Thursday, March 26, 2009

Memento Mori: How Player Death is Killing Horror Games (Part I)

In Part I of this article, programmer Nels Anderson explains why there's a fundamental problem with player death in horror games.

“YOU HAVE DIED” flashed on the screen, crimson and dripping with viscera. He furiously threw the controller against the coffee table. For nearly 45 minutes, cyborg loup-garou commandos had fallen to an unending hail of hollow points, grenades and rocket-propelled grenades. He had finally reached the last boss, the dhampyr cult leader masquerading as the Secretary of Energy, but ran out of mauve herbs before he could finish the creature. Seething, he pulled the disc from the console, tossed it in its case and grabbed his keys. It might still have a decent trade-in price at GameStop …

I picked up Deadly Creatures a few weeks ago and my reaction to it set me thinking a lot about horror and death in games. I absolutely love horror as a genre across media, but its most common manifestation in games, survival horror, far too often resembles the above. It may begin a bit creepy and unsettling, before shortly becoming an exercise in frustration. Yet Deadly Creatures still has me a bit on edge, and I think it is because the source of my apprehension in the game doesn’t become any less spidery as the game continues. But using death as a failure condition in horror games drains the reservoir of fear very fast. For horror games to reach beyond their current aspirations, which if anything have sunk in the last half-decade, some fundamental assumptions about the design of player death need to be dramatically reexamined.

(As a prelude, I want to repeat Chris Remo’s observation that debating what is and isn’t scary is akin to debating whether or not asparagus tastes good. It is wholly subjective and everything here is intended to be nothing but. This is my own observations and those I have heard echoed from others. As always, your mileage may vary.)

Creating horror, especially in games, is deceptively challenging. All the individual components are simple- have the player interact with something most people fear (death, isolation, the unknown), place them in an appropriately thematic atmosphere and make them feel vulnerable. The failure lies in the execution. Looking at Resident Evil 4, one of the most popular survival horror games of late, these failings are obvious.

For being, ironically, far past the 4th Resident Evil title, the game begins surprisingly well. The creepy, isolated European forest is a solid setting, and your police escorts are quickly dispatched in a fashion most gruesome. The villagers are not obviously zombies, but certainly bloodthirsty and dangerous. The reveal of the first chainsaw-wielding villager is especially unnerving, given that he is likely to decapitate the player regardless of their current health. Then the village’s church bell tolls and all your attackers immediately cease hostilities and simply shuffle off. It’s quite effective at combining the feeling of vulnerability with not having any notion of what is truly happening.

Unfortunately, the atmosphere cannot be maintained. The first encounter with an El Gigante may be unsettling, but it has become wholly rote by the seventh. Dying time and time again to invisible enemies is frustrating, not frightening. The Resident Evil tropes of abysmal writing and voice acting return, literal shooting galleries and ridiculous midgets torpedo the atmosphere further, and soon any sense of horror is gone. It’s still an extremely fun game, and one that I like a great deal, but it’s also exemplary of how nearly all horror games cannot reconcile their fundamental design decisions with the aesthetic and emotions they’re attempting to invoke.

The major challenge for creating horror in games is that the feeling of fear largely relies on the audience knowing their character is in constant peril. For the majority of games, death of the protagonist is narratively meaningless. The player’s free time is threatened, for if they die, they’ll have to repeat some section of the game. Fear of death becomes fear of wasted time. The intent is to create tension, to make the player anxious about dying. But the consequence is that player is now forced to repeat the same section of the game and it will not be nearly as scary the second, third, nth time. Groundhog Day does not chill the soul.

This is the fundamental design problem with using player death as a penalty in horror games. By attempting to make the game tense by increasing the consequences of failure, every time the player actually does fail, the fear of the unknown so essential to the atmosphere is dismantled.

Swinging the pendulum too far in the opposite direction, and removing the consequences for failure completely, is rarely successful. The genre of narrative games that most eschews player death is the graphic adventure. There haven’t been many horror graphic adventures, at least partially because tension is hard to maintain when the only failure state is being stuck on a puzzle. There were a few solid attempts, including some live action games (and wow, would you look at that cast?), but ultimately, they did not prove successful. Horror games simply do require that feeling of vulnerability and the tension it creates.

There have been a few instances of making the player character’s death permanent (and then giving them control of another character). Unfortunately, it’s also problematic due to the many times the player will die in the course of normal gameplay. Despite not being a horror game, Call of Duty 4 features some pretty powerful scenes involving the death of the player’s character. But the dozens of temporary deaths the player has experienced up to that point mute the potency of the moments when they die “for real.” Folks still enjoyed them, but it is clear they are not as powerful as they could have been.

Nels Anderson is a programmer at Hothead Games. He's probably the only game developer in Vancouver (and maybe all of Canada) that was born and raised in Wyoming. He writes about games and game design at


Ben Abraham said...

Great article Nels.

I wonder if something akin to the Left 4 Dead director could be the answer to solving the problem of repetition destroying tension. If, when you have to try a section again, something completely different and unexpected happens (even if it's that nothing happens this time) would that be enough to keep the player tense?

Nels Anderson said...

@Ben At GDC, I asked Dead Space's art director about that thing. He responded that it was a difficult problem to them to address. They tried some judo where they'd set up a set of player expectations and then intentionally break those expectations. It's definitely better than nothing, but it's still just pushing the problem out one level.

A completely procedural system like L4D's director could definitely help. It would probably have to be a bit more granular, but I think it would be a pretty easy way to improve on the status quo in a lot of cases.

Post a Comment