Migrated and edited from the Temple of Time site, where it was originally posted by me as I experimented with GitHub pages.
Wizard: I want to decipher the ancient languages.
DM: Ok, make a History(INT) check.
Wizard: Er, 14?
DM: You translate the book and it says…
We discuss how a skill check should really work in Dungeons and Dragons 5e.
You might say, I know that one! I roll a d20, add my skill modifier from my character sheet, and then the DM tells me I passed or failed! Mechanically, yeah, that’s the basic idea of making a skill check. But you’re dead wrong when it comes to what a skill check is.
Well, let’s discuss what a skill check is supposed to do, and how we as good DMs can accomplish that. Along the way we’ll create some simple guidelines for skill checks.
A skill check is not just a mechanic to throw random number generators/dice rolling at your players. And it is definitely not a “Do I know this random fact about something” roll (we’ll get to knowledge checks later). A skill check is a barrier, an obstacle that players want to overcome. And all obstacles need challenge, right? Otherwise, we’d just succeed at everything! Hm… I’m starting to feel part of our definition coming on. My first instinct would be to say, “All skill checks must involve a challenge,” but that’s a crappy definition. Dice rolling isn’t a challenge, and it won’t feel very good for your players if every time they attempt something they have to roll. Sometimes we need to move the action along (you’ll note that the example at the beginning does nothing to move the game forward). Which leads me to a better rule:
Rule 1: If a player wants to act, and performing that action results in no significant obstacle needing to be overcome (i.e., most people could readily accomplish that action), then the player acts!
We’ll see how that changes later (e.g., most people can light a torch, but it might be harder during the stress of battle). But you’ll notice that I’m giving a lot of auto-success out now. Want to ask the King for money? Just ask (he might just say no). Want to climb on that cart? Go for it! Want to do it unseen? Well, now we have an interesting action. See, in the case of getting to the cart unseen, we have some more interesting scenarios:
Note that in none of the examples did I say, “You failed and were seen.” I gave it more flavor, especially along the middle lines, where he partially succeeds. This is something I borrow from reading DungeonWorld’s rules, but it can add to the narrative drama of the game.
What’s different about this versus other actions? We have multiple resets! In a binary system (ignoring the middle two bullets), we have “success” and “failure.” And it’s up to the DM (us) to make success and failure not boring. (That said, failure should still suck. That may be because it’s boring, but may not.) I also managed to throw in some interesting halfway states. You might creatively apply these if a skill check misses by just one or two points, or if the character would normally be very good at this sort of thing and had a bad roll.
But wait! I said skill check! I implied that this was a skill check, and it is. For one simple reason: there are consequences that depend on how the task is actually accomplished. Some are good, some cost you, and some really hurt (I just wanted to hide in the cart…from the guards chasing me!). Is that a good rule? Sure! It seems simple, and it needs to be: I don’t want to consciously think about this too much. I have more important things to worry about. So,
Rule 2: Only call for a skill check if there are consequences dependent on the result.
Rule 2 (optional rewording): Only call for a skill check if failure is meaningful.
We might call this something like “opportunity cost”—you failed, and because you did that and not something else, this other thing happened. We don’t want to make it too severe (nobody is losing limbs because they rolled a 1 diving into a cart, at least not in my game), but we do want to allow the dice to actually have some shape in the game.
Alright, sweet, so now we have two simple rules: most of the time, our players just get to do things to keep the game moving. No bogging things down with “can I lift this 50 lb weight” (if you have positive non-zero Strength, sure) or “can I climb this ladder.” But, of course, climbing that ladder in time to stop the BBEG from sacrificing poor Princess Zelda has a cost. Climbing the ladder sneakily so the BBEGanon doesn’t hear the approach has a cost if unsuccessful. Now what?
We still have skills on our character sheets, and our players might ask “Can I apply my stealth to climbing the ladder?” For some, that’s a valid question. Me, though, it ruins my immersion, and it takes all the suspense out of the game. So, here’s the solution: get rid of skills in dialog. Angry does this very well in his article (see links at the bottom), so I’ll briefly explain the principle.
Nobody gets to talk about skills. You might describe your character as more acrobatic, or tell me you want to persuade the king to give you more money, but now we’re talking about in-game relationships, not meta-game stats. (There is nothing wrong with meta-gaming, i.e. with players using available information to inform their choices. This is just predictability!)
Done properly, this enables players to be more descriptive of their actions (and DMs to do the same). Let’s look at the ladder again:
RogueLink: So, hearing the sound of Zelda’s screams, I want to climb the ladder quickly but stealthily. I coat my hands in cloth and take my boots off so they make less noise.
DM: Sure You leave your boots at the base of the ladder and wrap your hands in, what, socks, or strips cut from you clothes? While you get ready for the climb though, the screams get louder, and suddenly sound strangled [time keeps moving…]
RogueLink: I’ll go with the socks. Not much time, gotta get in there and rescue her!
DM: RogueLink, you make your way up the ladder incredibly fast, concentrated very hard on not slipping because of your sock covered hand. Go ahead and make a Stealth (DEX) check, with advantage for your precautions.
Now, at this point, I can have a serious, tense moment. The player rolls, everyone waits to find out what is going on, does the rogue surprise the BBEGanon, etc. If he succeeds, he has a serious leg up. If he fails, I have very comic, very dastardly options. Maybe he slips and crashes to the floor, causing Zelda to be moved somewhere else while Ganon searches for intruders. Maybe he gets to the top, but bangs his head on the trapdoor because he was staring at the ladder rungs. Maybe, if he barely makes it, he sneaks up the ladder, but Zelda notices him and that gives Ganon a clue to turn around. Now the Rogue has to make a choice about how to react: stand there, dart behind a column, dive down the ladder, etc.
Our whole scenario got much more immersive. It also became more tense (though the hilarity of Link facing Ganon without shoes is not lost in the ensuing combat). We were focused on the game, not the mechanics of it.
Rule 3: Nobody can mention skills by name as skills.
Obviously, there’s some Grey area—a rogue might claim to be an expert at picking locks, while a bard says, “I generally perform well” (in who-knows-what context). But you can encourage players to be more descriptive. At least they’re describing the game world and not sheets of paper.
That pretty well covers it, right? Well except for maybe those pesky knowledge checks. In 5e, we have things like “Arcana,” “Nature,” “History,” and “Religion.” But some little die roll doesn’t influence what a character is expected to know. We have three basic categories:
This is very straightforward: if your players might not know it, either because they don’t remember what language the Ooks speak or because they haven’t needed to yet, just tell them.
DM: Guys, you see the Phoenix on the tavern door and know that it’s a safe place for Order Members.
If I have them roll to find out if they know this, they’ll never figure anything out about the world (because Goblin Dice and crappy luck and possessed RNG).
Besides, this is knowledge the characters should have.
This is still pretty straightforward, at least to me. It’s knowledge that only people with specific training would know. That sounds like proficiency to me. So people that studied History or have knowledge of the Arcane might remember details about the spellplague. A druid who knows Nature like the back of his hand might be able to identify plants based on their smell.
So, if random squiggles on the ground are relgious purification circles, I tell any player who has Religion proficiency that they recognize them as purification circles of ImportantGodPerson. I still have to tell people things, but who knows what can be easily identified: people with matching proficiencies and people whose background otherwise suggests they might know such a thing. For example, a young Noble whose parents spent extravagant amounts to educate him might know a lot about astronomy and geology (Nature-like things). If a player can argue that their background or backstory gives them certain knowledge (usually in less than a minute), I’ll give it to them.
This is a bit more tricky. Where does that teleportation sigil lead? What are the effects of this newly discovered plant? Now, you might think, skill check! And I agree! This is the perfect opportunity for a knowledge-level skill-check.
To get a leg-up, I’d say the players can
I still use proficiency/background to inform general knowledge (you know it’s a teleportation sigil, just not where it leads) or to give instructive hints (you read about bloodstained rattails in your training to become a priest, but you don’t recall the details). But the players have to figure out how to figure out the missing pieces.
If a spell affects your memory, I might ask for a check to recally very specific facts. If there’s a definite, interesting cost to not knowing something, skill check. If you’re attempting a precise ritual, such as creating a health potion, praying for advice from a god, or studying intricate runes, those are good times to let rolls happen. It’s a measure of attempting something challenging, and lets people that are skilled at those challenges do what they do best.
Keep the story moving!
This comes with the caveat that anything you read on theangrygm.com is filled with foul language: it doesn’t detract from the excellent advice Angry provides, but it may be offensive to sensitive viewers.