Why Tabletop Role Playing Games Need Skills


A recent episode of the Internet Office Hours: Role-Playing Games podcast was discussing the question of Skill Checks in Tabletop Role Playing Games (TTRPGs) and if they were actually needed. In designing my game I’ve spent a lot of time thinking about this, and I thought I’d share my perspective.

Skills in TTRPGs

What’s a Skill

Most TTRPGs tend to come in two flavors. Essentially all of them have a collection of core attributes like strength, intelligence, and dexterity. Some also have “Skills”

Some skills are pretty generic, like “investigation”, while some are very specific, like “lock-picking”, but all of them are a representation of things that can be improved with training and practice.

Playing Without Skills

Most games don’t have a “lock-picking” skill, but anyone can attempt to pick a lock. This is generally handled by having the character roll with their most relevant core attribute (Dexterity, Finesse, etc).

This is reasonable, and I’d say actually needed, because you’re never going to have a comprehensive skill list, and there are many things that would be reasonable for a character to do that the player wouldn’t think to list during character creation.

Why That’s Not Enough

Let’s stick with the lock-picking example. In D&D everyone can attempt to pick a lock. It doesn’t matter if the character has ever seen a lock before, or has any clue what a “tumbler” is or how they function in locks. Everyone can pick a lock. Because almost every character has an above-average dexterity, this means that everyone has a decent chance of picking the lock.

This is especially true because the difficulty of a test, in essentially every system, is based on how hard it would be for an average person with a tiny bit of understanding to accomplish a thing. So, it’s a lock with a difficulty of 15 regardless of if you spent your entire life in a tree or spend half of it practicing for the world lock-picking championship.

Realistically, no-one should have any reasonable hope of picking a lock unless they’ve practiced picking locks and understand how they work.

That’s where skills come in.

How Skills Change Things

Some people have been trained in lock picking. Some people can pick locks. Some people are better at it than others, and unless they’ve got a degenerative motor disease it really doesn’t matter how much “dexterity” they have.

Without skills there’s no way to represent this. Without skills, there’s no point in having a backstory that says you’ve trained in archery for your entire life. It’s irrelevant. In many games, someone who’s never picked up a bow, but is an incredible gymnast will absolutely kick the ass of someone with average dexterity who has practice archery their entire lives. Why? Because you’re rolling against dexterity not an archery skill.

Skills make backstory meaningful.

Classes like “thief” are just shorthand for a collection of skills. A “thief” is expected to be able to pick a lock, or a pocket because it’s part of the skill training to become a thief. At least, it is in Swords & Sorcery games.

Classes are a way to get around the need for explicit skills, but it’s only meaningful if other classes can’t do those things. For example, the “fighter” shouldn’t be able to pick a lock, because they’re trained in martial arts, not the ways of thieves. Sure, they can stick a lock-pick in a lock and wiggle it around, but unless it’s a ridiculously trivial lock, they should be guaranteed to fail.

What about Knowledge Checks?

I’ve gone back and forth about the value of knowledge checks. There’s a good argument to be made that the player and / or the Game Master should have an idea of the kinds of things that are reasonable for the character to know.

I went so far as explicitly removing them from my game, and instructing people to simply ask themselves if it’s reasonable for their character to know the thing you’re wondering about, and go with whatever the answer is.

I still think that that is how most knowledge checks should be handled. However, I realized that there are a lot of edge cases that aren’t reasonable, or fun to address that way. For example, pretend you’re a researcher who’s spent years reading old scrolls of arcane knowledge. Have you read every scroll? Probably not. Have you read the one specific scroll about an especially obscure bit of lore? Again, probably not but…. maybe?

What’s more fun? The realistic probability that no living person has read that particular scroll or saying “Probably not, but there’s a slim chance, so let’s roll to find out.” If the dice say you aren’t familiar with it, then you’re no worse off. If the dice say you are familiar with it, then the player gets to describe the circumstances under which they came across this piece of obscure knowledge, and the GM gets to give the players a piece of unexpected, and useful knowledge. A chance for a player to shine, and an unexpected reward for trying.

In this case you might be playing off of an “arcane lore” skill, or the character having a “arcane history degree”. The fighter doesn’t even roll, because it’s not reasonable to expect that their character would have ever been reading through obscure arcane scrolls.

What about Perception & Investigation?


Perception is interesting. The ability to notice things is a skill that you can practice and improve, but few people have any reason to. Your default is generally “good enough.” If you were a scout, a guard, a thief, or someone who’s profession demanded good awareness you’d likely be well trained at this and better than most. However, I’ve yet to see a game that treats it this way.

Perception checks are generally used to answer two questions “What do I see?” and “Did you notice that?”

I loathe the first with a passion. Just tell the player what they see. If they ask “do I see X?” just tell them. There’s either a chair in the room or there isn’t. It’s not like the chair sitting in the middle of the floor will only be noticeable if they roll high enough. Everyone who isn’t literally blind is going to see the chair.

The second has a lot of value, because you can’t just ask the player “Do you notice the assassin sneaking up on your character?” And it’s no fun to just declare that the guards you’re attempting to sneak past automatically notice you. It’s fine if they’ve got a “notice” or “perception” skill that’s radically higher than your “sneak” skill, but as a general statement there should be a roll to see if you can pull of the attempt.


I’m pretty bad at finding things. If there’s something that needs to be found, and I’ve failed to find it, I call in my wife. I’ve got a pretty bad “investigation” skill. My wife has a higher rank in “investigation” than I do. At the same time if I were to rank us on a scale of say, 1-20 I’d put us at roughly the same number. So, eschewing an investigation skill means “just use your intelligence” and now my wife and I are equally good at finding things.

Dogs, on the other hand, aren’t generally considered to be particularly smart animals (relative to humans) but they’ve got really good noses, and are great at finding things once you manage to convey to them what you’re looking for. Being good at investigation, is therefor independent of intelligence. You just need sensory input that’s good enough and a decent strategy for finding the input you’re looking for.

I could practice, and approach the process of finding things in a more methodical way, and then I’d get better, but my intelligence level wouldn’t change. Dogs find things better when paired with a human with a better strategy for investigating. Investigation really is a skill. Deriving it from intelligence doesn’t make a lot of sense.

We need an investigation skill, and dice based checks because of people like me. The thing might be “just inside the closed door” but who’s to say I’m going to think of opening the closet? The roll represents the result of a search. The investigation skill represents how likely I am to do a good job at searching.

We need a mechanism to determine if the thing we’re searching for is there.

The Gumshoe system provides a partial counterargument for this. In Gumshoe if you’re skilled in a thing, and you ask about it at the appropriate time, you find it. If you’re not skilled, you probably don’t. For example, assume your character is a detective who regularly investigates murders. You arrive at the crime scene and ask “do i see blood spatter?” The answer is “Yes, and here’s everything it tells you”. If you’re a professional dog walker the answer is “maybe”. You then roll. Even if you do notice it, it’s not going to mean as much to you as it will to the detective.


When I speak about “fighters” and “thieves” I’m speaking about the archetypes. There’s no reason you couldn’t have a fighter who spent their childhood in a house of thieves learning to pick locks. Any character should be able to have any skill so long as there’s a reasonable explanation and it’s something they’re physically capable of doing.

Also, this is about “most” TTRPGs, not all. In Honey Heist you have 2 attributes: Bear and Criminal. You don’t need any more for that game. Lasers & Feelings has just Lasers, and Feelings. People have had tons of fun playing both of these.

That being said, most popular TTRPGs are building on the ideas that started with D&D. Cypher System is radically different from D&D, but like most, it still has core attributes, and skills. It’s a fundamentally good way of describing the capabilities of a complex and nuanced character, who is more than just their core physical and mental attributes. Some games want that level of detail. Some don’t.


We need skills to make characters backstories meaningful. We need them to differentiate characters in a meaningful way. We need them to address the fact that regardless of how smart, dextrous, or strong you are there are things that each of us know how to do that our companions do not.

We need skill checks to address the fact that sometimes you make mistakes, and sometimes your training isn’t good enough, and sometimes you’re just unlucky.

A character that has a skill gives them, and their player, an opportunity to shine. That same character not having other skills gives the other characters an opportunity to shine. We want characters that aren’t good at things, so that they have obstacles. Having skills says “I’m good at this” but it also says “The skills my companions have, that I don’t, are things I’m not good at.” This enriches our characters, and our stories, and also provides the GM with a list of things that they can leverage to let a character shine, or maybe help take the spotlight off of a character who’s been succeeding a little too much.


Added the Clarification section based on comments on Mastodon.