Haunted House Adventure Design
I admit that I am not the most well versed in horror games. I have both run and played in a handful, but I am in no means an expert. However, in my short time with these games, I have noticed a that a lot of the adventures share a common element with haunted houses. Now, I love haunted houses, but they are a very different medium from TTRPGs, and this design element really has no place in the form. The element that I am referring to is jumpscares.
Jumpscares maybe isnt the best term for this (unless your GM is leaping across the table and yelling BOO), but it gets to the core of the issue. A jumpscare is a quick burst of something spooky that is ultimately of little or no consequence. Call of Chthulu's starter adventure The Haunting is full of these "jumpscares". The PCs are investigating a haunted house and the entity causing the haunting is doing spooky things to ward off the investigators. Among other things, the entity can: form pools of blood, scratch at windows, create thumping noises, and mess with electricity. This is all well and good to create atmosphere, but it offers little of substance for the players to latch on to. There really isn't anything for the players to do in the house aside from experience these jumpscares until they reach the basement and face off against the entity.
Despite not being very fun, haunted house adventure design is disapointlingly common among horror games. People get inspired by movies and other horror media, and try to translate that into their games without understanding that describing a jumpscares is a lot less scary than the real thing. So we get haunted house adventure design. Spooky things happening for the sake of spooky things.
To fix this, we need to add consequences to jumpscares. Jumpscares as obstacles and challenges are way more interesing that jumpscares as jumpscares. To go back to The Haunting examples, perhaps the pool of blood makes crossing a room difficult, or the players need to find a way to communicate or see at night without electricity. By making the scares consequential, they give players a challenge to overcome while providing the spook factor.
Sanity traps, however, are not good consequences. Calling for a sanity check when a jumpscare occurs is not a meaningful consequence, because there was no way to avoid it outside of luck. When using sanity, think of scares like traps instead. A trap has a tell, a trigger, a change to dodge, and a consequence. The chance to dodge and consequence are the sanity check and loss of sanity, but the tell and trigger will vary based on the scare. A tell is simply something that warns the players that a trap is present. In a dungeon, this could be a bloodstain, scorch mark, smell of acid, or magic runes. In a horror game, this could be scattered books on the floor, macabre objects like a skull, magic runes, or even "bad vibes". The trigger is what actually causes the trap, or scare in this instance. Touching the skull, picking up the books, or reading the runes can be the trigger for some larger scare to happen. The important part is that the players understand that interacting with a trap carelessly will cause harm.
Setting better goals is an easy way to add consequence to scares. In The Haunting, the only goal that the PCs have is "investigate the house". This is quite vague, and doesn't leave many points of friction for scares to latch on to. "Find the source of the entity effecting the house and make sure it doesn't come back" provides a lot more to work with. The players first have to find the entity, then figure out how to vanquish it, and make sure it has no way to return. Although this technically is the goal of The Haunting, it is never outright stated, meaning that many GMs and players will stick with "investigate" as the true goal.
I think these solutions could go a long way in making horror games more interesting to play, at least to me. Does this also bring them closer in line with the OSR? Yes. I know what I like.