PDA

View Full Version : An open discussion about skill and ranks



Indigo Knight
2019-01-28, 06:41 AM
For the moment, I don't have a ready and made change in mind. This is still the stage of 'figuring out'.

Also, most of the subject is around skills that can be used untrained. More the like of Jump, Intimidate or I don't know... Gather Information?

Mainly, I'm having a bit of tough time aligning skill use with parts of my own homwbrew. I think that this is mostly due to wrong mathematical implementation. Or at least wrong attempt of presentation of different cases. What do I mean?
I mean that it is possible for a character to rank up more and more and more during adventure and open such a gap with characters, who haven't invested in that skill, to such a degree that either they can't possible pass the DC, any DC, or that the former is able to auto-pass all checks in that skill.
More then that, the comparison of numbers when thinking of dynamic target (meaning opposed rolls) to that of a static one, which, once again, the latter one will rise and rise, and the former one will keep relatively in the same range.
In contrast to when characters wish to invest every level in their important core attributes – the rank number is supposed to keep increasing, but the average difficulty is set, meaning that at a certain point the skill rolls will overtake it. That is unless the target keeps rising with the levels and then why even have skill ranks from the beginning?

I think I need to change the progression to that of an asymptote function such that the more and more you progress in a skill, the less of impact there is of an increase in performance. The easiest way I can think of is an increasing demand of points:


1 point.
a points, where a > 1.
b points, where b > a.

etc.

The limit should probably be so that an untrained roll will be in range. I think that means something like;


0roll + limroll =< 20

Because 20 is the dice used for rolls.

There could be a bit of a difference between items with rolls of more dice and pick highest/lowest.


That's kinda where I'm at. If you know of a thread I might haven't seen, or an article about it, please share.
And, of course, feel free to comment.

MoiMagnus
2019-01-28, 10:53 AM
In most games (not D&D in particular), quadratic progression works pretty well, since it is both simple and effective:

0: 0 points
1: +1pt = 1pt
2: +2pt = 3pt
3: +3pt = 6pt
4: +4pt = 10pt
5: +5pt = 15pt
6: +6pt = 21pt
etc

Alternatively, the exponential works mathematically great, but usually feel unrewarding for the players:

0: 0 points
1: 1pt
2: 2pt
3: 4pt
4: 8pt
5: 16pt
6: 32pt
etc

Last solution is to make sure that "high level PCs" have big bonuses even in skills where they aren't proficient. The simplest example is the (infamous) V4 where, from memory, you had: skill = half-level if not proficient, and skill = half-level + 5 if proficient.
Pros:
+You don't have this gap that increase between skills.
+You don't have players trying to use their maxed skills in a weird way to make it believable that they are using intimidation to dispel magic.
+Your skill value in the maxed skills don't stagnate at high level, and you still continue to progress as quickly as at low level.
Cons:
+If not handled correctly at narrative level, it can give the impression than the PCs are universally good at everything for no reasons. (Since when do you know how to talk old-draconic ? Since I have succeeded at my skill check.)

If you want more gradation:
+ Use skill ranks as normally. (One skill point = one skill rank)
+ Add a skill named "Veteran", that cannot be chosen to be increased by any way other than the DM says "everyone gain 1 in Veteran skill"
+ Veteran skill can be used instead of any skill in situation that doesn't require very precise knowledge.
+ As a DM, make sure the Veteran skill remain high enough to be relevant compared to maxed skills, but low enough to not makes skills irrelevant.

Indigo Knight
2019-01-30, 12:04 PM
Exponential might turn to a frightening monster. Also, at a certain level it becomes so that instead of investing in that same skill choice, it becomes more profitable to buy every other skill existing (because you have a gazillion points). It seems fitting from a math viewpoint, but not that attractive from player's perspective.


But taking the first example, wouldn't that mean that skill progressing basically dies off at mid-levels? The need for more and more points will make it so that a certain level, the cost will pass the sum of skill points gained at all levels.
Meaning, by having a linear with positive slope (the skill rank cost) vs. a line with no slope (or 0; the skill point gain each level), the lines won't converge to the same number. They'll just cross.

What's the answer?
Could be that the math is too complex for a ToTM game.

MoiMagnus
2019-01-31, 11:43 AM
But taking the first example, wouldn't that mean that skill progressing basically dies off at mid-levels? The need for more and more points will make it so that a certain level, the cost will pass the sum of skill points gained at all levels.
Meaning, by having a linear with positive slope (the skill rank cost) vs. a line with no slope (or 0; the skill point gain each level), the lines won't converge to the same number. They'll just cross.

Though, you have integers, so either it grows to the infinite, either it stagnates to a maximum. The first method is easy to compute, by hand, but if you allow more difficult functions, I can suggest:



Rank
Increment
Total Cost


0
+0
0


1
+1
1


2,3
+2
3,5


4,5,6
+3
8,11,14


7,8,9,10
+4
18,22,26,30


11,12,13,14,15
+5
35,40,45,50,55


16,17,18,19,20,21
+6
61,67,73,79,85,91


...
...
...



Alternatively, you could just "cut" this to +5, and everything rank after 10 cost 5 points.

aimlessPolymath
2019-01-31, 12:19 PM
Exponential might turn to a frightening monster. Also, at a certain level it becomes so that instead of investing in that same skill choice, it becomes more profitable to buy every other skill existing (because you have a gazillion points). It seems fitting from a math viewpoint, but not that attractive from player's perspective.


But taking the first example, wouldn't that mean that skill progressing basically dies off at mid-levels? The need for more and more points will make it so that a certain level, the cost will pass the sum of skill points gained at all levels.
Meaning, by having a linear with positive slope (the skill rank cost) vs. a line with no slope (or 0; the skill point gain each level), the lines won't converge to the same number. They'll just cross.


I mean, that sounds like what you were asking for here:



I think I need to change the progression to that of an asymptote function such that the more and more you progress in a skill, the less of impact there is of an increase in performance. The easiest way I can think of is an increasing demand of points:

The whole point of an asymptote is that it levels off.

I guess I don't know what you're asking for in terms of numbers. Should a high level character continue to improve linearly? If he doesn't, then you'll reach this point eventually:

Also, at a certain level it becomes so that instead of investing in that same skill choice, it becomes more profitable to buy every other skill existing (because you have a gazillion points). It seems fitting from a math viewpoint, but not that attractive from player's perspective.

What sort of odds are you asking for? At what point do skills not advance anymore, and how does that compare to the cost of acquiring new skills?
How do you want to value a specialized character vs. a nonspecialized higher level character?
Should bonuses linearly reduce the difficulty of all checks at the same rate? Perhaps your chances of success at low-difficulty tasks "cap out", and so you'll literally never be able to auto-pass?

Stelio Kontos
2019-01-31, 12:45 PM
If quadratic feels too slow, and exponential feels too fast, Fibonacci would be juuuust right.

aimlessPolymath
2019-01-31, 02:27 PM
The Fibonacci sequence is also exponential. It's just somewhat slower about it.

Indigo Knight
2019-02-03, 09:06 AM
Though, you have integers, so either it grows to the infinite, either it stagnates to a maximum. The first method is easy to compute, by hand, but if you allow more difficult functions, I can suggest:
I'll have to run numbers on this and see what I come with.



The whole point of an asymptote is that it levels off.

I guess I don't know what you're asking for in terms of numbers.
That's exactly it - I don't know in numbers.



Should a high level character continue to improve linearly?
What sort of odds are you asking for? At what point do skills not advance anymore, and how does that compare to the cost of acquiring new skills?
How do you want to value a specialized character vs. a nonspecialized higher level character?
Should bonuses linearly reduce the difficulty of all checks at the same rate? Perhaps your chances of success at low-difficulty tasks "cap out", and so you'll literally never be able to auto-pass?

I think that a high level character would like to keep investing in his main skill, like in prior levels. But I don't want to increase his outcome number so that he would still be in range of characters that didn't invest. So that both him (close to 99.9999%) and the untrained (0.000000001%) have a some chance at rolling success for an untrained skill.
Because the chances are based on a d20 then the chances are based on multiplications of 5%. But this is seconderay to the need for both players to be in a rolling chance of success. If you can train so that your skills pass the untrained + 20 then there's too wide of a gap.
I don't know about aquiring new skills.
A specialized character need a roll of 2-20, because he's specialized. An untrained need a roll of 20. Other bonuses that are not from trained ranks can be had by all types of characters. So I guess they don't matter.
I'm... not sure.




If quadratic feels too slow, and exponential feels too fast, Fibonacci would be juuuust right.
heh... Fibonacci. ^_^

If anything, i'll be looking at
- x*x/100 + x/2 + 1

But that's no way something I'm going to toss at my players.

--------------------------


I'll try to give a specific example. Jump includes the following table.



Long Jump Distance
Jump DC


5 feet
5


10 feet
10


15 feet
15


20 feet
20


25 feet
25


30 feet
30




Right off the bat there's something massively wrong here. And that's because that coupled with the exchange rate of skill ranks (1=1), reaching 30 feet from 25 is as easy as going from 5 feet to 10 feet. In truth, without magical assistance, reaching world record at 30 feet is something only a few of the few can accomplish. That's why I think it needs to act more as a quadratic equation rather then a linear one.
This is not the only skill that behave this way. Consider ride for example.

Now, I could instead of messing with rank progression mess with DC. So that the table would look more like so:




Long Jump Distance
Jump DC


5 feet
5


10 feet
10


15 feet
20


20 feet
40


25 feet
80


30 feet
160



But the big restriction here is that the system is built around rolling d20, by scales of 5%.

johnbragg
2019-02-03, 11:00 AM
Exponential might turn to a frightening monster. Also, at a certain level it becomes so that instead of investing in that same skill choice, it becomes more profitable to buy every other skill existing (because you have a gazillion points). It seems fitting from a math viewpoint, but not that attractive from player's perspective.


You can fit that from a story point of view, if you want to. By 8th level, the Big Stupid Fighter has been around the block enough times that she has a solid chance to decipher the ancient runes (these DC 15 runes are a lot like the DC 15 runes the Scholar deciphered on a 2nd level adventure), the Clumsy Wizard has been told a dozen times the rudimentary basics of stealth, etc. So maybe a flat or very-slowly-scaling Veteran bonus to untrained or zero-rank skills, while the specialist keeps sinking points into hitting the increasingly difficult challenges? Septimus the 7th level wizard has a decent chance to sneak past a 1st level orc, but sneaking past Oculus the All Seeing Eye is something you need Whisperfoot the 7th level rogue to handle?

That's a sidetrack from the linear vs exponential question, but I think Veteran bonus (Flat +2? 1/3 level rounded down?) is an option to look at.


In contrast to when characters wish to invest every level in their important core attributes – the rank number is supposed to keep increasing, but the average difficulty is set, meaning that at a certain point the skill rolls will overtake it. That is unless the target keeps rising with the levels and then why even have skill ranks from the beginning?

Because maybe DCs over 30 are legendary-scale accomplishments. The high level thief CAN auto-succeed on picking a cheap lock or sneaking past a mook or fast-talking a random barmaid. Why keep investing? Because the DCs are higher to sneak past or fast-talk a dragon or unlock an arcane lock.

I'd argue that, in a magical world, everybody is magic. High level characters do more magic than low level characters--treat ridiculously high Skill Checks as feats of magic. Basically the idea of Epic Level Handbook skills, but better implemented. But a STealth roll of 50 could equate to a form of invisibility--you project a field of not-noticing, so that it's impossible to pay attention to you.

PairO'Dice Lost
2019-02-04, 10:10 PM
Right off the bat there's something massively wrong here. And that's because that coupled with the exchange rate of skill ranks (1=1), reaching 30 feet from 25 is as easy as going from 5 feet to 10 feet. In truth, without magical assistance, reaching world record at 30 feet is something only a few of the few can accomplish. That's why I think it needs to act more as a quadratic equation rather then a linear one.

The key question here is, where in the level scale would you say a real-world Olympic athlete falls?

There's a good case to be made (e.g. here (https://thealexandrian.net/wordpress/587/roleplaying-games/dd-calibrating-your-expectations-2)) that people in the real-world would be represented by characters in the level 1-6 range. And of course we don't have, y'know, magic in the real world (though performance-enhancing drugs are arguably potion equivalents :smallwink:). So reaching the world record for any real-world achievement should be doable by whatever level you determine the real-world equivalent would be, with whatever modifiers would be accessible to real-world athletes.

If you go with the "real world maxes at level 5" argument like the article at that link makes, and your formula for Jump distance is "DC X to travel X feet," then a dedicated athlete should be able to hit DC 30 at level 5 with some regularity (not automatically, but much more often than 5% of the time). If you decide that real-world maximum is 20th level (which isn't at all the case for D&D in any edition, but you can always adjust the intended level range in a homebrew system), then given that same Jump formula a character should be able to hit DC 30 with regularity at level 20.

In general, when working with a 3e-ish baseline (20 levels, ability scores in the normal range, expert progressions are around +1/level, etc.), you should never ever have DCs above 45 for anything you intend 20th-level characters to do with any sort of regularity. A character very dedicated to making a certain kind of roll will have around +20 from level, +13 from a maximized ability score (18 base, +2 racial, +5 level, +5 inherent, +6 enhancement), and +5ish miscellaneous (a +5 sword for attack rolls, +2 synergy +3 Skill Focus for skill checks, etc.), and can generally finagle a +2 circumstance bonus (high ground, Aid Another, a +2/+2 skill feat, Favored Enemy, whatever) most of the time, for a total modifier of +40; less-specialized characters, or characters rolling secondary attacks and skills, are going to have closer to +30 total (+20 level, +6ish ability score, +4ish miscellaneous+circumstance). There's a reason the big chart of sample skill DCs in the DMG tops out at DC 43 for an extremely niche task attempted by an extremely specialized character, and CR 20 monsters have ACs in the 35 to 40 range.

And keep in mind, that's a 20th-level character, who slays liches and dragons and demons on a daily basis. A 20th-level fighter doesn't, and shouldn't, give a single [bleep] about an Olympic long jump in the existing skill system, and he should continue not to care in any modified system. DC 40 for a 20-foot long jump is completely ridiculous when the world record for a long jump by an 18-year-old high school student is about 26 feet. It makes total sense that a high-level character can go from +25 to +30 with the same ease that a low-level character goes from +5 to +10, because the high-level character is more competent in every way than the low-level character and that includes being able to blow past world records that the low-level character struggles to meet.


If your main concern is that only a few specialized characters should be able to make high DCs and you reach a point where characters who haven't invested in a skill can't make those DCs...what's the problem, exactly? If a mid-level fighter who's invested in Str and Jump can attempt a 30-foot long jump and has a chance to make it, but a mid-level fighter with high Str and no Jump ranks or a mid-level wizard with Jump ranks but not Str doesn't have a chance of making that DC, that simply bears out that idea that making a 30-foot long jump is hard and not something everyone can do.

Indigo Knight
2019-02-10, 06:11 AM
You can fit that from a story point of view, if you want to.
Fixes that originate from story reasons are, to me, very problematic. Suffice to say that I won't be using them.



So maybe a flat or very-slowly-scaling Veteran bonus to untrained or zero-rank skills, while the specialist keeps sinking points into hitting the increasingly difficult challenges? Septimus the 7th level wizard has a decent chance to sneak past a 1st level orc, but sneaking past Oculus the All Seeing Eye is something you need Whisperfoot the 7th level rogue to handle?

That's a sidetrack from the linear vs exponential question, but I think Veteran bonus (Flat +2? 1/3 level rounded down?) is an option to look at.

Is this like the proficiency bonus? Yeah, I guess it could be used.
I'm still wondering about that. My gut tells me that this is not the solution I'm looking for. With it being weird in the sense that it's open ended (the gap keeps expanding, it would never reach a fitting equilibrium) and that it takes the agency away from the players.
PairO'Dice suggested something like that. And I'm still testing it out.



Because maybe DCs over 30 are legendary-scale accomplishments. The high level thief CAN auto-succeed on picking a cheap lock or sneaking past a mook or fast-talking a random barmaid. Why keep investing? Because the DCs are higher to sneak past or fast-talk a dragon or unlock an arcane lock.
Theoretically yes. But consider this, while players gain xp and level up, so does their relative foes. Well, not so much level up as simply swapping with Bigger Foes™ (which would be the same as leveling up). Placing the characters who increased in rank at the same relative spot as before. In actuality, the change is that non-trained are now more behind then they were before (sometimes to the points they are no longer able to participate), and that mundane daily tasks are trivial after a certain number of levels. Even for untrained characters.



Basically the idea of Epic Level Handbook skills, but better implemented.
I'm open to suggestions.



The key question here is, where in the level scale would you say a real-world Olympic athlete falls?
Hey dice. Thanks for joining.

1st place? At max ranks + all additional bonuses (synergy, luck, equipment, aid another). And a 20 roll is the WR.



There's a good case to be made (e.g. here (https://thealexandrian.net/wordpress/587/roleplaying-games/dd-calibrating-your-expectations-2)) that people in the real-world would be represented by characters in the level 1-6 range. And of course we don't have, y'know, magic in the real world (though performance-enhancing drugs are arguably potion equivalents :smallwink:).
An assumption that I'll gladly take. As well as giving the reason for E6 ruleset. Only that I think the levels shouldn't be capped at 6. The performance of skills should.
I argue that the better you get at a subject, the harder it is to get better at it going forwards. And the slope simply peters out.
My main argument is that one needs to invest more to get the same delta of improvement.
Which this
"DC X to travel X feet," contradicts by being a linear progression.



In general, when working with a 3e-ish baseline (20 levels, ability scores in the normal range, expert progressions are around +1/level, etc.), you should never ever have DCs above 45 for anything you intend 20th-level characters to do with any sort of regularity. A character very dedicated to making a certain kind of roll will have around +20 from level, +13 from a maximized ability score (18 base, +2 racial, +5 level, +5 inherent, +6 enhancement), and +5ish miscellaneous (a +5 sword for attack rolls, +2 synergy +3 Skill Focus for skill checks, etc.), and can generally finagle a +2 circumstance bonus (high ground, Aid Another, a +2/+2 skill feat, Favored Enemy, whatever) most of the time, for a total modifier of +40; less-specialized characters, or characters rolling secondary attacks and skills, are going to have closer to +30 total (+20 level, +6ish ability score, +4ish miscellaneous+circumstance). There's a reason the big chart of sample skill DCs in the DMG tops out at DC 43 for an extremely niche task attempted by an extremely specialized character, and CR 20 monsters have ACs in the 35 to 40 range.

This is a massive problem with the system as it stands now. I think everyone can agree on it. It is so bad, that straight of the gate, at level 1, it breaks the game. The giant wrote about it. You wrote about it. I encountered it in my games. This cannot be overstated enough. There is a glaring problem with the way numeric skill bonuses behave.



It makes total sense that a high-level character can go from +25 to +30 with the same ease that a low-level character goes from +5 to +10, because the high-level character is more competent in every way than the low-level character and that includes being able to blow past world records that the low-level character struggles to meet.

Unfortunately I'm having a hard time understanding that exact item. My hunch is that linear progression isn't a valid representation of skill improvement.



If your main concern is that only a few specialized characters should be able to make high DCs and you reach a point where characters who haven't invested in a skill can't make those DCs...what's the problem, exactly? If a mid-level fighter who's invested in Str and Jump can attempt a 30-foot long jump and has a chance to make it, but a mid-level fighter with high Str and no Jump ranks or a mid-level wizard with Jump ranks but not Str doesn't have a chance of making that DC, that simply bears out that idea that making a 30-foot long jump is hard and not something everyone can do.
That's not really what I'm grappling with. On the opposite. I strive for ways where untrainies will always have a chance of success at untrained skills with max DC. Or, in other words, that no matter how high the trained characters get, that those without ranks will always be within a rolling chance (in this system it's the d20, so -19).

rferries
2019-02-10, 06:22 AM
If you want even untrained characters to have a chance of suceeding, you could apply the "natural 20 = automatic success" rule to skill checks.

MoiMagnus
2019-02-10, 07:54 AM
A system that I do like quite a lot (but is FAR from perfect), is "the more you are trained, the smaller are the dices you roll".
For this example, I will assume that "you want to make LESS than the DC to succeed" (so "1" is critical sucess).
An untrained character uses a d20.
And then, depending on skill level, it uses d12, d10, d8, d6, or d4.

This system works better if you own "weird dices", like a d14, d16 and d18
It works even better if you use computer-dices (because you can chose any number of face)

Note that it does not prevent you from having effects like "using tools increase the DC by 3"

PairO'Dice Lost
2019-02-10, 09:16 PM
I argue that the better you get at a subject, the harder it is to get better at it going forwards. And the slope simply peters out.
My main argument is that one needs to invest more to get the same delta of improvement.
Which this contradicts by being a linear progression.

Firstly, you're not making an argument that the better you get the more you need to invest to improve, you're making the assertion that that's the case; "diminishing returns are necessary" is the premise of your argument, not the outcome. Which is totally fine--you should decide on the axioms of your system based on what you want the outcomes to look like--but not all workable skill systems have to be set up like that.

Secondly, for most things in real life you actually don't see constant diminishing returns for most tasks. Someone who decides to, say, learn to play the piano is going to struggle a lot in the beginning to make even a little bit of progress; they don't know what any of the symbols mean, sheet music is this intimidating page of impenetrable notation, and so on. However, once they get the basics down, it's actually going to be easier for them to progress: they have their scales memorized, they can read sheet music passably well, they know what "sounds right" so they can guess at chords and such more easily, and so on. The better they get at playing piano, the more easily they can learn harder pieces, explore new composers, learn more theory given that they have plenty of practical experience, and so on. It's only when they start reaching the limits of their abilities, when they reach the "professional concert pianist" level and performance success or failure hinges on a single half-second hesitation on a single note, that it gets much harder to improve.

Same with athletics (it's hard to start running or jumping on a track team, easier when you figure out the right form and start conditioning your body, hard again when you're eking out every last
inch of distance against other Olympians) and other such tasks. A graph of investment would look much less logarithmic, like constant diminishing returns, and much more S-shaped or logistic. (And of course when your high-end limits are "high-level D&D character" and not "real-life expert," the same patterns would hold but difficult tasks are things like "leaping 100 feet in the air to grab a swooping dragon" and breaking Olympic records would still fall in the easy-self-improvement bracket, but in this particular case it sounds like you're still concerned with E6 so that's less of an issue.)


This is a massive problem with the system as it stands now. I think everyone can agree on it. It is so bad, that straight of the gate, at level 1, it breaks the game. The giant wrote about it. You wrote about it. I encountered it in my games. This cannot be overstated enough. There is a glaring problem with the way numeric skill bonuses behave.

Bolding mine. I don't recall saying either that the current bonus progressions break the game at level 1, or that it's a particularly glaring problem overall. There's a big difference between "At 20th level, the breadth and cheapness of available bonuses tend to cause problems with setting fair DCs for various tasks" and "The skill system has a major fundamental problem."

I assume that by "at level 1 it breaks the game" you're referring to the fact that some characters might have +8 skill bonuses (4 ranks, 18 stat) while others have +0 (no ranks, 10 stat), plus or minus dump stats and Skill Focus and such. But a variance of 8 points among a party is well within the 20-point range of a d20, and against the low-DC stuff you expect to face at 1st level, that's pretty much fine. I think the issue here is that you're focused on the absolute highest DCs while also assuming that everyone in the party has to make those highest DCs, but (A) you rarely run into the absolute highest DCs, at least if the DM is following the usual rules on CR and NPC building and so forth, and (B) "highest DCs" and "everyone in the party has to make them" don't usually coincide.

Usually, only one person in the party has to make a given task, and it's usually the party expert (the rogue using Open Lock against a door lock, not the fighter; the fighter Jumping across a gap to tie a rope, not the wizard), in which case the weaker characters' modifiers are irrelevant for determining the challenge level. In cases when the whole party does have to roll (like sneaking past a sentry, where the lowest roll is what matters because even one person getting spotted will mean the party is discovered), then the gap between party members doesn't matter, it's the gap between the lowest party modifier and the enemy modifier that matters, so the DCs should be set based on that.


That's not really what I'm grappling with. On the opposite. I strive for ways where untrainies will always have a chance of success at untrained skills with max DC. Or, in other words, that no matter how high the trained characters get, that those without ranks will always be within a rolling chance (in this system it's the d20, so -19).

Why?

Not that there's anything wrong with that goal, but what specifically are you trying to accomplish by setting that goal?

Is it to ensure that people with +0 modifiers are willing to try rolling for any task instead of waiting for the expert to do it? Well, it doesn't tend to work out that way in play regardless; the party is going to tell the rogue to try Open Lock checks and the wizard to try Knowledge (Arcana) checks and only roll if they fail, because the expert is always perceived as having the best chance (except in e.g. low-level 5e where the "expert" is maybe +2 above the weakest party member and the outcome is basically random).

Is it to encourage people with +0 modifiers to try things as long as they have a chance of success? Well, "a chance" doesn't matter, most players won't try things if there's a large probability of failure; sure, if an un-sneaky character has a 20% probability to sneak past a sentry instead of 0% chance that's nice, but if there's 3 sentries and 4 un-sneaky PCs they're not going to risk it and just go around.

You've stated your desired outcomes, but they're somewhat vague in application and unintuitive compared to the benchmarks set by baseline d20. Without knowing your guiding principles, it's hard to figure out how to massage the numbers to arrive at those goals.

Indigo Knight
2019-02-12, 11:10 AM
If you want even untrained characters to have a chance of suceeding, you could apply the "natural 20 = automatic success" rule to skill checks.
That is one way. Sure. It simply place the chances of the untrained at exactly 5% to whatever he is trying to do. Do you think that's a good fit?



A system that I do like quite a lot (but is FAR from perfect), is "the more you are trained, the smaller are the dices you roll".
For this example, I will assume that "you want to make LESS than the DC to succeed" (so "1" is critical sucess).
An untrained character uses a d20.
And then, depending on skill level, it uses d12, d10, d8, d6, or d4.

This system works better if you own "weird dices", like a d14, d16 and d18
It works even better if you use computer-dices (because you can chose any number of face)

Note that it does not prevent you from having effects like "using tools increase the DC by 3"
This has the glaring drawback of inflating the proportional sway of numeric bonuses (or penalties) the smaller die you go. a +2 for racial advantage is a whopping 25% for the d8.



Firstly, you're not making an argument ... you're making the assertion that that's the case; ... premise of your argument, not the outcome.
You'll have to teach me the difference of the two because I don't understand.



not all workable skill systems have to be set up like that.
I don't know what you define as workable skill. Overlooking that, this is mainly regarding skills that can be employed by untrained (so no spellcraft, etc.) and are measured against static difficulty ( no sense motive). Also, it is highly theoretical.



Secondly, for most things in real life you actually don't see constant diminishing returns for most tasks. Someone who decides to, say, learn to play the piano is going to struggle a lot in the beginning to make even a little bit of progress; they don't know what any of the symbols mean, sheet music is this intimidating page of impenetrable notation, and so on. However, once they get the basics down, it's actually going to be easier for them to progress: they have their scales memorized, they can read sheet music passably well, they know what "sounds right" so they can guess at chords and such more easily, and so on. The better they get at playing piano, the more easily they can learn harder pieces, explore new composers, learn more theory given that they have plenty of practical experience, and so on. It's only when they start reaching the limits of their abilities, when they reach the "professional concert pianist" level and performance success or failure hinges on a single half-second hesitation on a single note, that it gets much harder to improve.
An excellent observation. You are (ofcourse) right. What would you say is the mathematical representation? Would you be able to chart an estimated graph? Would it be fun to simulate that in the game?



I assume that by "at level 1 it breaks the game" you're referring to the fact that some characters might have +8 skill bonuses (4 ranks, 18 stat) while others have +0 (no ranks, 10 stat), plus or minus dump stats and Skill Focus and such.
From experience, bigger than that. You left out bonus from tools, synergy, racial bonus, and on.
I consider that as a huge issue, that breaks gametime more than once or twice.



I think the issue here is that you're focused on the absolute highest DCs while also assuming that everyone in the party has to make those highest DCs, but (A) you rarely run into the absolute highest DCs, at least if the DM is following the usual rules on CR and NPC building and so forth, and (B) "highest DCs" and "everyone in the party has to make them" don't usually coincide.

Usually, only one person in the party has to make a given task, and it's usually the party expert (the rogue using Open Lock against a door lock, not the fighter; the fighter Jumping across a gap to tie a rope, not the wizard), in which case the weaker characters' modifiers are irrelevant for determining the challenge level. In cases when the whole party does have to roll (like sneaking past a sentry, where the lowest roll is what matters because even one person getting spotted will mean the party is discovered), then the gap between party members doesn't matter, it's the gap between the lowest party modifier and the enemy modifier that matters, so the DCs should be set based on that.
There are a lot of terms that get mixed here.
For starter, open lock is a no-untrained usable. So it's not something I'm referring to. Differentiate between the two types.
Second, only one character is usually tasked with rolling the skill. In that case, It doesn't really matter the relativity between characters skill caps and bonuses. It's also doesn't matter the number in the DC – only the chance to pass it, the percentage to succeed.
But! There are times where all players must roll and not all of them have invested in the skill. Perception, sneak, diplomacy, sense motive and even reflex saves. The instance of having two characters with a gap of more than the die roll is an issue. It means that one of them doesn't get to 'Play'. Either auto-failing at the task, or automatically succeeding, rendering the others efforts mute. I get that some people out there doesn't see this as an obstacle – but I do.



Why?
Because agency matter.
Player involvement is the most crucial item to the game. And (at least for me) engaging them is the first priority.
Most of the rules in the books have a footnote of "GM's discretion". Why? Because almost every item can be waived in or waived off according to the context the GM is creating. Before the dice even come to action, the GM decides how hard it would be, and most of the time the numbers change according to level, story, setting and other GM's 'constrains'.
Go look in the other thread about crafting skill and see that poster's answers mostly go by GM's hand waiving.
The definite DC of tasks doesn't matter. What matters is the relative percentage between the players around the table, and the story. Which means that probability of the dice is a bit important for both low-end and high-end.



You've stated your desired outcomes, but they're somewhat vague in application and unintuitive compared to the benchmarks set by baseline d20. Without knowing your guiding principles, it's hard to figure out how to massage the numbers to arrive at those goals.
True. That's because I'm mostly theorizing and not really looking at a set goal in mind.

rferries
2019-02-13, 02:13 AM
That is one way. Sure. It simply place the chances of the untrained at exactly 5% to whatever he is trying to do. Do you think that's a good fit?

It's as good as fit as the natural 20/ natural 1 rules are for the rest of 3.5. At some point verisimilitude is less important than fun. :)

johnbragg
2019-02-13, 09:00 AM
Secondly, for most things in real life you actually don't see constant diminishing returns for most tasks.


I'm responding to the idea PairO'Dice was responding to, that the 3X skill progression doesn't model the fact that it's harder to advance the higher up the skill curve you go. I'd like to point out that 3X *does* model this, it's just built into the XP tables rather than the skill ranks chart. (I'm sorry if I'm repeating myself, I swear I typed all this out but don't see the post).


Someone who decides to, say, learn to play the piano is going to struggle a lot in the beginning to make even a little bit of progress; they don't know what any of the symbols mean, sheet music is this intimidating page of impenetrable notation, and so on. However, once they get the basics down, it's actually going to be easier for them to progress: they have their scales memorized, they can read sheet music passably well, they know what "sounds right" so they can guess at chords and such more easily, and so on. The better they get at playing piano, the more easily they can learn harder pieces, explore new composers, learn more theory given that they have plenty of practical experience, and so on.

Hello, Pathfinder Class Skill bonus? (Perhaps tweaked to add Cultural/Racial Skill bonus for a few skills your learned the basics of as a child?) Actually I forget if the PF Class Skill bonus applies to untrained skills.


It's only when they start reaching the limits of their abilities, when they reach the "professional concert pianist" level and performance success or failure hinges on a single half-second hesitation on a single note, that it gets much harder to improve.

Which is why, in 3.5 it takes 1000 XP to add a 5th Skill Rank to a maxxed skill, 2000 to add a 6th Skill Rank, 3000 to add a 7th, 4000 to add an 8th, etc.


Bolding mine. I don't recall saying either that the current bonus progressions break the game at level 1, or that it's a particularly glaring problem overall. There's a big difference between "At 20th level, the breadth and cheapness of available bonuses tend to cause problems with setting fair DCs for various tasks" and "The skill system has a major fundamental problem."

When you say that I tend to hear "20th level D&D is borked in many different ways"


I assume that by "at level 1 it breaks the game" you're referring to the fact that some characters might have +8 skill bonuses (4 ranks, 18 stat) while others have +0 (no ranks, 10 stat), plus or minus dump stats and Skill Focus and such. But a variance of 8 points among a party is well within the 20-point range of a d20, and against the low-DC stuff you expect to face at 1st level, that's pretty much fine. I think the issue here is that you're focused on the absolute highest DCs while also assuming that everyone in the party has to make those highest DCs, but (A) you rarely run into the absolute highest DCs, at least if the DM is following the usual rules on CR and NPC building and so forth, and (B) "highest DCs" and "everyone in the party has to make them" don't usually coincide.

Yes. The highest DC stuff is the spotlight moment for the characters that spent character building resources on those tasks. What's the point of playing a high-Charisma Bard party face with maxxed Diplomacy, Bluff, etc if the Big Stupid Fighter who, in character, doesn't use the wrong fork because he eats with his hands, can Diplomacy just about as well as the Bard?



You've stated your desired outcomes, but they're somewhat vague in application and unintuitive compared to the benchmarks set by baseline d20. Without knowing your guiding principles, it's hard to figure out how to massage the numbers to arrive at those goals.

To be fair to OP, it's hard. If it were easy, WOTC would have done it in 1999 when developing 3.0



This has the glaring drawback of inflating the proportional sway of numeric bonuses (or penalties) the smaller die you go. a +2 for racial advantage is a whopping 25% for the d8.


I have to disagree here--if you use that sort of system, changing dice IS the bonus. So a +2 racial bonus would translate to going up from, say, a d6 to a d10. (Or down, if it's roll-under-target)

PairO'Dice Lost
2019-02-14, 03:32 AM
You'll have to teach me the difference of the two because I don't understand.

Sure. When you say "My argument is X," it implies that you're starting from some other premise (like the core 3e skill system), making observations about its outcomes, and stating things that can be proven or disproven; someone disagreeing with that means they accept the same premises but think that, as a matter of fact, you're mistaken about something. When you say "My premise is X," you're talking about the assumptions you're making or want to make, and then you can make arguments from there; someone disagreeing with you means they think that, as a matter of opinion, your premises don't lead to the desired outcome or should otherwise be changed.


An excellent observation. You are (ofcourse) right. What would you say is the mathematical representation? Would you be able to chart an estimated graph? Would it be fun to simulate that in the game?

A logistic curve looks like this (https://en.wikipedia.org/wiki/Logistic_function#/media/File:Logistic-curve.svg), starting off slowly, then increasing in slope (rate of increase), then leveling back off again.

As far as whether it would be fun to simulate, eh, maybe. It would lead to more involved skill investment and would require careful mathing-out to get right. On the other hand, most games top out at mid levels so making skills get easier to improve as you approach the 10th-12th level mark so you seem to get vastly better at your specialties would be exciting, and slowing things down a lot once you pass that would help rein in high-level bonuses, so it might lead to more player enjoyment overall. It really depends on the implementation details in this case.


From experience, bigger than that. You left out bonus from tools, synergy, racial bonus, and on.
I consider that as a huge issue, that breaks gametime more than once or twice.

That's covered in the "plus or minus dump stats and Skill Focus and such" part I mentioned. Yes, you certainly can get a couple other +2s (though not synergy bonuses, actually, since those don't come in until 2nd level) to widen the gap, but looking only at edge cases and worst-case scenarios isn't helpful since most of the time the range will be smaller. Sure, someone can theoretically have +16 (4 ranks, +5 stat, +2 racial, +2 tool, +3 Skill Focus) against -1 (0 ranks, -1 stat), but most of the time you're only going to have +1 to +3 stat bonus because it's a secondary or tertiary ability score, or only 2 ranks 'cause it's a cross-class skill, and no miscellaneous stuff because they're not a half-elf rogue with bunches of bonuses...and on the other side, when most opposed checks come up, you're generally going to have opposition with more than +0 stat, more than 0 ranks, and so on.

So yes, +16 vs. -1 is possible when an extremely focused expert faces the easiest possible opposition, but if most of the time you're looking at +6 (4 ranks, +2 stat) vs. +2 (0 ranks, +2 stat) or +8 vs. +4 (same as before, but each has a racial bonus) or the like, the game is hardly broken.


I'm going to step back here and point out that when people criticize 3e it's common to conflate broken (completely nonfunctional and game-ruining by default) with breakable (if you try to abuse things, you can push things toward an edge case and make it break). Bonus progressions in general and the skill system in particular are merely breakable: you can definitely find degenerate cases and build a character to crush a particular kind of check, but most checks work mostly fine most of the time, so it makes sense to focus on changing the stuff that pushes you out of "most" territory as much as or more than the core mechanics.

You think synergy and racial bonuses break things at low levels? Fine, ditch 'em; they're not a holdover from AD&D or anything, no one will miss them if they go except the people trying to stack up all the +2s to break stuff. Change the racial bonuses to "2 free ranks" instead of "+2 to checks" and don't replace synergy bonuses (they're mostly fiddly accounting anyway), for instance. Masterwork skill kits? Again, nothing is forcing you to keep them, or keep the bonuses they grant; maybe masterwork thieves' tools let you take 15 instead of taking 10, or let you reroll a failed check, or are just easier to hide on your person with Sleight of Hand.


But! There are times where all players must roll and not all of them have invested in the skill. Perception, sneak, diplomacy, sense motive and even reflex saves. The instance of having two characters with a gap of more than the die roll is an issue. It means that one of them doesn't get to 'Play'. Either auto-failing at the task, or automatically succeeding, rendering the others efforts mute. I get that some people out there doesn't see this as an obstacle – but I do.

That's the second case I mentioned: "the gap between party members doesn't matter, it's the gap between the lowest party modifier and the enemy modifier that matters, so the DCs should be set based on that."

Let's say you're setting DCs for sneaking past sentries and your average 10th level rogue has +20 to Hide. That's cool, buuut we don't care about the rogue! We're not going to assume average sentries have +20 Spot because (A) it doesn't make sense for every Joe Schmoe to be hyper-focused on spotting just because they're assigned to sentry duty, (B) that implies the average enemy has expert-equivalent modifiers in lots of different skills if you're talking about "average" opposition, and (C) it's the non-expert PC whose modifiers we care about more.

So let's arbitrarily set the sentry's Spot at +5 and the fighter's Hide at -2 (shame about the clanky plate armor, huh?). We end up in a situation where there's a greater-than-20-point difference between the rogue's Hide (+20) and the fighter's Hide (-2), but they're not rolling against each other so that's irrelevant, it's only the 7-point spread between the sentry's +7 and the fighter's -2 that matters for balancing purposes. And it doesn't matter if the rogue auto-succeeds on his own Hide check, because that just means he remains hidden, the rest of the party could still be discovered; it would only matter if it were the rogue doing the Spotting rather than the Hiding, because then a single success would indeed mean a "party success" in spotting enemies.

Essentially, you can't just talk about "auto-success" and "difference in modifiers" as a general rule, because different modifiers matter for different characters, and different success mean different things on different checks.


To be fair to OP, it's hard. If it were easy, WOTC would have done it in 1999 when developing 3.0

To be fair to WotC, they successfully achieved exactly what they set out to do. It just so happens that what they set out to do was "Make an edition where you can play 2e with vastly improved mechanics" (which 3e does beautifully) while what players set out to do was "Take a look at 3e without AD&D assumptions and smash the weak points with a +3 warhammer of RAW until they break" (which...not so much). :smallamused:

johnbragg
2019-02-14, 11:20 AM
To be fair to WotC, they successfully achieved exactly what they set out to do. It just so happens that what they set out to do was "Make an edition where you can play 2e with vastly improved mechanics" (which 3e does beautifully) while what players set out to do was "Take a look at 3e without AD&D assumptions and smash the weak points with a +3 warhammer of RAW until they break" (which...not so much). :smallamused:

Yes. We often forget that when 3E was developed, the internet was mostly dialup, CharOp forums were mostly limited to Usenet, searching-by-keyword was a crude tool at best, and nobody understood the implications of making magic-mart a full scale, easy to understand and operate thing by creating an explicit economy that traded GP for spell slots (single use, wands, per day, permanent).

Maat Mons
2019-02-14, 06:44 PM
I'd like to point out that none of the problems being discussed are specific to sills. Rather, they're problems of numeric advancement. Numeric advancement is when characters have number, and those numbers get bigger as the characters "level up," or whatever.

Here's the thing though. Do different characters advance a particular number at different rates? If so, in a system like d20, the difference between characters will eventually grow to the point that the non-specialist eventually has a 0% chance of beating the specialist. Or, if there's a hard cap one success/failure chance (e.g. "1 always fails, 20 always succeeds"), the cap will be reached, and further differences will become meaningless.

Alternately, does everyone always advance at the same rate? Let's say, some fixed base value, which differs between characters, plus 1/2 level? Well, that means if you take a pair of characters at 1st level and at 20th level, the fight between those characters has all the same odds at 20th level as it did at 1st level. So why are we even changing the numbers.

I can think of two possible solutions to this pair of problems. One is that, at some point, numerical progression caps out. And the cap is selected to occur before things get off the RNG.

The second possible solution is my dandy-dandy Sigmoid Dice System™ (patent pending). Under this system, rolling more dice is sometimes a good thing, and sometimes a bad thing. As your odds of success approach 100%, your roll more and more dice, and succeed as long as at least 1 of them is a success. As your odds of success approach 0%, you also roll more and more dice, but now you fail as if even one die shows a failure. Your odds of success can never reach either 0% or 100%, but there's always a step closer to either.

PairO'Dice Lost
2019-02-14, 07:00 PM
Here's the thing though. Do different characters advance a particular number at different rates? If so, in a system like d20, the difference between characters will eventually grow to the point that the non-specialist eventually has a 0% chance of beating the specialist. Or, if there's a hard cap one success/failure chance (e.g. "1 always fails, 20 always succeeds"), the cap will be reached, and further differences will become meaningless.
[...]
I can think of two possible solutions to this pair of problems. One is that, at some point, numerical progression caps out. And the cap is selected to occur before things get off the RNG.

People talk about inevitable divergence a lot regarding d20 advancement, but that's not necessarily the case; as long as things don't diverge by 20th level (or whatever other level cap you set), you're fine.

Standard 3e save modifiers, for instance, cap out at +25ish for good saves (+12 base, +13ish for a 36ish primary attribute) and +6ish for poor saves (+6 base, +0ish for tertiary or quaternary attributes), meaning it gets close to but never actually diverges completely, even in the extreme case, and since most characters don't jack their Dex/Con/Wis up to that extent in practice the difference is smaller. Save DCs against good saves are the same way, capping out at 32ish (10 plus +9 for a 9th level spell plus +13ish for a 36ish stat) while a good save with a tertiary attribute ends up at +12, once again still baaarely within the 20-point range and much closer together in the average case.

So it's entirely possible to come up with a progression that ends up with a smaller maximum range at the maximum level, either by tweaking the existing ones slightly (e.g. if you remove the assumption that anyone who wants to can get a +5 inherent bonus and +6 enhancement bonus to their primary attribute, stat bonuses top out at +7ish instead and the maximum range is much reduced) or by coming up with new progressions (e.g. DCs are now 10 + 1/3 CL + stat, poor saves are now +1 plus +1/3 level, bam, the range shrinks by 4 points), giving you the benefits of progression without pushing people off the RNG in the most extreme cases or requiring an arbitrary cap that people will quickly bump up against.

Indigo Knight
2019-02-19, 10:18 AM
It's as good as fit as the natural 20/ natural 1 rules are for the rest of 3.5. At some point verisimilitude is less important than fun. :)
I did use crit/fumble for the longest time. I don't know for sure if the game benefit or not from that. Either way – I think that it's possible to get a smaller chance than 5% with d20 mathematically. Now how hard it would be to simulate that is something else.



I'm responding to the idea PairO'Dice was responding to, that the 3X skill progression doesn't model the fact that it's harder to advance the higher up the skill curve you go. I'd like to point out that 3X *does* model this, it's just built into the XP tables rather than the skill ranks chart. (I'm sorry if I'm repeating myself, I swear I typed all this out but don't see the post).
I think you wrote that in the other thread.
And as well here as there, I actually don't think that it represents that accordingly because both rank 1 in climb and rank 16 in heal is the same "cost".



Hello, Pathfinder Class Skill bonus? (Perhaps tweaked to add Cultural/Racial Skill bonus for a few skills your learned the basics of as a child?) Actually I forget if the PF Class Skill bonus applies to untrained skills.
Citation?



Yes. The highest DC stuff is the spotlight moment for the characters that spent character building resources on those tasks. What's the point of playing a high-Charisma Bard party face with maxxed Diplomacy, Bluff, etc if the Big Stupid Fighter who, in character, doesn't use the wrong fork because he eats with his hands, can Diplomacy just about as well as the Bard?
I think he means high end level rather then high end skill.


[/QUOTE]To be fair to OP, it's hard. If it were easy, WOTC would have done it in 1999 when developing 3.0[/QUOTE]
Thanks. To his defense, though, I really don't have a ready-made set goal in mind.


[/QUOTE]I have to disagree here--if you use that sort of system, changing dice IS the bonus. So a +2 racial bonus would translate to going up from, say, a d6 to a d10. (Or down, if it's roll-under-target)[/QUOTE]
So you mean, abolish circumstantial bonuses altogether? Or what?


That's covered in the "plus or minus dump stats and Skill Focus and such" part I mentioned. Yes, you certainly can get a couple other +2s (though not synergy bonuses, actually, since those don't come in until 2nd level) to widen the gap, but looking only at edge cases and worst-case scenarios isn't helpful since most of the time the range will be smaller. Sure, someone can theoretically have +16 (4 ranks, +5 stat, +2 racial, +2 tool, +3 Skill Focus) against -1 (0 ranks, -1 stat), but most of the time you're only going to have +1 to +3 stat bonus because it's a secondary or tertiary ability score, or only 2 ranks 'cause it's a cross-class skill, and no miscellaneous stuff because they're not a half-elf rogue with bunches of bonuses...and on the other side, when most opposed checks come up, you're generally going to have opposition with more than +0 stat, more than 0 ranks, and so on.

So yes, +16 vs. -1 is possible when an extremely focused expert faces the easiest possible opposition, but if most of the time you're looking at +6 (4 ranks, +2 stat) vs. +2 (0 ranks, +2 stat) or +8 vs. +4 (same as before, but each has a racial bonus) or the like, the game is hardly broken.
Look, I'm mostly bringing example out of my own experience. And during my years DMing - most groups have this one player that is going to be on top of the numbers trying to maximize his abilities, aka min/max, aka optimizer, aka yougetit. (side note, the DMGII is fantastic regarding those. See player's incentive right at the start). Just this one guy is enough. Suddenly I'm confronted with this gap between the optimizer and his teammates. I get complains from the others that they can't do anything (and failing to realize the mathematical reasoning about why this is the case). One of them tries to act a skill he has no rank in (Swim), fails, and get frustrated. Another, wants to do this cool act, which he of course have no ranks in, and then goes back to begrudgingly perform as a bard, muttering that his class suck and he can only do this one thing. Meanwhile, this optimizer guy, sticks to what are his biggest numbers. Need to take the target hostage? Sorry, no go, he didn't build grapple build. He'll just quadruple crit the guy with chi-shadow-damage and drag his dead buddy. Need to pass bluff check against a routine guard patrol? Doesn't even bother. Smilbidrillion damage to the guard, to his companion, to his superior, to his underlings, to his nana, and another one just for good luck. Because that's where he invested his points. Need to go talk to the barkeep of ANY AVAILABLE BAR to hear the gossip? Won't even bother. Catch the nearest urchin. Smack him around. Continue the plot.
Now, putting aside the subjects of bettering myself as a master, bettering the participators as players and generally communicating better while gaming, put all this to the side; this is a thing that happens. I experience this situation. So with all due respect, to you and others, trying to tell me that "that is rare", "this is an extreme case", "that most of the time I'm going to have something else", "skills aren't used that way" etc.' - this is the starting point. This is what I have at the table. There's this one guy (sometimes it's the other guy) that's going to stretch the skills because he's able to (and also because it makes sense numerically). Two points that exist at wider gap then the dice rendering one of them useless/infallible. And this is what I'm trying to respond to. I observe the skill ranks functions as flawed and I wish to address them. Now stop telling me that it's not the case and work with me.
Thank you.
Sorry.
Thanks




I'm going to step back here and point out that when people criticize 3e it's common to conflate broken (completely nonfunctional and game-ruining by default) with breakable (if you try to abuse things, you can push things toward an edge case and make it break).
Point taken.
Breakable.



You think synergy and racial bonuses break things at low levels? Fine, ditch 'em;
Already done.
Though, I placed a cap on bonus to skill use. But this are mostly small chips.



That's the second case I mentioned: "the gap between party members doesn't matter, it's the gap between the lowest party modifier and the enemy modifier that matters, so the DCs should be set based on that."
For victory? Yes.
For player involvement? No.




I'd like to point out that none of the problems being discussed are specific to sills. Rather, they're problems of numeric advancement. Numeric advancement is when characters have number, and those numbers get bigger as the characters "level up," or whatever.

Here's the thing though. Do different characters advance a particular number at different rates? If so, in a system like d20, the difference between characters will eventually grow to the point that the non-specialist eventually has a 0% chance of beating the specialist. Or, if there's a hard cap one success/failure chance (e.g. "1 always fails, 20 always succeeds"), the cap will be reached, and further differences will become meaningless.
True.
If we deal with linear functions, and those functions aren't parallel, then they will always create a gap somewhere. It's just a question of when.



I can think of two possible solutions to this pair of problems. One is that, at some point, numerical progression caps out. And the cap is selected to occur before things get off the RNG.
Yeah. That is the easy solution. And if nothing better is made up then that would be the answer.


The second possible solution is my dandy-dandy Sigmoid Dice System™ (patent pending). Under this system, rolling more dice is sometimes a good thing, and sometimes a bad thing. As your odds of success approach 100%, your roll more and more dice, and succeed as long as at least 1 of them is a success. As your odds of success approach 0%, you also roll more and more dice, but now you fail as if even one die shows a failure. Your odds of success can never reach either 0% or 100%, but there's always a step closer to either.
Resembles the same conclusion I have.
I'll try to portray what I'm thinking:



The new sigmoid skill system.






Rank Number
Dice Roll


0
3d20, pick lowest number


1
2d20, pick lowest number


2
1d20


3
2d20, pick highest number


4
3d20, pick highest number


5
4d20, pick highest number


6
5d20, pick highest number


7
6d20, pick highest number




You can the probability spread in this graph (https://anydice.com/program/13ab3).


This could be (or not) coupled with increasing rank cost:

cost 1 point.
cost 2 points.
cost 3 points.
cost 4 points.
cost 5 points.
cost 6 points.
cost 7 points.





Thoughts?

PairO'Dice Lost
2019-02-19, 09:24 PM
Look, I'm mostly bringing example out of my own experience. And during my years DMing - most groups have this one player that is going to be on top of the numbers trying to maximize his abilities, aka min/max, aka optimizer, aka yougetit. (side note, the DMGII is fantastic regarding those. See player's incentive right at the start). Just this one guy is enough.

Suddenly I'm confronted with this gap between the optimizer and his teammates. I get complains from the others that they can't do anything (and failing to realize the mathematical reasoning about why this is the case). One of them tries to act a skill he has no rank in (Swim), fails, and get frustrated. Another, wants to do this cool act, which he of course have no ranks in, and then goes back to begrudgingly perform as a bard, muttering that his class suck and he can only do this one thing.

Meanwhile, this optimizer guy, sticks to what are his biggest numbers. Need to take the target hostage? Sorry, no go, he didn't build grapple build. He'll just quadruple crit the guy with chi-shadow-damage and drag his dead buddy. Need to pass bluff check against a routine guard patrol? Doesn't even bother. Smilbidrillion damage to the guard, to his companion, to his superior, to his underlings, to his nana, and another one just for good luck. Because that's where he invested his points. Need to go talk to the barkeep of ANY AVAILABLE BAR to hear the gossip? Won't even bother. Catch the nearest urchin. Smack him around. Continue the plot.

Now, putting aside the subjects of bettering myself as a master, bettering the participators as players and generally communicating better while gaming, put all this to the side; this is a thing that happens. I experience this situation. So with all due respect, to you and others, trying to tell me that "that is rare", "this is an extreme case", "that most of the time I'm going to have something else", "skills aren't used that way" etc.' - this is the starting point. This is what I have at the table. There's this one guy (sometimes it's the other guy) that's going to stretch the skills because he's able to (and also because it makes sense numerically). Two points that exist at wider gap then the dice rendering one of them useless/infallible.

And this is what I'm trying to respond to. I observe the skill ranks functions as flawed and I wish to address them. Now stop telling me that it's not the case and work with me.
Thank you.
Sorry.
Thanks

Ah, see, now this is a hippogriff of a very different color. The thread was originally framed in terms of theory and mathing things out, when you said "I think that this is mostly due to wrong mathematical implementation. Or at least wrong attempt of presentation of different cases." and started talking about progressions and functions. That's why you're getting discussion and feedback about the math side of things, and expectations and general system performance and such.

But it turns out that the actual problem you're trying to address is that you tend to run games for a lot of people who don't know the rules and who try to do fun/inventive/off-the-wall stuff all the time and get frustrated when things don't turn out as expected, and a lot of other people who powergame and don't care about the plot and just want to kill all the NPCs to show off how ub3r l33t they are, and these two types of people end up in the same group and have totally incompatible expectations and playstyles.

Basically, it's like if you had a group where every single arcane spellcaster only ever casts basic blasting spells with low save DCs and is getting frustrated that they can't kill anything, so you post a thread wondering how you can improve the sorcerer and wizard to be on par with the warmage, and everyone is talking about the tier system and good school specialization and battlefield control tactics and such, and you're getting frustrated that they aren't talking about wizards throwing fireballs and they're getting frustrated that you're claiming warmages are more powerful when they're factually weaker and both sides are talking past each other.


So, some observations:

1) Someone who tries to do one cool thing, fails, and then complains that their class is a sucky one-trick-pony doesn't understand the rules and their outcomes, and instituting houserules to help them out isn't going to make them understand the rules better; it's still likely that they'll try something else the rules don't support and get upset about that not working. Powergamers try to abuse the rules by definition (not to be confused with optimizers or min/maxers, but that's a whole 'nother topic), and instituting houserules to limit them isn't going to make them play nice with the rest of the party; no matter what kinds of houserules you implement, a powergamer is going to try to abuse those houserules.

You cannot fix playstyle incompatibilities with rules, unfortunately, so understand that changing up the skill system isn't going to magically fix the group dynamic problems, and this is something that would best be addressed out-of-game in conjunction with whatever houserules you decide to implement.

2) The issue of untrained/low-modifier characters seems to be less "I want them to still be somewhere on the d20 compared to an optimized build to avoid auto-success and auto-failure" and more "I want people to have fairly good chances of success for untrained tasks at all levels so PCs with 0 ranks can try cool stuff and succeed at it a reasonable portion of the time." That part can't be fixed with different progressions or long-term skill investment, because the issue crops up at low levels long before the high-level progression divergences.

Three good ways to address this come to mind, which can be used together or separately:
Have skill "progressions" start with (or consist entirely of) big up-front bonuses that outweigh skill ranks for a few levels, something in the +3 to +5 range like PF and SWSE do, and then encourage players to scatter skill points around instead of maxing out a handful of skills. This won't solve the problem entirely, but it certainly helps when you can put a single rank in something and suddenly be moderately good at it.
Implement a good "stunting" mechanic, that is, giving out hefty bonuses when players try something that (A) is fun and cool, (B) they describe well/flavorfully in-game, and (C) the character isn't already good at. That gives Mr. Frustrated an incentive and a reward for trying cool things and helps ensure that they succeed, without allowing Mr. Powergamer to stack that bonus on top of an already-high modifier.
Let PCs save skill ranks and invest them as needed instead of investing everything at level-up, within whatever limits you feel are fair. If a PC can say "I wanna swim over to that underwater statue and do a thing! *invests 5 of his 30 saved skill points into Swim* Er, I totally took swimming lessons last summer, why do you ask?" then Mr. Powergamer can meticulously invest his ranks as desired and Mr. Frustrated can throw a bunch of skill points at a daunting task and they both come out happy.
3) The issue of optimized/high-modifier characters seems to be less "I want to institute diminishing returns and higher DCs because I think it's logical" and more "My players are powergaming one or two modifiers and using it on everything, which is boring and frustrating." This is a bit trickier to address, because the example you gave was optimizing combat stats, not skills, and the skills that can actually break games--Bluff, Diplomacy, Hide, and UMD, mostly; no one's breaking games with Decipher Script and Use Rope--break things even without automatic success because the things they do are powerful whether you can do them 100% of the time or just 20% of the time. To give specific advice here, I'd have to see some examples of the kinds of skill-related breakage you're seeing

But one thing to keep in mind in the meantime is that not all skills have to scale indefinitely. It's entirely within your rights to rule that you can only ever have 8 ranks in a skill because skills replicate real-world capabilities and "real-world" tops out at 5th, or declare that regardless of how you stack bonuses your modifier is hard-capped at +10, or something like that; it's like doing E6, but just for skills.


Does that all help address the actual underlying issues?

johnbragg
2019-02-19, 10:00 PM
I have to disagree here--if you use that sort of system, changing dice IS the bonus. So a +2 racial bonus would translate to going up from, say, a d6 to a d10. (Or down, if it's roll-under-target)
So you mean, abolish circumstantial bonuses altogether? Or what?

Not abolishing circumstantial bonuses. Abolishing +X bonuses. Instead of a +2, a bonus means you change dice by one increment. STacking bonuses means to change increments multiple times.

It's a poor fit with d20 systems, but if you DID try to force a d20ish system to work with it, using 3d6 instead of d20, I guess each +1 would translate into increasing a d6 to a d8, and then a d8 to a d10, etc. So a BAB +5 fighter with 18 strength (+4), a +1 weapon, and a +1 morale bonus from bardsong (total of +11). Upgrade 3d6 to 3d8 (3 steps), 3d8 to 3d10 (6 steps), 3d10 to 3d12 (9 steps), then upgrade 2 of the d12 to 2d6 each, so d12+4d6. Or 3d14 with the right dice-rolling app.

That is, of course, an awful mess to deal with, because it's a bastardized system forced to do things it's not supposed to.

It would work better with a "roll under target" system, where an Expert (+1) Fighter(+1) with a magic sword (+1) attacks a PC wizard (AC 10). From a base d10, upgrade to d8 for magic sword, d6 for Fighter, d4 for Expert. If the target number is 10, it doesn't really matter--but if the target number is 5, those bonuses are huge.

Indigo Knight
2019-02-26, 09:14 AM
Does that all help address the actual underlying issues?

eh... no.



This thread is about theory and mathematics. What I said before fits with my mathematical understanding of the rules. What I observed personally, and I'd like to discuss those. You can feel free to add your own examples, ofcourse.
I'm not going to change anything for the players I play with. Whether they understand the rules or not. I was pointing at them since that is my chief source of experience with the game.
See, now I'm kinda regretting bringing them up at all. I'm not going to introduce new things for the group. I wasn't asking for advice about that. At the end of the day, the way the game is played is what it is and I don't intend on re-educating anyone.
I strongly disagree that a player who identifies and abuse the skill cap-less simply needs a pep-talk outside of playing. I'd much rather build a foundation which is harder to abuse. And I actually think that Maat Mons'es idea is a fitting option.
I always thought that Wizard's on-site articles with tips on how to keep the party together were ridiculous. Doubly so when the reason for schisms came from the "disparity between Wizards and Fighters" (with the rest of them classes in line). A gap that they created (probably unintentionally) with the wording of rules: the same foundation.
Bottom line - I see things and say: "Heeeeeey... Wait a second".

Anyway, disregard that example from before that I gave. I would edit it out, but I don't believe in hiding the past.




Not abolishing circumstantial bonuses. Abolishing +X bonuses. Instead of a +2, a bonus means you change dice by one increment. STacking bonuses means to change increments multiple times.

It's a poor fit with d20 systems, but if you DID try to force a d20ish system to work with it, using 3d6 instead of d20, I guess each +1 would translate into increasing a d6 to a d8, and then a d8 to a d10, etc. So a BAB +5 fighter with 18 strength (+4), a +1 weapon, and a +1 morale bonus from bardsong (total of +11). Upgrade 3d6 to 3d8 (3 steps), 3d8 to 3d10 (6 steps), 3d10 to 3d12 (9 steps), then upgrade 2 of the d12 to 2d6 each, so d12+4d6. Or 3d14 with the right dice-rolling app.

That is, of course, an awful mess to deal with, because it's a bastardized system forced to do things it's not supposed to.

It would work better with a "roll under target" system, where an Expert (+1) Fighter(+1) with a magic sword (+1) attacks a PC wizard (AC 10). From a base d10, upgrade to d8 for magic sword, d6 for Fighter, d4 for Expert. If the target number is 10, it doesn't really matter--but if the target number is 5, those bonuses are huge.

I see.
My biggest problem here would be that 3d14 and 3d6+11 aren't the same. And it's not a gap you can wave over. Look at the graph option of this:
https://anydice.com/program/13c66

There is also something to be said about how much they differ in granularity. A +1 bonus (or penalty) is the extreme case on any die. So even a 1d2 isn't small enough to replicate a nuanced bonus. It's too big of a step.

I really like the idea, though.



Btw, if I were to pick all classes and mix them to 1 averaged class (or, in other words, if I were to create a class for the day-to-day human), what would it's number of skills per level be?