5. Note that this does not mean "good", just "better".

>>52920892
PF is really not a good system, it is if you like casters and taking revenge against the jocks.

>>52920924
>PF is really not a good system, it is if you like casters and taking revenge against the jocks.

Yeah, but that problem exists in 3.5 as well. The problems with Pathfinder also exist in 3.5, but there are a number of problems in 3.5 that don't exist in Pathfinder (polymorph, nacht), and Pathfinder doesn't really introduce too many new problems.

>>52919755
>Anecdotes are not data.

Considering our sample size they're probably better looked at as practical experiments.

>>52920554
I'm iffy on batman wizard style theorycrafting, such as when people are talking about having the perfect options prepared to solve their problems, as though they can count on that to happen all the time (the least reliable source of information, unless they can somehow prove they will have the options they need available, or they factor in how frequently they will, and use that as part of their theorycrafting, or something).

I see probability statistics as the exact opposite of that (IE: it's the most reliable source of information), except when people start assuming they will have access to situational modifiers they may not have (which is why my initial numbers did not assume advantage. You can lose advantage).

>>52920581
How do these things prevents the other players from making knowledge checks after the first person rolls, exactly?

How is this not "Everybody rolls, take the best result"?

>>52922949
>which is why my initial numbers did not assume advantage. You can lose advantage

Sure, theoretically. However given that D&D is a party-based game that presumed 4 players and a DM, it is more than reasonable to assume that in any given situation where a character needs to make a skill check, someone will be nearby to use the Aid Another action.

>>52920485
It's more an issue with "if the group tries to accomplish something they need to succeed at together (such as sneaking past the guards) if one person screws it up, everybody faces the consequences", which makes the odds for success very low. Which the players pick up on after a few tries, and then they avoid doing those things in the future, which results in things like the rogue is halfway across the dungeon from the rest of the group and he dies alone because they saw a skill challenge as a class-feature access wall (extreme example, but for hyperbole's sake)

>>52920628
I may think he's a shit for the namecalling, but he at least tried to make an actual point, rather than just saying mixing the word autist with some vaguely insulting buzzwords as though that was an actual point.

>>52920581
>It's not a Type C check
>Different results which can be given from a knowledge check
How exactly does this prevent the whole party from rolling knowledge checks?

I'm missing what part of this makes it not Type C.

>>52920996
Pathfinder is definitely an improved 3.5.

It's also more accessible because d20pfsrd.

It doesn't fix everything wrong with 3.5, but it fixes a bunch of stuff.

It does nerf casters compared to 3.5. Unfortunately, it also Nerfs Combat Feats and Maneuvers, when it should have buffed them and made them easier to attempt.

And I can't help but feel Pathfinder Archetypes are just worse than the elusive 3.5 ACFs.

>>52923083
Pot, meet kettle. And your actual point is what?

Shooting down the well reasoned and articulated arguments because they're wrong and you're right?
Shooting down the posted examples proving the premise of proximity of successes of useless vs trained PCs is flawed because they're wrong and you're right?
Don't be a little shit.

So is 5e.pf done yet?

>>52923377
>>52923452
What, so I'm supposed to respond to insults with flowers?

Fuck that.

And I've yet to call anyone an autist. I've called people out for calling others autists, and stated my opinion about Mr. "Anyone who disagrees with me is a shit DM".

I also asked him to explain how his arguments have anything to do with the subject at hand.

>well reasoned and articulated arguments
What well reasoned and articulated arguments?

He covered different outcomes you can get from a knowledge check. Either he missed the point, or there's some piece of vital information he didn't provide which explains how that makes it not a Type C (Everyone gets a roll, and any success will do) check.

>>52924028
I dont think there's enough mechanical gripes with the core system for someone to try making a 5.PF before WotC discontinues it.

>>52924050
Ok, so you're the only ones who get to call the others autists and name callers, got it because you can do no wrong.

The well reasoned and articulated arguments throughout this thread if you actually bother to read if you can get off your high horse and can dislodge that stick firmly stuck up your ass, but selective attention seems to be your thing.

>>52924028
At least they're contained in their safe place here.

>>52924050
>I dont think there's enough mechanical gripes with the core system for someone to try making a 5.PF before WotC discontinues it.

Fix everything people have common issues with, add some stuff, rename shit, make some classes stronger and others weaker, add some out of combat utility on martials, put out more books then level 1-X adventures and put in new class choices, feats and fighting styles.

You'd have to be careful not to piss off WotC lawyers but it could be done and might even be able to make money.

>>52918171
Super specialization SHOULD be rewarded more than having more people with no training in them whatsoever, and at any rate that wasn't the problem with 3.PF's skill system nor was it what SAGA edition's skill system was trying to solve. 3.PF's problem was that the skills were poorly balanced against each other, that it was completely pointless in the face of the magic system, that most classes got far too few skill points, that there were too many point sinks for the same skillset like Move Silently/Hide and Spot/Listen, and that the stat distribution for useful skills favored spellcasters due to nearly every single one of them being Dex, Int, and Wis. SAGA was trying to solve the bean counting issue by making it an issue of trained or not trained and to make sure that level 20 Darth Vader could hit basic DC 15s more than half of the time, NOT to close the gap between specialists and non-specialists.

There is literally nothing wrong with specialists being able to autosucceed at checks untrained characters are struggling with while untrained characters are unable to even touch the DCs the specialist struggles with. That is how nearly every single RPG out there works and there's a reason for it.

>>52924349
>There is literally nothing wrong with specialists being able to autosucceed at checks untrained characters are struggling with while untrained characters are unable to even touch the DCs the specialist struggles with. That is how nearly every single RPG out there works and there's a reason for it.
Think people have been saying this was precisely the reasoning that Wizards may have been moving away from, that is enabling everyone in the party to contribute.

>>52924474
Which they did at the expense of making specialists good at their specialization until at least midgame, and people who are otherwise maxed but unable to get double proficiency bonuses are even worse off. It's shit for the same reason that level 3.5's low level skill checks and attribute checks in general are shit.

>>52924115
I have seen a ton of insults from your side. On my side, I called someone a shit (for namecalling, specifically), and someone called someone else an illiterate for making arguments that were pre-refuted right in the OP.

As for the well reasoned and articulated arguments:
>There was a good argument for why you should consider assuming the trained character has advantage from the Aid another action, rather than discounting it because it can be taken away.
>There was math done to see your odds at different DCs, with different bonuses, assuming you have Advantage. (But no comparison to see how it matches up to an average party who has more than a +0 between them, as yet).
Those were the only real argument I've seen that was well reasoned.

There were a whole lot of people claiming the math and its effect on the game was irrelevant, and some others who made arguments which seemed to be about other topics entirely.

Then someone claimed that Type B and C scenarios shouldn't ever even come up, so I asked how exactly you manage that. They spit out a bunch of potential outcomes to a knowledge check, without any explanation what that had to do with "How do you make up for the ridiculous math that occurs for 4 rolls vs 1 roll".

And here we are.

>>52924165
I mean, I know what I would want out of that, I can't speak to other people.

>>52924563
If those are your only conclusions from reading the thread, you are confirmed to have selective attention, have extreme bias and there's no helping your autism.

>>52924349
>3.PF's problem was that the skills were poorly balanced against each other
This is true. Some skills are worthless. Other skills, it's not worth bringing above a +5.
>That it was completely pointless in the face of the magic system
Also true. Diplomancers aside, high skill bonuses held little value.
>That most classes got far too few skill points
Also true. 2/Lv a shit.
>That there were too many point sinks for the same skillset like Move Silently/Hide and Spot/Listen
They reduced that problem in Pathfinder when they condensed the skill list, but they didn't go far enough.
>And that the stat distribution for useful skills favored spellcasters due to nearly every single one of them being Dex, Int, and Wis.
Yeah, more or less. Except for UMD, which is CHA Based.

"The range of potential skill bonuses is too high" is not a concern I had ever heard voiced by *Anyone* until 5e.

>>52924710
It's a good thing that 5e is different to PF and the genre it occupies is different, isn't it?

>>52924634
There were a lot of shit tangents and side arguments that had nothing to do with the OP, if those are what you mean.

There were also some suggestions about "Avoiding Type B and Type C situations as much as possible", but that was more about making contrived scenarios to try to avoid the problem coming up (at least where you can control it) than resolving the problem.

If there are other more salient points (which actually relate to the thread topic - either in addressing the underlying math issue or ensuring the situations with the underlying math issue never occur), that I actually missed as this thread happened, feel free to point them out, I am pretty confident I haven't missed any relevant points to consider; but it's certainly hypothetically possible I missed something.

And I would like to see an extension of OP's comparison table with columns for advantage. >>52905063

>>52924745
You say that like that makes up for stupid situations cropping up due to mechanics.

The only problem with 5e is that it is dull, you cant fix it like PF with 3.5

i know PF didn't fix shit and bla bla bla

>>52924798
More so due to your side this time, actually read through the thread without bias this time.

A point of discussion was that op's maths and assumptions are flawed, which also makes the maths irrelevant.

>>52924803
That's more to do with the DMs and contrived situations.

>>52924848
No, it has to do with the game's mechanics. Just look for any kind of knowledge check in an officially printed adventure and you've got pretty good odds of finding one that will cause exactly the issue that has been complained about.

>>52924745
>It's a good thing that 5e is different to PF and the genre it occupies is different, isn't it?

Of course it's a different genre. One is fantasy supers. The other is action movie fantasy dungeoncrawling. PF's genre is mostly vacant (PF, 3.5, 3.0, HERO, and some really high powered GURPS - that's about it), whereas 5e is more in the same (more crowded) genre as: (Buffy/Angel RPG, Dungeons and Zombies, GURPS Dungeon Fantasy, RQ Classic Fantasy, FantasyCraft, Savage Worlds, Iron Heroes, and a low point HERO game, and to some extent: d20 Conan/ 2d20 Conan/FFG Star Wars).

That doesn't mean there's something *wrong* with PF having high skill bonus totals.

Does that genre shift make 5e a better game? Not in the least. It's just a genre shift.

The many poor design decisions in PF, on the other hand - well, I still GM 5e. I'm no longer willing to GM PF (but I'll play it, and mine it for ideas)

>>52922971
Everyone who makes the check gets an answer based on their result. It's up to the players and the circumstance how it's used. Or you could just have them metagame and follow whoever rolled the highest number. Then you'd have to run it as a group check.

>>52924902
Could you cite an example or examples from the officially printed adventures?

>>52924926
Which leads directly the scenario of someone in the party outperforming the Wizard's check on average, which is something I will never, ever agree with no matter how much you try to downplay it as the result of being a different genre in PF.

>>52924980
Would appreciate a party of at least 5th level PCs, competing against a DC, say knowledge arcana. Thanks. Otherwise it'll be that anon's argument that anyone can make an argument with supposed statistics.

>>52907408
>>52914330
>An example
The high level party wants to move across a rope while balancing over a chasm. This is a hard challenge, and they are in an anti magic field or something.

The amazing rogue wont have to roll for it, unless he also wants to help those behind him.

Meanwhile, the 'amateurs' have to roll a dc 12 to get across. if the guy helps them, they get advantage and he has to succeed a dc 12 test. This should allow most characters to have a fair shot of succeeding, especially if they get advantage. Yet there is still a good chance of failure.

If they were low level facing the same challenge, the rogue will have to roll, too. Increase the DC to 18. The rogue probably has +5 or 6.
The low level party is not likely to succeed this daunting task. Perhaps it would be better to find another way?

>onto your point
Also for your point about a team of amateurs, consider most challenges adventurers face. A lot of these can be better accomplished by adding manpower, rather than adding skill.
Breaking open a door; add more people (accumulated strength).
Have to get over a wall; add more people (human ladder).
Get over river; stay together (fight the current).

Of course there's stuff like disabling traps, but you usually only get one shot of these anyways (one character can attempt it). Then there's stuff like knowledge checks where you would think a single skilled character would be better than a bunch of dabblers. However, you can easily fix this by giving the guy with proficiency in a knowledge skill more info for the same rolls, or letting him autosucceed as I described.

>Final assumption
All in all, I think a simple assumption deflects most of the weight of these claims. If all adventurers are at least a little bit competent in adventuring, then this "narrow gap" is not much of a problem.

All adventurers probably have some experience in the wilds. They have encountered things that increase their knowledge of the world. They have negotiated.

>>52924931
No. I got rid of my 5E books years ago and don't feel like downloading a PDF.

>>52924848
>A point of discussion was that op's maths and assumptions are flawed, which also makes the maths irrelevant.
I mentioned that. They made a good case saying OPs math should have assumed advantage because Aid Another. I may run the rest of those numbers myself later to see how they wash out.

>>52924848
>Due to Your Side
Which side do you think is my side, exactly? Do you think this thread only has two sides? The way I see it, this thread has three sides, and one of them is consistently conflating the other two as if they're the same.

So far as I see it, this thread has more than 2 sides; it's got at least 3.

>1. Everything works perfectly, anyone who says otherwise must be an autist (no claims of not understanding this time, because people were dropping hard numbers presumably).
>2. "The Numbers Are Too Damn Low! Everything should scale Faster!"
>3. The mechanics for group checks don't seem to work out very well in play, and we should look at that and see what's going on there, and brainstorm if/how they should be changed to make the game more fun.

My side is #3.

1 & 2 have started a lot of shit, and side tangents. I haven't seen the same from my own side.

All 2 & 3 have in common is they both disagree with you that "everything is perfect" and are willing to actually discuss things civilly (at least with eachother) #2 seems to get in those obnoxious side tangents with #1.

>>52924980
Think of it as everyone being able to contribute in a situation, and okay for when the party is split. The team should be acting as a team.

>>52925033
Okay, you got nothing and would just like to whine. Good to know.

>>52924931
Not that guy, but I'll take a look at my copy of 'Abyss' and see if I can find an example knowledge check of some kind.

Someone else will have to provide a sample party.

>>52925049
That doesn't help.

>>52924710
>The range of potential skill bonuses is too high" is not a concern I had ever heard voiced by *Anyone* until 5e.
4e lowered the gap in skill bonuses, and there was arguments about whether it was too much or not enough. 5e is pretty similar for its maximum gap, though that's rarer and only grows to that point at higher levels. You might have just ignored these complaints, but they have been there.

>>52925035
>1. Everything works perfectly, anyone who says otherwise must be an autist (no claims of not understanding this time, because people were dropping hard numbers presumably).
No one was saying they were autists if they believed otherwise, stop injecting shit into discussion.

You were also really quick getting onto the insult bandwagon but that's you taking the burden of the moral high ground. Again, don't act like a little shit.

>>52925049
Nobody objects to the rest of the party being able to contribute.

They object to the 4 untrained characters outperforming the trained ones under most circumstances (Again, whenever Type C checks happen, which is typically the "roll to know stuff" type of check).

That is, outperform them sufficiently that arguments were made that it's not even worth taking proficiency in the knowledge skills (including investigation), and you're better off taking other skills instead.

However, I haven't seen a side by side for the hypothetical party of untrained characters mentioned in the OP (just a "everyone else has +0").

>>52925087
Why? Did you forget it's meant to be a team game?

is the guy who hated 5e here? i dont remember his name but he haves a lot of stories

>>52925095
A Wizard in 4E can end up with +40 Arcana rather trivially while an unspecialized character with INT as their dump stat can end up with +15 or +16 to it. The gap between a starting character and a completely untrained, 10 in the relevant stat character can start off at a difference of +10, so obviously that's not the problem.

>>52925172
That's not an argument. You don't see this issue in Shadowrun and that's far more of a team game than 5E ever will be.

>>52924926
>Everyone who makes the check gets an answer based on their result. It's up to the players and the circumstance how it's used.
>Or you could just have them metagame and follow whoever rolled the highest number. Then you'd have to run it as a group check.
Hmm. I can't see them *Not* metagaming if they know who rolled the highest, and therefore just assumed it was unavoidable, but you've got me thinking: I suppose knowledge checks could all be moved behind the GM screen, with the GM knowing everyone's bonuses on those types of checks, and then telling the players what they know, and never telling them who got the highest roll...

What are peoples thoughts on that?

Is there a reason that's a terrible idea?

>>52925162
>They object to the 4 untrained characters outperforming the trained ones under most circumstances (Again, whenever Type C checks happen, which is typically the "roll to know stuff" type of check).
Please give us this comparison party, would be good to see.

>>52925203
Shadowrun is also known for its overly convulted mechanics, which 5e does not have.

>>52925157
>No one was saying they were autists if they believed otherwise
Calling someone an autist is an insult, full stop; whether you personally believe it to be true or not. If you call someone an autist, you're trying to derail things and start a fight, not proving their claim wrong.

>>52925177
Virt?

>>52925177
He showed up last thread, I dunno if hes in this one though.

>>52925220
+0/+1/+1/+2 was the party used in the OP's examples.

But someone mid-thread proved a party of +0's *Still* outperforms the trained guy (assuming you're not taking advantage into account).

Removing ability score limit can be fun

>>52925266
So your calling someone a shit is okay? Has that hit too close to home?

Why are you this sensitive? Do you start frothing at the mouth and ignoring the other points because someone injected autist in there? Calm your tits.

>>52925244
Which has very little to do with the issue at hand. It's *not* an issue that fundamentally comes from being a team game nor does it actually help it be a team game, it's an issue that comes from fetishizing small numbers while using a large die size. Again, this isn't anything new, you saw the exact same thing happen with 3E's stat checks.

>>52924798
Ok.
>>52908356
>>52919742
>>52920101
>>52920213
>>52920364
These aren't actually all mine. Just mostly.

Group check is a game mechanic in the book, in case you're confused by your own illiteracy and selective reading. You can look it up.

>>52924980
If everyone tries on their own, wiz will have the highest result out over his party of dicks (0,0,1,2) half the time at level 5. If you can't stand to share the spotlight on roleplaying even half the time, I'd understand why the rest of your party treats you like this. If Dumdum instead helps the wizard, it's 75%.
http://anydice.com/program/b797

>>52925266
I'm genuinely concerned you are autistic.

At character creation pick 1/3 of your skills/tools that you are proficient in. You have advantage on these checks.

>>52925388
I really think they don't want to share the spotlight. This is a concern as a DM.

>>52925301
>Do you think calling someone a shit for namecalling is different than calling someone an autist?
Yes, 100%. One is trying to bait people into derailing to start an argument, and the other is calling them out for doing so.

>Do you start frothing at the mouth and ignoring the other points because someone injected autist in there?
No, in fact I responded to his other points. I then asked how those points actually addressed the issue, and got a response, which made me ask if rolling such checks "behind the curtain" so the PCs don't know who passed (and therefore can't metagame) was a terrible idea.

Speaking of:

>>52925217
>Is that a terrible Idea? Why?
>It seems to me, to be all-win (aside from the DM having another job to do).

>>52925415
I'm fine with sharing the spotlight. It's why tier 3 3E is the best game possible in it. Those classes blow everything within their field of specialization out of the fucking water guaranteed while still being able to contribute when it's not relevant.

I'm not fine with specialists being limited to two specific classes, specialization meaning nothing until halfway through the game, and everyone who's between that point struggling to beat a team of retards on checks they shouldn't have any issue with.

>>52925463
Your justification to your own behavior is baffling, the pot calling the kettle black is eminently applicable. But you do you.

>>52925183
It starts at up to 5 from proficiency, ~5 from starting stat allocation, and adds 3-4 from stat gains on leveling and 3 more from skill focus. I think there's still racial +2s. Which gives you up to an 18-19 point gap between two epic characters before items anyone can use and powers with limited uses.

>>52925217
Bear in mind the first hurdle is that the wizard reads the runes and tells you he knows *something* then everyone else jumps in to translate the runes too for some reason, presumably just because they saw the wizard rolled low. It couldn't hurt to try it. There's even the passive check mechanics that suggest you should know your PCs bonuses.

>>52925474
5e doesn't even have buying of skill points, with only their proficiency and expertise, if applicable, applying to skills. But that doesn't mean the PCs with non-proficiency skills are retards, that's disingenuous.

>>52925561
Untrained, proficiency, expertise, assuming the latter two have the correct primary stats for it. I was also talking about *literal* retards, as in 8 or lower INT. I think there's a problem with a game when one on one they have a chance of beating out the second category of characters because bonuses are so compressed, and especially think there's a problem if the game's mechanics mean that having enough retards around means that they'll outperform you on average.

how about instead of adding profecciency to two skills, you have 4 skill points?

>>52925650
I'd actually argue literal retards are those with Int 7 and lower, or just starting from Int 6, which means quite a bit of a negative modifier. Int 8 still means average Intelligence I'd argue.

>>52925698
Asian average is 11
White average is 10
Black average is 8
Retard is 7

>>52925740
Wat.

>>52925698
It's a -2.

It's fucking nothing on a d20.

>>52925798
It's a difference of 4 if the "specialist" has +2, which is unlikely.

Phoneposting, so bear with me.

Wiz (level 5, +7 arcana) rolls against DC 15. There's a 65% chance you get the result you want and you're done.

The other 35% of the time, rogue (+2 int) rolls his own to check the wizard's result. No wait, he helps the wizard because he's not a dick, or because the player knows 2 dice at +7 is better. So now in total, 87.75% of the time you have the result you want and you're done.

The remaining 12.25% of outcomes, the other three dumb shits in this example want to roll too (+1, +0, -1) and now this is a group check with two in the party already failed. These three need to all pass to pull the total successes over half. For DC 15, that works out 2.625% of the time as you might imagine. Sonce this only even comes up when the geniuses have failed, there is a 0.3215625% chance for the peanut gallery to come up with the answer when the smarties fail, or 2.1223125% if the -1 can help the +1 and not be counted (so saving the party only needs two successes in three rolls.) Taking the best case, the party gets the result you want 89.8723125% of the time. 65% wizard, 22.75% help, 2.1% peanuts.

In comparison, the same party with their wizard not proficient finds overall success 79.33125% of the time. 50% wizard, 25% help, 4.3% peanuts.

Is that sensible? Is that acceptable?

>>52925820
I was comparing it to a normal person.

D&D stats act like they mean something, with shit like "Oh, you have 6 CHA, you must be a hideous autist that has trouble connecting with peoiple and expressing himself!"... when compared to an average 10 it's just a -2. Literally just as good 90% of the time.

>>52926146
I thought the comparision was meant to be against a specialist char, where the difference would be much greater.

>>52926259
For the thread, yes, I just made an off-topic remark to express my annoyance at the difference between what people think stats represent versus what their in-world impact is (especially in relation to skills).

Please excuse me for the derail.

>>52926320
No worries, I don't think D&D really matches up to the real world though bro.

>>52926351
It doesn't have to be real world simulation, I'd just like if the implications matched the mechanics.

OD&D, for example, had the skills roll under your stat. This meant that stats for skills were twice as important as now, so 6 INT would actually make you a lot more stupid then, than it does now.

>>52926376
Oh, and being untrained in a skill made you have to roll under half your stat (don't remember rounding but am going to assume down).

>>52925988
I personally think it's pretty great. It's also giving great evidence that the skill parity isn't as close as it seems.

>>52925498
>Your justification to your own behavior is baffling
I'll simplify then:
I don't object to namecalling. I object to the use of namecalling to try to derail a discussion or as though namecalling will "discredit" someone's well articulated points.

>>52925560
Yeah, I was thinking of passive checks there. Wasn't sure if passive checks would be appropriate, or if you should be rolling each PCs check for them to do it.

>>52925758
Anon should go back to /pol/ rather than trying to start shit.
He's talking about racial averages in American IQ tests.

>>52926690
It's subjective, don't be so fucking sensitive. The denigration is part of the point being made, you don't get to be the final arbiter.

>>52925988

Interesting breakdown.

A couple questions:
1. Can these characters retroactively roll to assist, after the wizard has failed? I would have thought that would need to be declared in advance. The book doesn't specify either way.
In my experience, Wiz fails, and then (because obviously you cant improve an already failed check without a special power to do so*) everyone else simultaneously rolls checks, at which point someone succeeds.
2. I think Group Checks are more to mitigate Type B (All must Succeed) scenarios, not Type C (One success is enough).

The math in the OP shows peanuts (at 4 dumbshits instead of 3) succeeds 82.25% of the time. Not using "Group Check" though, presumeably because Group check doesn't really make sense in the circumstances (you don't need *everyone* to get through, you need *anyone* to get through).

>>52926782
>In my experience, Wiz fails, and then (because obviously you cant improve an already failed check without a special power to do so*)
You could fail at a check but get help from another team member, thereby getting advantage.

>>52925988
Wha's the breakdown for the skill? +3 prof, +4 Int?

>>52926747
The only point it makes responding to rational arguments with insults, is that the person in question really wants to act like a twat.

If you make a point in your post, the point is no more than your logic, arguments, and evidence.

Most of the posts with insults in this thread, last thread, and /tg/ in general come down to broad, empty, unsupported claims, weird tangents with no explanation of how they connect back to the rest of the thread, or nothing at all beyond the insult.

Hence, insults are of no value in a debate other than to give everyone the impression that you can't back up your claims.

>>52926807
Yes, that's the question.

Most times I've seen that happen, the DM has said "No, you already failed, him "helping" doesn't justify you getting a second chance you wouldn't otherwise be able to take".

Made us real careful to remember to declare when we're using Aid Another before anyone rolls.

>>52926816
At level 5, that would be it, yes.

>>52926839
Can you just peace out? We really don't need your autistic spewing.

>>52926857
Then that's a pretty bad DM, just because you failed once shouldn't preclude you from another assisting you to roll again.

>>52926886
Well if you didn't know the answer to something. You don't know it. You've done everything you can think of and you've got nothing.

How does another character help now? It would make more sense for him to make the check with you helping him.

>>52926782
>Can these characters retroactively roll to assist, after the wizard has failed?
If not, that only lowers the group's overall chance to succeed, not the chance the wizard succeeds.

>2. I think Group Checks are more to mitigate Type B (All must Succeed) scenarios, not Type C (One success is enough).
They are for any scenario where a group result makes sense. A pub trivia game would be a group check as your table is only actually giving one answer. Or maybe it's a family feud style 3 strikes limit. Point being that wrong answers detract from your overall success, but may not prevent it.

>presumeably because Group check doesn't really make sense in the circumstances
I think most circumstances can be a group check if you want. It depends on what model of results you want. I gave a link to another example earlier for free for all roll results with the specialist bringing the best total half the time or more.

>>52926857
Kind of a dick move on the DM's part. They (dicks) often don't bother to justify not allowing further effort(leading to automatic success with 10 times the time) or a retry.

>>52926912
He says something, related or not, that gives you an inspiration or jogs your memory, like it's done in every mystery show ever.

>>52926976
Is there any reason to have the second most competent guy use the help action?

Anyone who can theoretically succeed at the task could make the help action, right?

>>52927171
He's just the one naturally most likely to intervene. It's essentially in character, but a lot of situations could allow anyone to help. And he helps because that is the mechanically most beneficial.

>>52927200
>Most effective.
How is that? When I looked at it, there doesn't seem to be a roll to help, you just say you're doing it, and if you're eligible to do so according to the DM, he gets advantage.

Or did I miss a thing?

>>52927318
Not him, pretty sure the assister makes a case and then you make the roll again with advantage. If the DM is reasonable they'd let this happen.

>>52927318
If you only need one success to pass, then rolling the higher bonus with advantage is better than rolling two separate individual checks.

Proficiency vs non proficiency should really be thought of as class vs cross class skills from 3e.