In Defense of the Perception Check

Apologies, humans, I am in fact a geek.

For those of you who don’t know what a d4 is, or who have never dungeoned a dragon, my apologies. This will get into the nerd-weeds fast.

When I’m revising, I read a lot of tabletop RPG rules, and I’ve been revising a lot recently. I find it calming to page through a rulebook and think about how I’d run the game, what stories I might spin off of an adventure hook, or how a particular class or playbook might handle itself in play. Reading game rules and setting information gives me the immersive thrill of encountering a new world without the constant backseat driving commentary track that runs when I read others’ fiction during those intervals in which I spend my desk time weighing and judging my own work: “Nice scene button.” “Not how I would have handled that moment.” “This sentence is the wrong way ‘round.” “Presuming more affection for the character than I feel at this time.” “Augh! The beats! They’re in the wrong order!”

(This is also what happens in my head when I read my own work. No one is safe!)

But I’ve noticed, reading modern games: some of y’all out there in game land really don’t like perception checks.

(If you’re not a tabletop rpg person, this next part won’t get any more comprehensible, but a ‘perception check’ is what happens in many tabletop roleplaying games when there’s a chance that a player character might, or might not, notice something in the fictional world they inhabit—someone hanging from the ceiling, say, or trying to sneak past a guard. You ‘check’ whether any successful ‘perception’ happened, hence perception check. The player rolls dice, modifies the roll or the result based on how good their character is at noticing stuff, or noticing this kind of stuff, and if the total clears some arbitrary threshold, they succeed.) 

I see rulebook after rulebook explain tenderly that we don’t do perception checks around here, that game masters should present characters with any and all relevant information, that the perception check is a fundamentally bad game system since it’s just an excuse for GMs to withhold information in a way that leads to players not knowing how to navigate the scene, and so on. But I barely recognize the sort of stunted perception checks these rules describe—and often, I find in play that the absence of a perception check creates a hole in the system (or the meta-system, the conversation at the table within which all “systems” operate) that requires a great deal of GM skill and sensitivity to address.

As a GM, a perception check is not a pure dramatic skill challenge in the “roll to see if you jump the chasm or tumble to your death” sense—it’s a communication tool and an opportunity. (Maybe that’s all skill checks are in the end—a question beyond the scope of this essay.) A perception check should never be a bar to narrative progress, and when I see it used that way it is frustrating, a total bummer. But a perception check is a rich opportunity to shape the story to the players, and the table as a whole. It fills so many niches that I have to break out the bullet list:

• A GM-initiated perception check is a streamlined, mechanical way to manage the pace of exposition, while giving the players agency and keeping them engaged. Sure, as the GM I can provide a two paragraph box text description of everyone important at the Duchess’ Masquerade Ball, oh AND the mysterious trio skulking around behind the catering table toward the Authorized Personnel Only door—but by the time I’m finished, at least one player has started constructing the Leaning Tower of d8s, and the Bard is about to hit on Sven the Impaler out of sheer boredom (or desire to be impaled, you never can tell with bards). There’s nothing ominous about being told by the GM that a menacing figure lurks in the corner. That feels like railroading. By contrast: sketch the masquerade ball in a few vivid lines of detailed description. If there’s something to notice, something that a television show might emphasize with a special shot: ask for perception checks. The roll breaks your description into two dramaturgical elements: scene setting, roll, then dramatic reveal. The highest roller will feel good, and they get the nice narrative cookie of information they can use to drive the action and reveal their character based on how they respond: “someone else is trying to infiltrate the masquerade ball!”

• This is also great because characters notice things in different ways. You narrate the Bard noticing the rival burglars with a different flare and emphasis than you’d use to describe the Space Wizard or the Giant Lizard Lady noticing same. This is a great chance for Giant Lizard Lady’s infrared vision to come into play, making her feel like the star for a moment—and to signal that you remembered this salient feature about her character, and care about it, which his as much as to say, a signal to the player that you care about them having a good time. 

• “But Gladstone, what if the players don’t roll well and miss the clues?” My friend, my darling, my GM In or Out of Christ, nobody but you is there to tell the players what DC they have to hit. The modules has a DC? The module isn’t running your game. If you have a standard four-player table, the chance that nobody rolls high enough for you to engage in defensible exposition is rather low. If you’re playing a game with standard DCs—well, I’m going to talk about Blades in the Dark in a moment, but there’s no reason you can’t ask for a controlled / standard roll or a chance roll modified by a skill or stat.

• And let’s say everyone flubs. Even if you want to let the players “fail” a perception check, by asking for one you have given them an indication that there is something to perceive. That creates tension on its own, and increases the chance that players will take active measures to find out what’s wrong.

• You can also use a perception check to test not whether the players notice something at all, but whether they notice in time. In our Masquerade Ball example: perhaps the players don’t notice the suspicious characters in time to stop them, or they only notice the Authorized Personnel Only door swinging shut—and, huh, the caterers who were working the catering table have disappeared, and their manager seems confused… Oh, and the lights just went out.

• A perception check is a non-confrontational way for your players to tell you that they expect to see something—a tool which also gives you a critical beat to figure out how you can use that expectation. “I want to check the General’s desk for traps before I search it!” They roll, do basic addition, experience an emotion about the outcome—you’ve just had somewhere between thirty seconds and a minute to come up with a paralytic dart launcher concealed in the eye socket of the bronzed alien skull on the General’s desk. Player feels good (“gee, I’m glad I checked for traps”), GM feels good. Player fails the check? Maybe they don’t notice the trap. Maybe the trap didn’t exist. Or maybe the General’s files were coated with a slow-acting paralytic poison, that won’t affect anything until the daring escape, when you want to throw just one more fun complication into the players’ lives—which also rewards the player for having checked for traps, even if they didn’t find any. “I knew the office intrusion was too easy!” Flip side: if the players have already complicated their lives enough, maybe the contact poison doesn’t come into play. The check was an expression of interest in a potential outcome—it’s up to the GM to decide whether the game is proceeding in a way that would make it fun, or dramatic, or whatever to explore that outcome.

I could keep going all afternoon. But to sum up: for me, perception checks are a tool for players and GM to negotiate exposition and scene development in a character-revealing and player-driven way. They also create a pleasant tempo which allows space for ideas to develop. When I play games without them, I observe both players and GMs reaching for the tool that’s not there. GMs under-describe or under-prompt (“You sneak into the back office.” Well, okay, what do I see?). Players need more information to act (“Where is the safe where they’re keeping the classified documents?”) and are unable to frame their characters’ actions in the fiction until they get this information—so they have to hand the reins back to the GM. GMs, though, tend to expect that players will do things, and , and, when pressed (“Where is the safe?”), may feel that a check is called for. But what if the GM doesn’t want finding the safe, or noticing it, to be the crux of the scene? What if they’re more excited about the interesting traps that wait within, or the shocking revelation the documents contain? A player asks for a perception check because it’s not immediately clear to them what paths are available in the story; a GM asks the player for a perception check because the GM wants to signal that they are about to interfere with the player’s goals, or change the dramatic situation. These signals, properly deployed, can be as useful as a telegraphed swing in a stage fight—or as bidding conventions in Bridge.

One argument against the standard perception check that I do respect, is that having a single perception attribute doesn’t feel coherent in fiction. The Cleric shouldn’t be the one noticing the ambush just because Perception happens to be modified by the Wisdom stat. I respect this position—but the easy answer, it seems to me, is to use relevant skills (Burglary, Warfare, Sorcery, whatever) as notice checks, with low consequences for failure.

I quite like the Gumshoe approach of having ‘investigative skills’ separate from dramatic skills—it removes the “sorry, you don’t notice anything” prop and forces GMs to play and plan more skillfully, though I miss the fine tempo of waiting for the player’s roll and result, the tiny feeling of victory that comes from a player taking action to move the story—especially if the relevant detail wasn’t part of the adventure before the player informed me, by saying that they wanted to make a roll, that they were interested in something being there. But I see the logic.

Other arguments I see against perception checks boil down too, the player should only be rolling when failure has dramatic consequences. (This kind of case gets made quite often in Blades in the Dark and PBTA-style games, which really want to contain rolling to the dramatic crux of a scene, the moment that resolves its underlying tension.) I could write a whole other essay to write about this issue—but in brief, I think it oversimplifies the ritual of the roll in tabletop roleplaying, in a way that leaves a lot on the cutting room floor. (And markedly increases the felt impact of chance on play, since you’re making fewer rolls per session. If you build your character to succeed at, say, acrobatics 90% of the time, and you make one roll in a session, and you fail it, then subjectively you’ve just failed 100% of your checks in the session—which sucks, since you built your character to be good at this task!)

Perception checks are character-revealing atmospheric checks with low consequence that let players take the satisfying action of dice-rolling, while shaping the narrative. That’s a good thing! They’re also an easy way for a GMs to highlight character competence, while providing players with a nice shorthand that prompts the GM to develop the fiction while offering the GM a little time to think. They’re great tools, when used properly, and while I understand why people try to design around them, it would be neat to see more steering into the skid.

I’m going to do one of those trite “transition to the comments section” things here, but I really am curious: if you game, do you like perception checks? Hate them? Notice them? (Sorry/notsorry.)

And if you don’t game—thanks for following along. Next time I’ll write about Proust or something, maybe.

---

Forthcoming events!

• By the time you read this I’ll be in Hanover NH at Speculation by Design: the Dartmouth Speculative Fiction Project—in addition to the signing (Thursday), the project hosts a faculty panel Friday—really a series of micro-talks by Dartmouth professors about their research and the future of humanity—and an author panel Saturday, in which the authors in attendance will talk about the more speculative, and fictional, side of things.

• In Lincoln NE, at 6pm on April 20, I’ll be reading from Dead Country and signing whatever you have to hand at Francie & Finch books.

• That same weekend, April 21-23, I’m Author Guest of Honor at Constellation NE. Nebraska folks, come say hello! This is a particularly joyous occasion, since I was originally scheduled to be the GoH at Constellation back in 2020, which as you all may remember was… not a year for holding public events. This isn’t a sign of things being back to normal—I still wear a mask indoors & among people, in part to protect the broader human race from any pathogenic warfare agents incubated by my kid’s preschool, and in part because, you know, pandeicm—but it’s a moment of level-finding, of seeking out ways to be in the world we live in now.

• And, of course, my new novel Dead Country is available wherever books are sold (like, say, Amazon, B&N, and my local indie bookshop)!

Take care of yourselves, and happy reading!