Design Origins -- Part 5: Skills

Today I'm going to talk about the origins of the skill system in TNP, which I think is particularly important, due to the bigger role it's taking within the overall designs.


Filling in the Grid
Probably the most wide-reaching question to ask, is where did the "Core Skills grid" come from? There are a few points of influence, and (as with many things in this system) the short answer is that it's a blend of various editions of D&D, plus my own previous designs.

The actual list of skills I went with, was mostly just a blend of 3.5 and 4e -- unifying some skills where 4e did, but adding in some of the old skills it had left behind. The idea was mostly to have a slightly bigger toolbox on hand, in order to deal with non-combat situations. Professions and crafting were still left by the wayside, since there isn't any itemization or money used in TNP -- without which, those skills are largely just fluff.

After a few iterations, I would also separate what I would call "ribbon skills" (Knowledge and Performance) from the Core skills; ultimately, I never thought they were as valuable, and they never really fit into the grid (i.e. having them linked to attributes.) What made more sense, was to borrow from 4e -- by having Knowledge skills more closely linked to Power Source; Arcane classes typically got Arcana as a trained skill, Divine classes got training with Religion, etc. This also helped to give Power Sources more mechanical weight than just interacting with re-specs; generally they didn't do much in 4e, until you got deep into feats and optimization (which is a design space that I wanted this game to avoid.)


The Advent of Skillsets
As I've touched on in one or two other posts before, 4th Edition D&D had a group of class skills that were sort of ubiquitous to all "Fightman"-type classes (namely Athletics, Endurance, Heal, and Intimidate.) Similarly, the Thievery skill was limited to a very small niche of classes, and that was something I sort of wanted to preserve; there might be many classes which could be good at stealth, but deeper forms of skulduggery should still be more limited in their availability. So I was starting to lean more and more into the thought of a "skillset" really being a reflection of "what skills are iconic to this class?"

The idea of allowing classes to be built as "good at a skillset" was intended to avoid M.A.D. -- the design conceit of requiring a character to have several high ability modifiers -- in this case, to be good at their class skills (i.e. Athletics using Strength, Endurance using Constitution, Heal using Wisdom, and Intimidate using Charisma.) Essentially, the system should let you make a character that's good at all the skills in the set, without needing to be "good at" each attribute that the skills are tied to.

Another part of the thought process behind grouping skills into sets, was to allow each group to include skills for each type of encounter (combat, social, and exploration.) This would allow a character to always be able to contribute meaningfully, even if they were only good at one skillset. Ultimately, I would end up keeping skillsets aligned a bit more thematically, but I still think this goal ends up being achieved; generally, each class gets training with at least two groups of skills, which helps to cover the different encounter types.

Now, in the previous game I had worked on, there were 6 skillsets (which at various points in development were coupled and then de-coupled from attributes.)

• Athletics
• Acrobatics
• Influence
• Information
• Deception
• Detection

We can see how this setup clearly influenced TNP. Eventually, Information would become Knowledge (which was then moved to the background skills.) A friend who I had showed the designs to, asked me why I had separated Athletics and Acrobatics, and eventually I decided to just put the two together; conversely, the Deception skillset covered too much space, and would later be split into Subtlety and Infiltration.

The Fifth Element
I can remember when 5th Edition D&D was still in the playtesting phases, one of the notions put forth was that of combining any skill with any ability modifier. Ultimately, I would say that's not what 5e actually implemented, but then neither did TNP; instead, I kind of pivoted slightly, to the idea of distinct skills being the intersection of an attribute, and a skillset.

For example, instead of "Deception" just being a Charisma skill, it would be a skillset; where Deception and Charisma intersected, would be skills like Bluff or Disguise. Other "Deception" skills could be found by intersecting with other attributes, such as Sleight of Hand using Dexterity, or Forgery using Intelligence.

4th Edition D&D flirts with this sort of "mixing and matching" as well, largely due to the necessity of prioritizing your combat stats above all others. This manifested itself in the idea of replacing one ability mod with another, for certain skills. The example that always came up was the feat that allowed half-orcs to use Strength (an attribute which they got a bonus to) for Intimidation, in place of Charisma.

5e allows DMs to use this as sort of a general... piece of advice (I'm not sure you could call it a rule) when appropriate. What all these workarounds solidified for me, was the simple need to get rid of "traditional" ability scores -- the kind that govern both combat, and non-combat. Instead, by using "ability scores" (Attributes) as part of the skill grid, they serve as a second method for grouping skills -- creating another vector by which classes can be distinguished from one another.


Modifiers & Math
In 4th Edition D&D, the design around skill difficulty was something I felt was a problem, and it entirely comes down to the stacking of modifiers. When the game was released, I think the intention was that having either a high ability modifier or skill training would be sufficient to be successful with a given skill (in most cases.)

The problem was that the rest of the system strongly incentivized maxing out one or two stats, and also that there was nothing preventing a high stat from being combined with skill training. When the DCs were revised later on, it was (seemingly) done so, to take into account this sort of rampant overspecialization -- with the obvious side-effect of making it so that combining a low stat with training became suboptimal (and almost pointless, if the stat was low enough.)

So the seeds for the idea of "skill training or ability modifier, but not both" had already been planted in my head. How this manifested in my designs, was by allowing bonuses to a skill to come from either its Skillset, or its Attribute, but having each bonus apply only once -- even if the character had the same bonus to both groupings.

In the earliest incarnations of TNP (with the original slate) there had been skill mechanics based off of each class die; I quickly came to the decision that I wanted a more unified system for skill resolution (particularly as I began to add more classes.) My first thought was to use d6 and/or d10 -- again, leaning on ideas from the previous game I had written; the existing skill mechanics for all 5 class dice could simply be "rounded" up or down, to fit.

I was never sure of the exact implementation I wanted, but when I went in and crunched out the probabilities, it became obvious. The d10 on its own (when added to a d20 roll) pushed the crit chance way too high, and when combined with Expertise, it nearly pushed failure completely off the table. Since the d6 roll combined with Expertise produced comparable results to an unmodified d10 (without the drastic increase in crit rate) axing the d10, and just sticking with d6, presented itself clearly as the simplest solution. I also like that this uses one of the most common dice, sort of keeping in line with the game's more tactile-based design choices.

---

That's all for today! I may dig into some comparisons of the execution of skill systems a bit more, in another post.
In any case, check back on September 29th for the next update.