Horror obviously involves fear, but many genres involve fear. At the core of the horror genre is to reach an intense level of fear using shock.
Shock is an important part of horror and I don't mean that in a vague way—I mean specifically being surprised about something and that something is bad is an important part of what defines the horror genre versus othergenres. In horror either the audience is shocked or the characters are or both.
If there's a criminal with a knife that's a crime thriller, if the audience knows that when they stab a person with the knife it's going to be gory or just terrifying in a way they can’t anticipate that's horror.
If a character is scared because there’s a werewolf that's fantasy if they’re scared because there's a werewolf and because they didn't know werewolves were a thing that's horror.
So in horror, it’s not just a bad thing is happening, someone is surprised at the bad thing happening.
Horror Vs Thriller and Fantasy and Horror-Fatigue
Crime thrillers and fantasy both posit a world where the raw material of the genre is common—just as a thriller reminds us there’s a knife in every kitchen, in fantasy, the supernatural is everywhere.
In horror, the fact shock (a species of surprise) is part of the genre can actually limit the protagonists. After they’ve seen a certain number of horrors, they are no longer shockable. Walter and Jesse can keep getting into sticky situations with armed gangsters each week on Breaking Bad and it’s a thriller every week, Conan can keep fighting wizards and it’s fantasy every month, but you’ll notice the X-Files or Hellboy stop feeling like horror at a certain point, despite the presence of horror elements. This is one reason American Horror Story changes its set up every season and why series’ about “Vampire hunters” or the like tend to feel more like fantasy or superhero stories than classic horror and why George RR Martin writes 7 books about the same people and Shirley Jackson doesn’t.
There is also the question of mystery: defeating a horror usually involves answering a lot of “Why?” questions—and once these are answered, a mythology develops (“Vampires fear holy water”) and it ceases to be mysterious. you climb more toward simple noir (without the supernatural) or fantasy (with it).
The Extended Campaign
There are a few ways to deal with shock fatigue and loss of mystery in an extended campaign:
-Go full gore: Make the mortality rate so high that new characters are a near-constant presence, and total party kills are frequent. You have to have a certain kind of player for this.
-Live with it: Hellboy is a fine comic, even if Hellboy isn’t usually scared and neither is anyone else. Allow your characters to adopt the jaded mien of seasoned monsterfinders.
-Ascending scale of supernaturalness: This requires some planning and discipline on the GM’s part. The first adventure is about a hitman, the second is about a serial killer, the third is about a necromancer, the fourth about a demon, the fifth is about Satan, the sixth is about the Apocalypse. The key is each threat is not only bigger than the last, but, more subtly—breaks the rules of “normal” more fully.
-All the horrors have a common source: Again, this requires planning and discipline. In this scheme, the idea is that the investigators eventually find that all the horrors are part of one larger horror, like the Ascending Scale the horrors get worse, but the idea is the surprise comes from how many new things the larger horror is capable of. Like in the story Call of Cthulhu, the titular monster is responsible for a psychic’s nightmare, for an artifact, for a cult ceremony, and then the creature appears itself.
The Horror Sandbox
A sandbox is a kind of game defined by two characteristics: first: players can, to some degree, pick their objectives and second: the world responds somewhat like a real one would to their actions, possibly on a very large scale.
The classic sandbox is a fantasy game trope: a great map is spread out, with tantalizing rumors about treasure and dangers in various places and the players decide which sound fun—exploring any one of these options creates a cascade of consequences, changing the world and then the players need to decide again what to do in this changed world. The appeal of the sandbox is it offers a great deal of player choice, and unpredicted player actions have unpredictable consequences.
It’s relatively unusual to have a pure sandbox in a horror game, partially because having multiple objectives at once (a werewolf in Warsaw, a ghoul in Ghana) can lead to horror fatigue or a loss of a sense of mystery, but even moreso because an investigation scenario requires a GM to essentially give the PCs an absolutely existentially essential objective (solve the mystery or we all die) and then do a certain amount of work to make achieving that one objective interesting and full of interesting choices. If objectives are optional, they're not really horror. Horror is "Dracula must die" not "Dracula could die if you felt like it".
However, again, with discipline and planning, it is quite possible to make a horror sandbox if that kind of game interests you and your players. Essentially it just means taking the kind of choice presented in the Investigation as Dungeon and scaling up.
The way that most cleanly respects the horror and investigative genres is this:
1. There’s a conspiracy of some kind
2. Near the beginning of the sandbox campaign (or the sandbox phase of the campaign) the characters discover a more-or-less explicit list of separate conspirators.
3. (Unbeknownst to the investigators but likely suspected by the players) the conspirators each have some connection to a horror of their own. Or they are a horror.
4. The list of conspirators makes clear some interesting and at least slightly juicy fact about each conspirator: their neighborhood or city, their job, crimes they may be suspected of, etc.
5. The list also makes clear they are interconnected—exposing/killing/locating one will cause the others to change in some way or ways. Perhaps it is a cult and they all must sacrifice a creature on a given night—and if one doesn’t it is known that another must recruit a new member to take their place. Or perhaps it’s a smuggling organization, and getting rid of one link in the chain means operations will shift to somewhere else. Give the players as much information about these mechanisms as you can without spoiling a mystery you want them to solve—it will make the decision and planning more interesting for them.
6. Treat the investigation of any given conspirator as basically one adventure to design, but with one important additional emphasis: whatever happens to this conspirator affects the others in some interesting way.
7. Season to taste and serve.