Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 21, 2014
arrowPress Releases
October 21, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Antagonistic Horror Design
by Josh Bycer on 04/24/13 02:45:00 pm   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


Horror design has been called out by mainstream designers in the past for not being marketable. Today's post examines a common horror movie narrative and how it can impact horror design.

Reprinted from my site: Game-Wisdom

A few weeks ago, I had something that I haven't experience in a long time: a nightmare-- Specifically a "wake up with your heart racing" nightmare. It's been a while since a nightmare has actually made me wake up tense as I've reached a point where nothing really scares me in real life.

But enough psychological talk, the element of the nightmare that relates to game design is an unexplored area of horror in game design: antagonistic.


A Game of Cat and Mouse:

In my last post on the subject of horror design, I focused specifically on the protagonist, looking at alpha heroes. This time, we're going to flip it around and talk about the thing (or things) that are trying to kill you.

Going back to horror movies while there are two types of protagonists, there are also two types of antagonist : horde and mastermind. Horde is where the hero is dealing with mindless creatures on the hunt, or some kind of rampaging monster as the focus. We can include everything from zombies, to aliens and just about every Sci-Fi channel original movie.

If there is some kind of main villain who is either controlling them or is a monster itself, it is still considered a horde type as the main threat comes from it unleashing its minions at the hero. The necromorphs in Dead Space, while supposedly under control by the marker, were still left to run around and cause destruction.

A mastermind antagonist is someone with an actual agenda and not just about running around causing trouble. These characters will do their best to stalk their prey, get the group separated and are usually the stand out character in the story. Characters like Pinhead or Freddy Kruger as an example.

We could probably debate if the thing from "The Thing" would be either a horde or mastermind type monster. On one hand the creature did everything it could to blend into the group the best it could. But then again, it wasn't that smart once someone found out and attacked it.

Incidentally, stories that feature horde type antagonist are ones where the hero is usually an alpha character, while the reverse is true for mastermind antagonist. The reason is that when the antagonist is the stand out character, the viewer becomes invested to see how they are going to get rid the survivors and how said survivors will hopefully get away.

This kind of cat and mouse game is one of the main sources of terror with mastermind antagonists: as you know someone is going to die, the only question is when. But for such an interesting take on the genre, it's one we don't see in video games often.

Evil AI:

Antagonist villain design is an interesting concept for a horror game as the player is not the one in control, but the villain. More importantly it would get around one of the biggest barriers of creating scary games: linearity. If you had a game where the AI would actively track the player and do whatever it can to get the player into a trap to kill them, there would be no set attacks in that case.

Imagine playing a horror game where you could never really be safe as whatever you came up with, you knew the monster would be coming up with a workaround. I bet for a lot of people reading this that last statement made them excited about the possibilities but there are a few catches.

Haunting Ground was all about hide and seek from the chasing antagonist, while trying to solve the game's many puzzles.

First is that there were very few games that went the mastermind approach. With the only three that I could think of were Resident Evil: NemesisHaunting Ground and the Clock Tower series.But while each game did have a reactive villain, the titles were still largely linear.

You as the player had to complete tasks in a reasonable order and there were cases where the enemy would just show up at set points.

Trying to make a game with a mastermind villain in today's setting is especially tough, as we have seen from most major developers that they believe that horror is out. Both Resident Evil and Dead Space have become more action-horror then survival which go against the mastermind format.

The player should not be the one in charge just plowing through everything; they need to be weak and forced to react to the situations around them. But of course that goes against the concept of having an action based hero.

Now the tougher issue is on the technical side, as you would need a highly complex AI to effectively "play" the game against the player. Something like the AI director from Left 4 Dead taken to the extreme, but that is not an easy task. Strategy game designers, who make some of the most complex titles around still haven't figured out a way to make a completely reactive AI.

And that is the key ingredient in making a horror game where the player is truly being hunted by a superior force: Where attacks are never predetermined and the path through the game is always different.

The indie game: Slender that was released a few weeks ago was a step in the right direction. Requiring the player to find random objects while being hunted by a monster. 

As mentioned, many major developers feel like the horror genre has been played out, but maybe they're just looking at the wrong kind of horror. Left 4 Dead's experiment with the AI director showed the basic concept of having an AI affecting the experience. Taking that and expanding on it could open the doors for reviving the horror genre in the eyes of mainstream designers.

Related Jobs

Treyarch / Activision
Treyarch / Activision — Santa Monica, California, United States

Senior UI Artist (temporary) Treyarch
Treyarch / Activision
Treyarch / Activision — Santa Monica, California, United States

Lead UI Artist
Treyarch / Activision
Treyarch / Activision — Santa Monica, California, United States

UI Artist/Visual Designer (temporary) Treyarch
Treyarch / Activision
Treyarch / Activision — Santa Monica, California, United States

Senior Character Artist (temporary) - Treyarch


Michael Stevens
profile image
I think the biggest problem with horror game enemies stems from our expectations for the protagonist. When I fire up a new Silent Hill game I know my character is going to survive for at least ten hours. Anything an enemy does to kill my character is non-canonical and a result of my failure as a player and not intrinsic to the reality of the game. This severely defangs enemies. If the worst thing an opponent can do is kill me, and that doesn't count then I'm just playing a grim version of Peekaboo. The current design trends don't let individual monsters be terrifying in the way they would in real life.

L4d offsets this slightly by having shorter scenarios and win states where one or more characters just don't make it. It undercuts itself a bit by only having the four characters, inexplicable revive closets, and a loose level order. It's (gratefully) got those as concessions to multiplayer, but in terms of building dread it's at cross purposes.

Josh Bycer
profile image
"The current design trends don't let individual monsters be terrifying in the way they would in real life."

This is a good point and a big problem with current "action-horror" gameplay. The first time you fight an enemy can be scary, the 50th time, or a room filled with ten of them, not so much. There is really only two ways I can think of to get around this.

1. Make individual enemies do more then just run at the player. Have them track the player, break down barriers and have them do whatever they can to keep after and hunt the player down.

2. Limit the # of enemies and instead have a small number of "boss enemies". Such as having 5 mastermind enemies in the entire game and no one else. Design each one to be completely unique and put the player in a situation where they are being tracked and hunted down while trying to figure out a way of dealing with them.

Luis Guimaraes
profile image
Good that you mentioned "action-horror", Josh. It comes perfect with the exemples I was going to use anyway: Terminator and Predator.

Both are great mixes of action and horror, with character that usually fit the alpha-kind as you put it, and with the solo antagonist premise. All you can do again the terminators is slow them down, and the predator is just playing with you for most of the time and could finish you at any time.


This brings us the the third way to manage the same situation, which is the same done in Predator 1 & 2, RE3 Last Escape/Nemesis, Silent Hill 2, and First Encounter Assault Reckon 1 & 2: The use of contrasting enemies and situations to deal with the flow aspect of the game. Which can also be simply called the "food chain" aspect of the game.

The Predator example goes more into stalker foe ground is not among my favorites because it falls in the same category as Amnesia, Ju-on, Night Of Sacrifice and so on, which I personally call "Adventure Horror", where you have a lot of "scary" scripted events and almost no real threat (check Thomas Grip's talk at GDC, in the "No Death" section). The fake threat is too transparent to the player, so such predator game could come to life very similar to FEAR 1 & 2.

The trait of Power Fantasy is that the player is always on top of the food chain, while in Survival Horror the player is on the bottom of the food chain or at least at the same level of most enemies. Having enemies both bellow/equal and above at once is also a way to deal with it, but rarely well explored in games.

Alfa Etizado
profile image
I think it is a problem that goes beyond action horror games. On my post down there I talk Amnesia up, but paradoxically that game lost all its tension when you suffered the worst thing that could happen: meeting enemies and dying at its hands. Amnesia was at its scariest when you didn't die.

First time you are killed you aren't scared anymore. You've seen the enemy, you know where it'll be, how to avoid it. Though Amnesia did right by not resetting your state. Whatever you used between a save and your death was lost forever. That's the key.

Permanent death can be very scary, but it limits the game in many ways. So instead of going all or nothing, I think partial loss is better. I believe it is better if enemies don't cause game over.

Partial loss can be a bunch of things, just going to throw some random ideas. Enemies make you lose useful items or resources, that's what Diablo 2, Dark Souls and many other games did. It doesn't even need to be items, it can be characters dying.

That characters dying thing could also become story "losses". Failing encounters could mean the story taking a different course. I think this can be tricky to pull off though.

I think the player must not clearly be aware of how the story branches, so it shouldn't feel too mechanic, like your success decides if you follow path A or B. It's better as something cumulative, if it isn't based on pre-defined paths, if it isn't based on good or bad endings and if it relies on player set goals, like "I want to reach the end with everyone alive", if someone dies the game can go on but not like desired. Once again this can be tricky, I believe that no path should lead to a less satisfying game, even failure shouldn't do that. It should be bad because of the player's desire, not because the game becomes limited or anti climatic.

Another idea, body horror. Failing an enemy encounter deforms the player's character in some way. It could also reflect in gameplay, for an instance making the player slower.

There are many ways to do partial loss.

Josh Bycer
profile image
The concept of partial loss mechanics sounds interesting but would have to be carefully balanced.

On one hand if you make a game with only temporary saves and each time the player dies they lose something. If you're not careful you could create a lost cycle: where each time you die will make the act easier to happen next time.

On the other hand if you make it so that you can save at will, then what will stop the player from setting up saves to fall back on and get around the system.

I think the story aspect could be a good fit: It's punishing the player for dying/failing, but would only matter to the players interested in seeing the whole game and not "punish" someone who doesn't care about it. The idea of limb removal was something I was playing around in my head a few years ago for a game concept: where if you fail enough times and get caught by the enemy, you'll be punished by losing a limb.

Re: Amnesia I came to the same conclusion about the horror mechanics. At the end of the day, if you want the player to be scared while playing, then there has to be something on the line. It's one of the reasons why people have found playing a game like Dark Souls scary as one mistake could kill you and you risk losing all your hard earned souls.

Alfa Etizado
profile image
I think the limb thing works better if it makes the player character disgusting. I mentioned that slowing player down example but later on I realized it would suck to run around slowly for the rest of the game. Similarly it would suck to remove an eye and blank half the screen or remove a hand and make aiming a gun worse.

These would permanently affect the way the player interacts with the game, they'd have to make controlling the character worse and that can't come out well. Maybe it can but I have no good ideas for it. It is more drastic than killing off characters, removing resources.

Just making the player character repulsive I think could work. From time to time throw in an absolutely disgusting moment, like making the player take care of a wound or something.

Michael Stevens
profile image
Dismemberment is exactly what I had in mind. I wouldn't have it be at a damage threshold, but a percentage chance on grabs. If you have shorter scenarios you can significantly hinder the player in ways that would just be annoying in a full size game. If I've only got one arm maybe I can't hold a shotgun anymore, if I'm missing both then maybe I can't use doorknobs and I've just got to kick all the doors down and play the last 20 minutes as pure flight.

I don't think you have to take a hit on the narrative to pull off permadeath. Part of the reason it would seem that way is because so much of game narrative is binary "kill all the monsters or die in a sub-basement" stuff. If you open yourself up to think about what else the characters might care about you can build in a lot of equally tiered objectives and let the player decide what winning is. The Nippon Ichi PS2 Srpgs did this frequently; their cut scenes were so asset-lite that they could do massive diversions and all they really needed to make was the dialogue.

Josh Bycer
profile image
I haven't played it yet, but one thing I thought was neat about ZombiU is what they did with the character system. Each time you start a new game you become one of the random survivors and when that person dies you have to become another survivor.

If someone could take that concept of body switching and put it into a narrative with a small group, now we have a horror game with a limited lives system. And there is also the dynamic of having certain members able to go to different areas weaving that narrative back into the gameplay.

Alfa Etizado
profile image
One thing I think is wrong with enemies is that I know what they'll do the moment I see them. They are too descriptive in every way, even name. The Spitter in L4D2, the moment you see her you know she'll spit acid and die in a few bullets. Playing through Dead Space, almost every enemy I could tell what they'd do and how I should kill them, from just looking at them the first time.

This can be good, but not in a horror game. In a horror game, the unknown is much scarier than a big tank monster with glowing joints chasing you.

On first sight, the player should have no clue of what an enemy does or how to kill it. Preferably, the enemy should surprise the player with new actions in future encounters. The player shouldn't even know if the enemy is dead or not. Going even further, the player shouldn't know if it is an enemy in the first place.

Despite its many obvious enemies, the new RE games do that to some extent. After being killed enemies might evolve to one of the many different mutations. There is so much variety, this keeps them unpredictable. But that's what RE did.

Other games can make enemies behave differently based on where you find them, on what other enemies are around, on what weapon you use or you can just give the enemy a wide range of actions but play your cards one at a time. There are tons of ways to do this.

Even though all of this can be become a one pony trick, I believe it is much better. Horror works by keeping the tension high and having one big terrifying moment, those are the ones that stick. When you first pit the player against a new obstacle, that's the big moment. If you face the player with a predictable enemy, that moment isn't as good as it could've been.

Amnesia did that very well. You never truly faced enemies, so you didn't know what they were capable of. It had just a handful of enemies, some you never met again. Remember that water monster? Easily one of the most terrifying moments in any horror game for me and you can't even see the monster, how genius was that?

gard skinner
profile image
I believe so much comes down to level of the linear aspect of a game. The biggest scare moments are intricately scripted, but therefore you have to be pushed to the predetermined place and time to activate the sequence. When you look back at the early survival horror games, it was so new, RE and Silent Hill, it felt open at the time, the gamer didn't have expectations as they'd never been set up like that before.

The challenge is the gamer also wants important decisions that affect the sequence of events. So as soon as you open the world, the options for a big cinematic event slip away. I think you need to ask more questions like what do people really fear? It's not just gruesome graphics. Loss. Family strife. Solitiude. Lack of control. Poverty. Abuse. The list is long.

The idea of a "monster coming up with a workaround" is great, but the monster can be anything that you have shown the player that they need to be afraid of.