Percentile based Skill system

[Mith]

Hello the Playground,

When it comes to skill mechanics, the idea my brain always comes back to is using percentile dice with modifiers. In part because I like how percentiles work, and would like to use them more, and because I think that it can work to communicate the difficulty of a task fairly easily. If the average person (no modifiers or penalties to the roll), has a 50% chance of success at the task, then the DC is 50%.

The problem I see coming with percentiles is that 1% increments mean that one can miss rolls by a very small margin, even if that is not intended by the GM. To that I would suggest the idea of tolerance, where if you roll below your target, but within tolerance, you succeed due to your skill level. Within 5e, this can be determined by your proficiency bonus.

Tolerance of failure is based on systems where you roll for a target number and if you roll under you have consequences of "you succeed, BUT...", "you fail, BUT...","You fail", or "you fail AND...." depending on how far you missed your mark.

While this idea is theoretically system agnostic, I am going to use 5e as an example case, as I am more familiar with it. With this change, you roll percentile dice plus a modifier based on your attribute bonus and proficiency. Thinking about this a bit more, it could be interesting to change the skill system a bit farther:

1) Upon character creation, you get your list of skills that you can choose to be proficient in, and skill points to invest. Instead of being points like 3.x, these determine your level of proficiency. Class features or feats that allow you to be an Expert allow you to invest two points into chosen skills. When you reach the next tier (proficiency bonus changes), you gain skill points to allow you to raise a skill to the next level, broaden your skill set and pick up a new skill at a lower level, or if you have a skill you are an Expert in, you invest two points in to boost that skill twice. You can never have a higher proficiency bonus in a skill than an Expert at your current tier. This sort of change would likely require a larger skill list if you want to prevent a character from filling out their skill list, but they won't be as successful as a more focused Expert, so it may work out well in the end.

2) You roll percentile plus proficiency bonus and stat modifier for your result. I personally like the feel of proficiency dice for proficient rolls and the standard "double regular proficiency bonus" for Experts, as it makes Experts more consistent, but that's my opinion before playtesting, and may change.

3) The tolerance of failure is set by your proficiency bonus. So a +2 proficiency gives you 2% increments of failure,+3 3% and so on. An Expert gains no boost to this tolerance, as they get a means to boost through two categories of failure with the double proficiency.

What this (hopefully) looks like in play:

The DCs from the regular system is converted to this system by determining the % chance of failure, and rolling above it. So DC 5 is a 25% chance of failure, DC 10 is 50% chance, DC 15 is a 75% chance and DC 20 is a 95% chance.

A Level 1 rogue as an Expert (Acrobatics) and a +3 Dex mod is looking to walk a narrow ledge quickly. The DM figures this is a DC 15 in the old system, which translates to roll over 75% chance. The rogue rolls a percentile, and adds +7 (+2 proficiency*2, + Dex modifier) If they roll 66% or higher, they will succeed with no problems, as they have a 2% tolerance range. A roll of 64 or 65% means they succeed but with complications. Rolling 62-63% means they slip and may need to make a Reflex save to catch themselves, with below a 62% they fall off the ledge. There would be no "You fail AND..." option to this, unless the drop is small enough that they could potentially reduce damage of the fall through an Acrobatics check. Then the last degree of failure would deny them that check.

A note that the above example was chosen off the top of my head. I realise and would give that more careful crossings could be auto success with no roll required. However rushing a process would trigger a roll.

Upon reaching 5th, 9th, 13th, and 17th , the rogue gets skill points equal to the number of skills they are initially proficent with (in the case of the now level 5 rogue, 6 skills from class and background, plus two expertise points for 8 total). They can choose to spend these points to upgrade all their skills to the next level, (+3 proficency, with two +6 Expertise), or to not upgrade some skills in return to obtaining others skills. On reaching the next tier, the rogue will always get 8 skill points to spend. The only case I can see for potentially getting more skill points is to spend a feat to get more skills, allowing you to obtain more skills. I would probably allow for anyone to gain Expertise from investing skill points, but Expertise features make things easier as they give you the extra points to maintain your Expert status.
The trade off at 1st level will probably be using class/race/background overlap to reduce your pool of proficennt skills to gain Expertise (eg. Half Orc Barbarian uses both Half Orc proficency in Intimidation and Barbarian profiency in Intimidation to gain Expertise, but they now only have 1 class based skill to invest).

I hope this idea is at least interesting to others as it is for me, and I would like to hear other people's thoughts on the subject.

Sincerely,

-Mith

[Knaight]

Percentile systems are pretty common, though the way it usually works is that your character skill is the percent chance of accomplishing a default task (so if you have something like hacking 35% you have a 35% chance to hack a "typical" system), with bonuses and penalties applied directly to the skill, often in 20% increments. Technically yours is less a percentile and more a d100 roll and add, but that's just pedantic.

As for the concern of missing by small increments, that's not really a concern, and is just how failure chances tend to work. Having some tiny 2%-6% tolerances is just fiddly and obnoxious. Take your rogue example. That first 2% tolerance is literally just a bonus applied weirdly, as they still succeed. It's them effectively having a +9 % instead of a +7%, applied more obnoxiously. Then those 2% increments start coming in.

On top of that these bonuses are tiny. You've multiplied the die size by 5, but left the modifiers unchanged (well, boosted by +Prof%). That +7 on a d20 is equivalent to +35%.

So, lets convert D&D 5e to percentile here, making a few decisions along the way but preserving most of the math. First, the odds of succeeding at a DC 10 check with a +0 bonus is 55%, so that's the baseline. That gives us this structure:

Ability Checks: Assign abilities as normal. To determine base success chances for ability checks multiply the attribute bonus by 5%, then add 55%.

Proficiency: For skills with proficiency multiply proficiency by 5%, and add to the ability success chance.

Bonuses: As Proficiency.

Task Table:
DC 5: +25%
DC 10: +0%
DC 15: -25%
DC 20: -50%

Add the task difficulty to the roll, then roll under the success chance to succeed.

That's where you'd start making changes from, like the margins of failure based on skill level - which is a fairly proven mechanic, generally, though I can't say I've seen it in a percentile system before. There's options for how to do this, but I'd strongly avoid 2% margins. My first inclination would be to calculate margin from success chance overall, probably just dividing it by 10 and rounding down. At the very highest (max attribute, expertise, max proficiency) that +17 becomes a 140% chance, so the margin is 14%. That's pretty huge, but that's also a skill with a lot of investment. Then just don't use the "you still succeed" margin.

[Mith]

Percentile systems are pretty common, though the way it usually works is that your character skill is the percent chance of accomplishing a default task (so if you have something like hacking 35% you have a 35% chance to hack a "typical" system), with bonuses and penalties applied directly to the skill, often in 20% increments. Technically yours is less a percentile and more a d100 roll and add, but that's just pedantic.

As for the concern of missing by small increments, that's not really a concern, and is just how failure chances tend to work. Having some tiny 2%-6% tolerances is just fiddly and obnoxious. Take your rogue example. That first 2% tolerance is literally just a bonus applied weirdly, as they still succeed. It's them effectively having a +9 % instead of a +7%, applied more obnoxiously. Then those 2% increments start coming in.

On top of that these bonuses are tiny. You've multiplied the die size by 5, but left the modifiers unchanged (well, boosted by +Prof%). That +7 on a d20 is equivalent to +35%.

So, lets convert D&D 5e to percentile here, making a few decisions along the way but preserving most of the math. First, the odds of succeeding at a DC 10 check with a +0 bonus is 55%, so that's the baseline. That gives us this structure:

Ability Checks: Assign abilities as normal. To determine base success chances for ability checks multiply the attribute bonus by 5%, then add 55%.

Proficiency: For skills with proficiency multiply proficiency by 5%, and add to the ability success chance.

Bonuses: As Proficiency.

Task Table:
DC 5: +25%
DC 10: +0%
DC 15: -25%
DC 20: -50%

Add the task difficulty to the roll, then roll under the success chance to succeed.

That's where you'd start making changes from, like the margins of failure based on skill level - which is a fairly proven mechanic, generally, though I can't say I've seen it in a percentile system before. There's options for how to do this, but I'd strongly avoid 2% margins. My first inclination would be to calculate margin from success chance overall, probably just dividing it by 10 and rounding down. At the very highest (max attribute, expertise, max proficiency) that +17 becomes a 140% chance, so the margin is 14%. That's pretty huge, but that's also a skill with a lot of investment. Then just don't use the "you still succeed" margin.

Thank you for your response.

I can see your point about die size vs. Percentiles, and the concept of using margins was to try and effectively shrink the die size. I was also trying to think of a way that allows for an in character assessment of "How difficult is this thing?" that isn't just a d20 number. That can still be done here I know, but it isn't quite the same as "untrained person has this failure rate, roll above." feel I was going for, even though my execution of concept was terrible.

So while actual dice mechanic is out of whack, is the change in how one's skills are gained and improved worth it, in your opinion?

[brian 333]

White Wolf/Runequest had a very good d100 system. You might look into it for ideas.