Skill Systems: Tweaking "The Middle Road"

Inspired to stitch together this post out of my languishing House Rules document by this post over on Papers & Pencils.

I also like to use “The Middle Road” approach, with a few modifications:

Skill Rating	Skill Die/Dice
UNABLE	1d4
ROOKIE	1d6
NOVICE	1d8
VERSED	1d10
EXPERT	1d12
MASTER	2d6
LEGEND	3d6

The “Ratings” are expanded a little bit, success is still on a 5 or more (making UNABLE impossible for the player without circumstantial modifiers...which are generally kept low/stingy). The bell-curve comes into play with MASTER which really increases the chance of success, but with a cost.

I don't tie skill improvement to level in any meaningful way most of the time. This is mostly to mitigate the “20 HD MASTER Blacksmith NPC” and “I gained a level and am suddenly an EXPERT at Tracking” issues, although Thief classes and formal training-as-cash/time-siphon can open up some additional improvement avenues during downtime with some good fictional explanations. Instead I like to tie Skill Rating improvement to actual play.

Any successful Skill Roll (5+) prompts for another skill roll immediately, and on a maximum die/dice result, the first letter of the next Rating is written down. Once you spell it out, you've achieved the next Rating (this is why all Ratings have 6 letters).

This has diminishing returns: It becomes more difficult to improve/master a skill as you get better as the chances of success increase, the chances of rolling the maximum also decrease. I also like that improvement is actually tied to “Doing the thing,” so players are encouraged to attempt it, even if the odds may not be great (Practice makes perfect!).

Skill Rating	Success Chance	Improve Chance
UNABLE	-	25%
ROOKIE	33.3%	16.67
NOVICE	50%	12.5%
VERSED	60%	10%
EXPERT	66.7%	8.33%
MASTER	83.33%	2.78%
LEGEND	98.15%	.46%

 If it's too harsh, you could also allow improvement on a “max roll” on the attempt itself, but this could possibly increase the speed of advancement a bit and homogenize expertise more quickly.

Thief types have some additional Rating advancement options to help protect their niche, and to ameliorate the malevolence of the fickle dice gods. Upon gaining a level, the Thief may choose one:

•                 Immediately Roll on All Skills for a Letter
•                 Three free letters to be distributed to any skill or skills the player chooses.

If they are part of a guild structure, they may also have easier access to Teacher/Mentor types for certain, non-competitive skills.

If you use class-based XP advancement (with faster advancing Thieves), another option is to allow Thief types the option of immediately spending 100 XP for an additional improvement roll on a Skill success. But my feeling is generally these Classes will likely be making the most Skill attempts anyway, their opportunities for improvement occur more often.

Another option for all classes is to seek out a Teacher/Mentor of a higher Skill Rating in the same skill that is amenable to training the character (often an adventure in itself once you’re seeking EXPERTs and above). After a week of downtime training, the Teacher rolls their Skill die/dice, and on a success the player is permitted to make a Skill roll.

If successful (over 5), they gain 1 letter, if unsuccessful they lose a letter as they have to shed or “unlearn” bad technique. If they roll the maximum (as for standard improvement), they gain 1 letter for each step in Rating difference (so a MASTER would grant the NOVICE three letters). This can be repeated (within reason, and time/resources permitting…training isn’t always cheap or free…I enjoy quest-dispensing Mentors), until the Teacher fails their Skill Roll, or the player reaches the Rating directly below the Teacher. At which point, the Teacher has no new wisdom to impart the student. I’d probably also prevent further training if a player is ever reduced to UNABLE in this way due to poor rolls. No risk, no reward.

Another option is to make the player hunt down a Teacher/Mentor with not only sufficient higher Rating in the Skill the player wishes to train, but also a TEACHER skill high enough to reduce the chances of failure. By using the TEACHER skill to transfer knowledge, instead of the skill itself the learning process becomes much more dicey. This makes finding a good Teacher almost as important as a proficient one. 

All starting characters receive 1 Skill at NOVICE based on Background, they can improve this skill to VERSED and/or take another at NOVICE by taking a voluntary UNABLE in a Skill (max 2 UNABLE skills). Anything else attempted is done at the ROOKIE rating. I’ve toyed with the idea of replacing/supplementing Ability Score modifiers with additional skills or letters (so someone with a -2 in Strength has to take two UNABLE skills involving Strength… or  someone with a +3 in Dexterity gets three free Ratings or letters to a skill or skills involving DEX). Thieves get more starting skill options, but I’ll probably need to cover that in a separate Class post.

I do still really like the idea of a Skill system emphasizing what the player's “Can't Do” rather than what they can (which I touched on here)...to prevent that character sheet paralysis that occurs with really granular skill systems and an invisible DC. When the odds of success are telegraphed to the player with everything they can attempt they may not even attempt the thing. Even though very little resolution is intended to be Player Facing in older versions of D&D (I touch on this problem in this ancient post), when improvement is tied to attempts, players may be more likely to try.

I briefly flirted with expanding Ratings a bit more, but I'm still not sure of the die/dice I would use:

ROOKIE              

TAUGHT

NOVICE              

TESTED             

ADROIT

VERSED

EXPERT             

MENTOR           

MASTER             

LEGEND             

SAVANT

As always, I’m open to thoughts/suggestions.