Emotions in Games, part 1: Fear

I’ve been thinking a lot lately about how well (or rather, how very, very poorly) games create emotional reactions in players, and since I tend to be a verbose bastard I’m going to separate my mind-dumping into posts on separate emotions, talking about how different media approach these emotions, and how games generally fail to do so, with a constructive eye on how this can be improved.

H.P. Lovecraft wasn’t only the past master of gothic horror; he also wrote one of the seminal works on the subject: “Supernatural Horror in Literature”.  His contention therein was that the type of horror that most effectively engages and frightens people is fear of the unknown.  I strongly agree, and one can see Lovecraft’s influence on everything from Stephen King and Dean Koontz to slasher flicks.  But let’s drill down a bit further into the ways horror, and especially the fear of the unknown, can be manifested in the media.

Fundamentally, I see three components to horror: suspense, wrongness, and startlement.  I’ll approach each in turn.


Suspense is generated by building tension in the reader/viewer/player.  It’s most effective when it builds slowly, adding onto itself with labyrinthine twists and turns until we beg for release.  Suspense arises from our perception that we don’t know something important, something crucial to the plot or to our own survival.  In order to build it slowly, clues have to be dropped, one at a time, so we have the sense that we can almost but not quite see the whole picture.

A lot of poor writers (usually scriptwriters) try to create suspense by having characters act in miserably stupid ways, e.g. shining a flashlight ahead of themselves as they slowly back down a dark hallway.  This builds tension sophomorically, relying as it does on the idiotic and unrealistic behaviour of the characters in question, but it can be effective; the tension generated comes from our desire to force the characters to act more intelligently.  “Don’t do that!” we cry, wincing as we feel the danger coming ever closer.  This approach is far more common in film and TV than in books, and is very rare in games.  Typically in games we share the perspective of the character we control, and the designers can’t rely upon us as players to act in such stupid ways, except of course in cutscenes.  This kind of suspense often is hard to distinguish from frustration and can easily bleed into annoyance.  Roger Ebert calls this one of the elements of the “idiot plot” – a plot that only works because the characters involved all have to be idiots.  The suspense generated by this kind of stupid behaviour in films and TV is almost always released through startlement (see below).

A more sophisticated form of suspense is one that doesn’t rely on characters acting in unrealistic ways, but rather sets the stage for the protagonists to slowly uncover the underlying mystery.  I’ve rarely seen this done effectively in games, to be honest, but it seems like a good fit for the medium nonetheless.  The basic requirements are simple: 1) the mystery must be crucially important somehow, often because of a physical or mental threat to the protagonists or to people or things that they love; 2) the mystery must be uncovered slowly through reasonable and realistic means; and finally 3) the mystery must be internally consistent and the clues should lead to the answer.  Red herrings are fine, but the final solution should make sense and should explain all of the clues that were uncovered.  This leaves the reader/viewer/player with the sense that they have a task to perform (i.e. figure out what’s going on) and that they have a reasonable chance of doing so.  A lot of mystery writers (I’m looking at you, Agatha Christie) fail the final test there by “cheating”, that is, they withhold clues that are essential to solving the mystery, and so the experienced reader quickly becomes disengaged from the tension the author tries to generate.


Wrongness is the sense that things are not as they should be; this quite often underlies the building of tension, and thus suspense.  You can however have one without the other, so it deserves its own explanation.

Lovecraft was especially gifted at conveying a sense of wrongness; his protagonists often quite literally went insane as a result of learning too much about the eldritch truths hidden from mankind as a whole.  Often there was no “final reveal”; the reader doesn’t usually have any direct evidence of the underlying problems, but only sees the symptoms.  Wrongness can be generated by something as simple as a bloodstain or torn piece of clothing, a strange sound, an odd smell, or even a premonition.  Like suspense, wrongness is based on small individual clues that lead the reader/viewer/player to the conclusion that Things Are Not As They Should Be.

Wrongness is quite easy to generate in all forms of media, since it merely relies on phenomena that are out of context, and thus are jarring to the percipient.  Video games can convey this readily, from the bizarre reddish skies of Angmar in LotRO to the graffiti left behind by the Joker in Batman: Arkham Asylum.  Some movies and games take this too far, such that the player becomes inured to the weirdness (e.g. American McGee’s Alice… which is initially creepy but then just seems silly after a while).  When gore is everywhere, we become desensitized.  A light touch is often best here; less is more.


Startlement is generally a cheap trick, founded on the sudden appearance of a threat (or something initially confused for a threat), or simply a loud noise.  Cheesy movies and TV shows rely overmuch on this approach to horror, and I’d argue that what’s generated isn’t true horror at all, but merely surprise.  Startlement is relatively rare without an initial buildup of tension; if the viewer/player isn’t already tense, then the sudden appearance or loud noise is just a strong stimulus and won’t cause the viewer/player to jump.

Startlement is quite uncommon in books, partly because it generally relies on the speed of the stimulus, and partly because it relies on visual or auditory stimuli.  It’s not very startling to read “and then there was a sudden CRASH! and a dark figure lurched out from behind the swaying door, bloody knife in hand”.  It’s a lot more alarming to see and hear that than it is to read it.

Startlement is quite common in both film/TV and in games.  Good examples of the latter include almost every first-person shooter (at some point) and as far as I know, every single horror-themed game of any sort.

Just as in wrongness, less is more with startlement.  If things jump out at us all the time in video games, we come to expect it… unless there’s a very high level of underlying tension this results in the startling phenomena becoming ordinary.  The ordinary isn’t frightening.  Again, less is more; it’s fine to startle players but do it sparingly for greater effect.

So now what?

OK, so there are three different building blocks that content creators can use to generate horror.  What underlies them and how do we use them?

My belief is that the underlying issue for all three building blocks above is control.  All creatures seek to understand and thus predict their environment, so they can maximize their own survival and success.  When we see things out of place, when we start to think there’s something important that we can’t perceive or understand, and when things surprise us, we view these things as threatening to our wellbeing.  The more immersed we are in the media we consume, the more readily we’ll translate this loss of control to fear.

What this DOESN’T mean is that the best route to scaring players is to overtly take control away from them by having their characters do things the players don’t want.  This generally results in a loss of immersion and by extension a loss of interest.  I’ve seen this in a lot of cutscenes, and I can’t think of a single time when it pleased me as a player.

Instead, horror is perhaps best generated by hinting at, rather than showing what’s wrong.  Allow the clues to build up over time, and make sure the player feels like they’re given a fair shot at understanding the underlying problem.  Note: the player only needs to have the perception that they have a fair chance; they don’t actually need to have a fair chance.  The more powerful a player feels, the harder it is to make them feel threatened; this can be overcome by having more powerful foes, but that’s a crude approach that usually fails in the long run, since most games allow player characters to grow more powerful over time.  A better way to overcome the player’s feeling of power and control is by lateral thinking; don’t give them problems with easy and obvious solutions, but rather suggest that something truly horrible is probably out there right now… and who knows what would be needed to stop it?

We return here to Lovecraft’s underlying argument: true horror rests on the principle that we, the viewer/reader/player don’t know what’s going on.  Fear of the unknown is much more powerful than fear of the known, especially in video games, where “the known” might surprise or even kill us, but will rarely if ever frighten us.

Now, fear of the known does of course exist, but it requires a very high level of immersion (e.g. real life).  Fear of the known is very, very simple: it requires a direct threat to one’s wellbeing or to the wellbeing of something very important to us.  We fear suffering, death, and loss – all of which are “known” quantities.  Since death in video games is usually not very important to us as players, and since most video games don’t create a high level of immersion, fear of death simply doesn’t work well most of the time.  Suffering and loss in video games is a very tricky problem, and one I’ll address at length in part 2 of this series.  Every time we enter into combat in a video game we’re directly facing these fears, and it’s extremely rare for a player to experience actual horror when confronting the known.  The unknown has far greater potential than the known in video games to create real fear in the player.

1 comment so far

  1. […] Like Fear, we’re dealing here with an emotion that’s experienced by the percipient – the reader/viewer/player of whatever medium is being experienced – and not by the characters in that medium.  How, then, can percipients be made to feel sorrow?  I’ll approach the subject by dealing separately with suffering and loss, and then I’ll briefly address how sorrow can be part of PvP experiences. […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: