GM Resources

Tribal Knowledge: The Thrill of Horror

Welcome to the first in my new column, Tribal Knowledge, in which I answer questions from readers and friends about gaming, and especially about GMing.

Precipitating a Panic among Players asks:

I, and others I know, are working on horror games.  Outside of the GM stripping down, what can be done to help instill that thrill of horror in the players at the table?

Dear Precipitating,

Inspiring fear in people who are, to the ordinary senses, sitting around a table in someone’s home, in comfortable chairs, and drinking a beverage of their choice is harder than it looks. To be honest, it’s not especially easy in a darkened movie theater, in chairs that have known the intimate fluids of unnamable horrors from beyond the veil of stars, drinking whatever your life savings let you buy from the concession stand. Personally, The Ring scared the crap out of me, but not many other horror movies even give me pause… not that I watch a lot of horror movies, though.

I can tell you for sure that there is no right moment to go au naturale in a tabletop game. In my case, it’s because I set out to inspire fear. Pity is rather off the mark, I’d say.

The gold standard of advice for horror GMing is Kenneth Hite’s Nightmares of Mine. (I know it says it is for Rolemaster. It is in fact system agnostic.) A lot of what I have to say is a recapitulation of things you’ll find there.

First, figure out if your players are:

  1. Signed up for a horror campaign. If so, open with a conversation in which they explicitly agree to let their characters be scared and/or horrified by things they discover.
  2. Not specifically showing up for a horror campaign, but open to allowing their characters to be scared and/or horrified, and aware that horror is a seasoning on an otherwise non-horror-based campaign.
  3. Dead-set against letting their characters be scared and/or horrified. This is, unsurprisingly, a really difficult case, and trying to horrify players who are actively resisting is probably more of a recipe for GM frustration than a surmountable challenge. Also, there’s a whole separate conversation about trying to do something to someone else without their enthusiastic consent. This is not the place for that ethics argument, and I’m not sure how I feel about it in this precise context in the first place.

There are also a bunch of different related emotions to consider. When I talk about “engaging with” an emotion, what I mean is that ultimately players have to choose to let themselves feel something, or not. Engaging with an emotion is permitting yourself to feel it and let it inform your character actions, rather than shining it on to prove what a great hero you are.

  1. Dread. This is the “best” of the list, because it sets a whole tone in the player’s mind that lowers the bar for achieving other emotional payoffs. Generating dread is a matter of long-term and subtle foreshadowing.
  2. Horror. This is fear from revelation: putting all the pieces together, or something suddenly demonstrating its true nature. Generating horror is the major end goal of your question, but it’s challenging over time because one revelation after another desensitizes the players to the awful things they learn or witness.
  3. Scares. I’m using this term for lizard-brain fear responses, like jump-scares in movies or sudden loud noises. Scares have the benefit of being more involuntary than other emotions, so you might be able to scare a player who hasn’t signed Yes, please scare me on the dotted line. They may not thank you for this. Also, it’s harder in tabletop than in other gaming media.
  4. Tension. This is the “heroic” response to overwhelming odds and situations gradually getting worse. This is a building-block emotion: there are a lot of players who resist engaging with fear and horror, but still engage with the sense of challenge in a scene or storyline. If your players aren’t willing to at least engage with tension, my advice in all seriousness is to talk it out and convince them otherwise, because (judging by the question’s premise that the GM is interested in a non-silly game) you’re at quite a gulf from the players’ emotional state. Getting on the same page in your group’s social contract needs to be Item 1 on the agenda.
  5. Anger. (At NPCs, not at the GM. Anger at the GM is trivially easy to generate.) Since conflict is the core of all scenes and stories, anger is as easy as falling off a log; just have the antagonists, er, antagonize, and you’ll have all the anger you need. It is, of course, not very useful, because in a game where violence is an acceptable solution to a problem, anger is a positive emotion, a precursor to agency. Dread and Scares, and to a lesser extent Horror, are about the absence or failure of agency.
  6. Confusion. There’s a good kind of confusion and a bad kind of confusion here. You want the one where the player has the basic knowledge of what the tools are and some of the ways they might be used, but does not know any of the specific combinations of uses that might lead to a positive outcome. To put that another way, you want them to know what nouns and verbs are available, but not what to say. Without the basic knowledge of the tools and some of the ways to use them, you have more frustration than fruitful confusion. I’ll come back to this.

The Fast Track to Fear

The single quickest path to tension is the self-inflicted wound: turn the players’ agency against them by showing how they have overextended themselves and underestimated the opposition. The important thing here is that the players have to have initiated the idea. If there’s an out-of-character awareness that there was nothing they could have done to avoid this conflict, most players I’ve ever met react with out-of-character anger at the GM, because it feels like punishment without cause.

This can be as simple as tactical concealment of a portion of the enemy force. For example, I once got a great reaction out of my players by having enough creatures for a “normal” encounter out in the open, when there was no particular reason to suspect hidden enemies. There was a ten-foot-high wall nearby. The players committed to the fight, and one of them decided to be clever and flank his way around that wall. There were, in actuality, three drakes hidden behind that wall, elevating it to a “difficult” fight… all the more so because that characters was now cut off from the rest of the party, and while he could have taken one drake all right, or held his own against two while the party cleaned up the other foes, three drakes could eat him for lunch (and did). To be fair to the player, if he had guessed right, he would have had a commanding tactical position, et cetera. The important thing here is that it got a good spike of tension and fear from the player, without the player just being mad at me or feeling like I had screwed him over with a false promise.

In this example, all I’ve done is bump up the tension in one way, and once the players re-assess the situation, they can probably figure out that while that one character might have a problem, they’ll probably get there in time to save him. They didn’t know the specific stats of the drakes by any means, but I didn’t go out of my way to describe the drakes as exceptionally high-danger creatures.

Some GMing advice would tell you to describe everything in as bombastic and terrifying of terms as possible; personally I think this undermines the players’ trust in your descriptions. Once they take down something you’ve described as a sanity-rending chthonic terror in three or four normal hits, your credibility takes a hit, and that is the kiss of death.

To ratchet up the tension further, introduce a second (and third, fourth, whatever) new threat later on, or introduce sources of tension that are not quantifiable like the drakes were. With additional new threats, make sure the PCs have had a round of actions between the arrival of new threats, or it just feels like a rider to the first new threat, and loses something in the process. Also, don’t wait too late in the fight to introduce the new threat, if it is generally quantifiable; that runs the risk of feeling like enemies trickling in slowly enough to be convenient for the PCs. Well, okay, it’s a matter of scale: if they have to go directly from one fight into another that is equal to or greater than the first, without a chance to rest, that’s a big boost in tension. The point is, small threats belong in the ramping-up portion of a fight, where they can magnify the effect of the increasing tension. Big threats belong wherever the hell they want to be.


I’ve mentioned the GM’s credibility a few times, and I want to return to that for a moment. There is nothing more important to establishing an environment for horror than demonstrating in detail that you will carry through with threats. In a game, no one ever shows up to miraculously rescue a horror protagonist, unless it’s one of the other protagonists. Players who have gotten accustomed to GMs going out of their way to save characters from death may be unhappy with this, especially if they are permanently losing a character they have developed from the start of play. There are various available solutions here, all of them outside the scope of my response. The important thing is, once you establish a conflict and stakes, follow through no matter what. The only punches it is okay to pull are the ones you haven’t revealed yet.



This is a story about a little-known mathematical formula: c * p = T. c is confusion as to what needs to happen, and physical disorientation. p is pressure: combat pressure in this case, but other kinds work as well. T is terror.

In the Dust to Dust LARP, we ran an adventure that started with the door slamming shut and sealing itself once the last of the PCs was inside a darkened, fog-filled building. (Darkness and fog cover, and generate, a multitude of sins. This is as true in tabletop as it is in LARPs.) We had set up a simple circular path using tarp walls, and blocked the path at several points with locked or secret doors. A concealed staff member used a mic and speakers to talk to the players as a disembodied voice. There was pretty constant combat pressure from a variety of creatures, ranging from a minor variant of a common monster, to weird new monsters, to a major campaign villain the players had seen exactly once before. I don’t want to overstate the effect that all of this had on the players, but especially when the aforementioned major villain mind-controlled one party member and liquefied another, I felt that the adventure successfully hit the mark for Horror.

The things that made this work could cross-pollinate to tabletop gaming pretty well. In a tabletop game, it’s easier to alter the map of the play area to create disorientation, though having a map of the gameplay area reduces disorientation. If you’ve drawn a map for them, it needs to be honest, or else it’s the GM lying to the players rather than the characters’ senses lying to them. I’m splitting a hair here, but I think it’s important. Instead, use your description of the area to tell the players what they can’t see – describe the limits of their perception, whether from smoke, fog, or leaves, and hint at what lies beyond. Once you establish that limit, allow time or player actions to push past it one detail at a time.

Bombarding the players’ and characters’ senses is harder in tabletop. Darkness and fog are good, but in tabletop they’re mostly descriptive terms that modify combat rolls. This is a good place to inject uncertainty into your descriptions, though. The disembodied voice is surprisingly difficult to do well, because almost all of the players’ game-relevant sensory input is auditory (aside from accessibility solutions for the hearing-impaired, obviously): the GM’s voice. As Dungeon World puts it, a tabletop game is a conversation. The disembodied voice could either exist as a recording that makes it hard to hear the GM, or as interjections in the rest of the GM’s description. The first option misses the mark by having one conversation talking over another, while the second still operates in just one channel of communication – not exactly overwhelming.

In tabletop, look for non-auditory ways to communicate information, which you can use while applying in-game pressure. Combat is traditional: solve the puzzle of this fight while, you know, fighting and taking damage and keeping the bad guys from dismantling your whole team. Skill challenges of many kinds work too: your ship is falling apart thanks to that black hole you flew too near (the Kessel Run is not for you), so you’ve got to figure out the repair schematics (presumably a text prop that you’ve put into the game) as systems fail. Protect Engineering and Life Support, if you know what’s good for you.


Establish De-Protagonizing Stakes

I’m going to catch hell for this, but: set things up so that the best victory the PCs can hope for is not all that great. Maybe escaping is the best they can do. Maybe saving half of the (crew, townsfolk, PC party) is the best possible outcome. Because you’re running a horror game, you can block off paths to better victory conditions. This is for games where the PCs have agreed to horror and/or desperation as the game’s tone. The horror is in making the PCs do awful things in order to prevent worse things.

When you’re building the setting, put the forces of Good on the run, in hiding, or otherwise rarely seen. Probably not totally absent… but the PCs need to believe that if they do nothing, Evil triumphs and things get worse. There probably aren’t other NPCs who could take care of that, or if there are, having them do so makes things worse in a different way.

One great application is showing two enemies duking it out, both of which are too powerful for the PCs to handle. You can lavish detail on the horrible things they’re doing to one another – it’s an ideal chance for exposition on your Big Bad. Make sure the PCs know that whichever one wins this fight, they do not want to be here when the dust settles, because even with injuries, it’s still more than a match for them. If you can show how its victory in that fight makes it even more dangerous to the PCs, so much the better.

A feeling of agency is the direct counter to fear. It’s still possible to experience horror at the same time as full agency – horror at one’s own actions, or the sudden revelation of the villain’s power and plan. The villain’s revealed nature loses its fear, though, if the players are well-positioned and motivated to do something about it.

On the other hand, don’t set things up so that the PCs can’t make some positive difference. If they lose hope, you have nothing left to crush! To take best advantage of the whole medium of gaming, where the PC actually is the protagonist and takes full responsibility for the protagonist’s decisions, you want to embrace agency while also subverting it: unexpected and dire consequences of player actions, or presenting awful solutions to worse problems, with time pressure so that they can’t seek better solutions.


Use Your Own Fears

If you have any phobias, spend some (very unpleasant) time thinking about exactly what you find disturbing about that situation or thing. Bring the PCs into contact with that, and make yourself as uncomfortable as possible in describing the scene. The reason Lovecraft’s writing works is that he was scared of damn near everything, especially people at all different from him, and he communicated that in text. Stephen King has written about how The Shining is specifically about his fear of snapping and hurting his children.

In Dust to Dust, I’ve poked around in one player’s phobia by accident. Unaware of her phobia (which I do not share), I spent five minutes or so describing creepy-crawlies burrowing under her skin. She did not thank me for this, but (fortunately) she also did not deck me. I believe that it is very rude to do something like this on purpose, because phobias are (as if by definition) the kind of fear that the player does not enjoy, because it is not separate from their real, out-of-character experience. I apologized for this misstep afterward.



There’s a huge amount more to say about horror and fear in games, but I’ve got to end this somewhere. I hope I’ve provided you with a new angle on unsettling your players. It’s difficult to walk the line of “scary” without someone reflexively making a joke to relieve the tension, but I think that if you stick with it and never ever allow yourself to be the source of the joke, the tone of the humor can become to bleak that it gets wrapped up in the horror and starts to reinforce it.