Around 10 years ago, a lot of very interesting and ground breaking horror games were released. These include Silent Hill (1999), Fatal Frame (2001), Forbidden Siren (2003) and a few more.
Since then not much has happened in the video-game horror genre and little has evolved. So what exactly can be done to push horror in video-games further? To answer that I will here present a list of my top 10 things I think could take horror game to the next level:
1) Normality
In most games the player usually starts out in some strange and not very normal situation. In our own game, Amnesia, the story takes place in early 19th century and has the protagonist waking up in gothic castle. Not something very easy to relate to. Other games see the player has some secret agent, has them trapped in a spooky town/village, etc. All of these are very abnormal situations, and something few of us will ever find ourselves in.
However, much of the good horror in other media starts off very mundane. They build on the having the audience strongly relating to what is taking place and being able to draw close parallels to their own lives. For horror games this would mean establishing a very familiar situation and then slowly introducing the horror there.
The goal is for the terror to not just be inside the game's virtual world, but to reach into the real as well.
2) Long Build-Up
Most games want to kick off the action as soon as possible. Even games with a drawn-out introduction, like Silent Hill 2, introduce the horror elements very early on.
The problem is that sustaining a really high level of terror is only possible in shorter bursts, and the more the audience has to contrast to, the greater the peaks intensity will feel.
Ring (Japanese version) is a prime example of this. While it does kick off the horror early on, the whole movie is basically one long build-up to a final scare moment. Horror video games need to embrace this sort of thing more, but in order to do so two common traits need to be let go.
First of all, the game must rely a lot less on a repeatable core mechanic, since we want the player to deal with actual horror elements as little as possible. Secondly, we must perhaps revise the game length and be satisfied with an experience lasting three hours or less, so that all focus can be on establishing a single (or just few) peaks of terror.
3) Doubt
Many of the best horror stories raise the question of whether a phenomena really exists. Is the protagonist really seeing ghosts, or is it all in her mind?
Since other media like film and books are very grounded in our reality, this sort of thing comes naturally (although it is still not always easy to sustain). However, in video games the player is in a virtual world with its own rules and entities, and this leaves little room for the player to doubt if anything could really exist.
Solving this is no easy feat, but I think a first step is to embrace the previous two entries in this list: normality and a long build-up. If the player relates to the game as "real life" and gets enough time to establish this idea, then she will eventually start to compare any features of the virtual world with the real.
Eventually she might start doubting if the ghosts, monsters or whatnot are really there. Also, some sort of sanity mechanic can also do the trick, but it must be a lot more subtle then any previous attempt. The player cannot see it as a game system, but has to view it has a feature of their own mind.
This is not an easy thing to establish, but that is not the same as it is impossible.
4) Minimal Combat
I have talked plenty about this before (see here and here for instance), but it is worth stating again.
The worst thing about combat is that it makes the player focus on all the wrong things, and makes them miss many of the subtle cues that are so important to an effective atmosphere. It also establishes a core game system that makes the player so much more comfortable in the game's world. And comfort is not something we want when our goal is to induce intense feelings of terror.
Still, combat is not a bad thing, and one could use it in ways that evokes helplessness instead. For instance, by giving the player weapons that are ineffective, the desperation of the situation is further heightened. This is a slippery slope though as once you show a weapon to the player it instantly puts them in an action game mindset.
That does not mean weapons and combat should be abolished, but that one should thread these very carefully. Finding the right balance is a big challenge for future horror games.
5) No Enemies
By this I do not mean that there should be no threats to the player lurking about. What I mean is that we need to stop thinking of any creatures that we put into the game as "enemies." The word "enemy" makes us think about war and physical conflict, which is really not the focus in a horror game.
It also makes us think less about why these creatures are in our virtual world. The word "enemy" is such an easy label to put on other beings, and then not worry about anything except that we need to destroy or avoid them. This is how wars work, after all.
If we instead think of these creatures as merely inhabitants of our virtual worlds, we need to ask ourselves why they are there, what their motivations are, and so forth.
This brings a new depth to the game which is bound to color the player's imagination. If we can establish our hostile beings as calculating, intelligent beings with an agenda, we vastly increase the intensity of any encounter and can make the terror so much stronger.
6) Open World
By this I do not mean that horror games should strive to be GTA-like sandbox experiences, but simply that they should allow more freedom of movement. Most horror games set up a very strict path for the player to follow even if they have, like Silent Hill, a large world to explore.
Instead, I think future horror games should allow for the player to skip certain areas and to go about in the world in a free way. This increases the player's feeling of being in a real world, increasing any emotions associated with it.
This is also closely related to the goal of achieving normality. Without a forced structure and more open world, it should be easier to give the sense of everyday life.
7) Agency
Horror games are so effective because they can make the player feel as though they are there when the horror happens. Other media, especially in the horror genre, have to try really hard to accomplish this, but for games it comes almost automatically. It is then a waste that many horror games does not take advantage of this properly and destroy the sense of agency in all kind of ways.
By far the biggest culprit are cutscenes, especially when they take away control at scary moments when the player's actions should matter the most. Another problem is connected with the open world entry above, with the player constantly being fed where to go and what to do.
The way to go forward here is to make sure that the player is involved in all actions that take place. The scenes that are so often left out (and replaced by cutscenes) are often vital aspects of the horror experience. Whenever possible, the playing should be doing instead of simply watching.
8) Reflection
The video game medium tops all others in giving its consumer a sense of responsibility. If something caused by the protagonist happens on the screen, the player has been part of that. This opens up for the game to be able to reflect itself upon the player and to make players think about themselves while playing.
Games have tried to do this in the past, but I do not think it has come very far yet. So-called moral choices are very common in games, but are hampered by being obvious predefined selections (choose A, B or C) and by being connected to the game dynamics (making the choice more about what is best for the player stats wise).
I think that the choices need to come out as much more organic for the player to truly feel as if they have caused them. To be able to do this, a strong sense of agency (as mentioned in the previous entry) must be achieved, and the player must truly feel like it was their own choice (which ties into the "Open World" entry above).
I also think that this can be taken a lot further than simply testing the player's ethics. It can put player in very uncomfortable situations, and really make them evaluate themselves as human beings. The game could also lure them into mind states that they never though they had in them. It can explore the nature of good and evil and similar subjects in a way that would be impossible other medium. In the end this can lead to some really personal and terrifying experiences.
9) Implications
What really brings some horror home is when it has implications in real life. This can be something like the fear of TV sets that Ring manages to achieve, or the bleak and disturbing universe that Lovecraft's stories paint.
Elements like these are almost entirely missing from video games, and again it ties into other entries on the list. Normality is probably the most important, and if we are able to achieve that, it will be much easier to tie elements of the game into everyday life.
A game that can achieve this successfully takes the horror to a new level, by being something that the player carries with them long after having put down the controller.
10) Human Interaction
The final entry will also be the hardest one: to bring human drama into the game's actions.
Most horror in other media does not have the phenomena/situation per se as its focus, but instead its effect on people. The Exorcist is a great example of this, and so is The Shining. However, in video games the main actions still revolve around inanimate objects or brainless foes. By typing the player's player's actions directly to other people, the horror gets so much more personal and intense.
Achieving this is not an easy task. My opinion is that it is not a technical problem, but one of design. It places a larger burden on the player's imagination.
Simulating a fully (or at least seemingly) sentient human being is a really hard problem. Simple solutions like dialog trees come often out as stiff and prefabricated. Instead, one should go the route of simple actions, like the hand-holding in Ico, and build upon that by being vague and hinting instead of trying replicate a book or movie.
Exactly how to go about is an open question, but the any steps closer to success can mean a lot to evolving the horror experience.
End Notes
That concludes my 10 steps for better horror games. It will be fun to see if they are still valid 10 years from now or not. If you have any other ideas on how to evolve horror games, please say so in the comments!
[Thomas Grip is the co-founder of Sweden's Frictional Games, developer of the Penumbra and Amnesia horror series. This article reproduced with permission from In the Games of Madness, Frictional's game development blog.]
It would be nice to see games come out that have a short focused story. The only problem is that people won't pay a premium price for a short adventure. Developers who want to venture into this area have to be willing to accept less money for their efforts than a normal AAA title. However, I think there are many people that if given the chance would spend $10 on a 2 hour game the same way that millions of people spend $10 on a 2 hour movie. Dear Esther is a good example of this working(Although 1 of the biggest complaints was the length of the game.).
I don't see why a three hour story has to mean a three hour game. I think a compilation of short horror stories would probably make sense to audiences in a way that a collection of action games wouldn't.
I think developers would probably be drawn to make something like "Tales from Silent Hill" or some sort of Lovecraft deal with a location-based horror revisited over time/ by different characters, but you could just as easily have them be totally unrelated. Stacking concepts and characters like that would probably make players feel a lot more vulnerable, particularly if you combined it with the sort of nested endings you get in jrpgs like Metal Saga or NIS games.
I agree just because the main story is only 2 or 3 hours long doesn't mean that the game can never be touched again. I would hope that if you really focused on a good 2-3 hour experience it would be good enough to revisit multiple times, the same way a really good movie has more value than just watching it once. The only problem I see about making large compilations is that the larger you make the game(Adding more stories makes the game larger.) the less focused the game becomes. It could take away from that short experience you were working on refining. Its true you can keep adding more and more stories to the game, but why not just release them as individuals games rather than one big collection of stories?
It would be interesting to see a game with the same sort of structure of a film like Juon, which was more like a collection of 3 short films united by a common geography and theme. This would, I think, particularly suit a game since it could be both short, and yet offer a satisfying overall length.
It reminds me also of the changing controllable characters in Sanity's Requiem.
There are two problems with open world horror (neither total blockers):
- You have to provide something to keep the player entertained while you're building up (that normalcy). Now you're building two games (or at least one and a half).
- I'm not sure you can sustain the horror when the player can just drive away. Now perhaps as it spreads it can get harder and harder to get away - your car just dies, or hulking beasts that can stop cars. Since you already know the area (from running around in it earlier) you still have a slight advantage you don't normally have in horror games.
Deadly Premonition tried this, and it mostly failed. I played it, but it certainly wasn't for the horror - which was contained to (mostly) closed in linear areas. Red Dead Redemption: Undead Nightmare took the tack of literally taking an existing full open world game and bolting zombies onto it. Fun, but not very scary. Alan Wake also fun but not scary, and Resident Evil hasn't been scary for decades.
Really the only scary game I've played recently was Amnesia. The slow controlled experience did the trick even when I could see how they were manipulating me.
The same game mechanics can be applied in two different ways. The build up then can be something of the tutorial. Take, for instance, the point and click adventure. You can feature a "going about their life" intro which you then disrupt with some horrid event.
I think the only way to make an open world work in horror is to make it a dynamic open world. That way, you can establish early on there is an open world, and when you need to, restrict it significantly. Silent Hill sort of does that in a few cases by dropping you in an apartment building. All of the doors are interactive, so the building seems really big, but most are jammed shut. You go from door to door hoping to find someone or something who can help you. Then, when you fall into the slightly creepier dark world, everything changes. Changing the maze while the player is inside it is very disorienting. They are lost in a newly hostile environment they just managed to get a handle on, and they are wondering who it was that unlocked all these doors (and why would someone do that?) Both give the game the appearance of less linearity and so, as per the other points, more horror.
In an original game, maybe you could subtly change the handling and acceleration of cars depending on some element. Or periodically change the controls and then change them back based on an invisible insanity stat (imagine inverting the Y axis for 1 or 2 seconds with no visual cue when a monster jumps out at the player). Do it right, and you make a mild scare and outright hectic situation. Although do it wrong (and too often) and your audience says the controls are bad.
Horror devs should pay more attention to the soundscape, too. It's almost(?) more important than the visuals. There have been few things more goosebump inducing than hearing those zombies wheeeeeezing somewhere nearby in Thief - which is actually a great example of taking an existing great game and slowly adding horror elements. But that's 1998, right in the time period you were discussing.
Agreed. I have the greatest respect for sound designers I know, due in no small part to the power they wield and how well they wield it.
In lieu of fingers, sound best conveys the sense of touch. After all, sound waves physically impact us (especially if we have a killer sound system). In a visceral horror game, then, sound is paramount.
IMO, Siren 2 is still the finest horror game to ever be released. It took the basic framework of Siren 1, removed a lot of the clunkiness and added some cool new ideas of its own (like the blind guy stages). However, Siren 2 came out 6 years ago. It's high time for someone to surpass it.
Siren 2 came between Siren 1 & the PSN episodic game and is the best one of the lot IMO (although the PSN game has better graphics since it's a PS3 game and not PS2). It was never released in North America but it was released in English in Europe as Forbidden Siren 2.
I also find that horror games (and movies) seem to think that gore = scariness. Although some kind of gore is almost always used (and can add a lot), gore is more likely to gross someone out, rather than actually scare them.
Interesting note in agreement here. At no point in the infamous shower scene in Psycho do we see a knife cutting flesh or even a bleeding wound. It is a terrifying murder done entirely without gore.
The amount of combat and enemies aside (which IMO is not really related to amount of horror...otherwise why are there many scary zombie movies?), I really feel like many of these points were covered very well in the original F.E.A.R, and it makes me wonder if the author has played that game? What was most terrifying about that game was the long periods of build up to something happening, the personal relationship of the characters, and having to come to grips with your own identity (when it is revealed at the end). There are many moments of the type of agency he refers to, and the final level makes you feel so confused and helpless it's truly terrifying.
This is something I've been thinking of a lot in the last few months since our next game may be a survival horror type game. I clearly remember playing the first Resident Evil (PS1 version) late at night and getting a few good jumps. Then I graduated to Silent Hill and really saw what a horror game could be. It as fun but had a lot of boring/repetitive bits in between sirens. I would agree that the genre has been pretty stagnant for a long time. I as actually really surprised with Dead Space. Yes it doesn't follow a lot of the traditional survival horror scope, and tends to follow more of the slasher movie genre, but I can't think of many other games of late that spooked me as much that I enjoyed playing (I'm looking at you Deadly Premonition). I ill say tat one of the most scary moments of Dead Space 2 was when the player has to go back onboard the ship from the first game and even though not much happen onboard the ship there is the memory of the first game around every corner.
I remember one game called The Thing for the XBox. I think it really did get human interactions down, because you never when one of them was going to turn on you. Also, they didn't trust you much either, so there was a whole dynamic to it. Sure, there was a lot of action involved in the game, but I remember that I was legitimately scared the first time I played it.
I agree Kevin, I think the most effective parts of the The Thing were dealing with the uncertainties of the other humans. The action was okay, but everything fell apart when things went from non-action to action. It was a shame seeing how scripted things were (cross this line, and no matter how much testing you do on a human, he will turn into the Thing because he's scripted to do that). But overall the human parts were unique enough to make the Xbox The Thing still memorable for me.
It feels like I've been stuck in horror mode for the last 2 weeks. Between writing a design doc for a horror game and a series of articles on horror game design, I can't seem to escape it :)
From the list: #5 I agree with completely. If you just have the player fighting mindless enemies one after another, then it just becomes a slow paced action title. Enemies should be able to hunt the player and mess with them. With so many horror games that have action in them, why haven't we've seen a game where the enemies can knock the weapons out of the player's hands or just simple grab them and use them on the player?
With weapon design I think a mix of weak and power can work to elevate the fear. Let the player bash some minor enemies around without much trouble. Then introduce something that can shrug off a knife to the head and throw the player though a wall. There should be a mix of combat and stealth, to force the player to decide between fighting or hiding.
Personally I think the key to future horror design is to take a page from the rogue-like genre. Introduce scenarios that will not play the same way each time, with monsters that are constantly reacting to the player's actions. The designer should mess with the player's perception every chance they get, set up rules then break them when the time is right. Repetition should be avoided no matter what, as that is one of the main ways to kill horror in a title. Give the player freedom to explore a wide environment, instead of leading them from one room to the other.
The design doc I'm writing is basically taking all my thoughts on horror design and trying to create the ultimate horror title. As I want to make a game that would scare even me, the guy who came up with the idea from scratch :)
I am currently replaying Silent Hill 2 (HD Collection) and I must say it still does the trick. Something I like about this game is the normality of the character who I can easily relate to.
Also, what's great about this game is that there are tons of disturbing things that have the player's imagination go wild, in addition to pure horror elements.
I don't think the cutscenes take away from the game. They can be well used. Actually when you have a scary or gory cutscene and you don't know when you're gonna be given the controls back, the effect can be quite horrifying.
On a side note: it's a problematic point I know, but the clunky controls of Silent Hill 2 contribute quite much to the game's scariness (especially the inability to quickly turn 180°).
I don't think the controls in Biohazard and Sillent Hill are clunky with the intention of handicap the player weaker than enemies and unable to fight, because it's definitely easy to fight enemies in these games. The point is to slow the overall pace of the game and the combat.
Piramid-head, Tyrant, T-101 and Birkin are all very slow foes. Slower than the player, and with very delayed reaction times. The point is not making the player weak, but keeping pace of combat inside the suspense territory, for the player to be able to concentrate not solely on combat, but have just enough time to think about it, what should the player do, what can go wrong.
Horror and suspense should not end just because a combat started, without that it just becomes an action game. If player and enemies where twice as fast and mobile, there would be not enough time for the player's mind to dig deep into the situation psychological details.
It doesn't matter how disturbing a monster is in Dead Space, how much threat it presents, you're just searching for their limbs, there's not enough time to "get it".
Of course that's because those bosses are inspired by Friday 13 and Halloween, but still works, and there are other ways to elate the same emotions of imminent unavoidable danger/evil.
Its not about the game space, its about the player. Go outside the magic circle and get as much info as possible about the player, then you will know how to scare them.
Great article and great read. I am huge fan of horror movies yet find myself incapable of completing most horror games just due to massive cowardice on my part - it is the Agency that makes the experience that much more terrifying for me.
Lone Survivor does some interesting things around your third point about Doubt in games. I have yet to complete the game but it is doing a very good job of blurring the line between what is real and what is not. It also ties in nicely with point #1 about Normality: the moments of doubt are when this post-apocalyptic world suddenly shifts into seeming normalcy. There is a scene early on in the game where you stumble upon a party where people are casually listening to music and drinking cocktails while you have had to evade and fight unimaginable monsters to get there. It's a great scene and is more disturbing on some level than the more traditional horrors that precede it. 'Lynchian' get thrown around a lot, but in this case it applies quite nicely.
I think developers would probably be drawn to make something like "Tales from Silent Hill" or some sort of Lovecraft deal with a location-based horror revisited over time/ by different characters, but you could just as easily have them be totally unrelated. Stacking concepts and characters like that would probably make players feel a lot more vulnerable, particularly if you combined it with the sort of nested endings you get in jrpgs like Metal Saga or NIS games.
It reminds me also of the changing controllable characters in Sanity's Requiem.
- You have to provide something to keep the player entertained while you're building up (that normalcy). Now you're building two games (or at least one and a half).
- I'm not sure you can sustain the horror when the player can just drive away. Now perhaps as it spreads it can get harder and harder to get away - your car just dies, or hulking beasts that can stop cars. Since you already know the area (from running around in it earlier) you still have a slight advantage you don't normally have in horror games.
Deadly Premonition tried this, and it mostly failed. I played it, but it certainly wasn't for the horror - which was contained to (mostly) closed in linear areas. Red Dead Redemption: Undead Nightmare took the tack of literally taking an existing full open world game and bolting zombies onto it. Fun, but not very scary. Alan Wake also fun but not scary, and Resident Evil hasn't been scary for decades.
Really the only scary game I've played recently was Amnesia. The slow controlled experience did the trick even when I could see how they were manipulating me.
I think the only way to make an open world work in horror is to make it a dynamic open world. That way, you can establish early on there is an open world, and when you need to, restrict it significantly. Silent Hill sort of does that in a few cases by dropping you in an apartment building. All of the doors are interactive, so the building seems really big, but most are jammed shut. You go from door to door hoping to find someone or something who can help you. Then, when you fall into the slightly creepier dark world, everything changes. Changing the maze while the player is inside it is very disorienting. They are lost in a newly hostile environment they just managed to get a handle on, and they are wondering who it was that unlocked all these doors (and why would someone do that?) Both give the game the appearance of less linearity and so, as per the other points, more horror.
In an original game, maybe you could subtly change the handling and acceleration of cars depending on some element. Or periodically change the controls and then change them back based on an invisible insanity stat (imagine inverting the Y axis for 1 or 2 seconds with no visual cue when a monster jumps out at the player). Do it right, and you make a mild scare and outright hectic situation. Although do it wrong (and too often) and your audience says the controls are bad.
In lieu of fingers, sound best conveys the sense of touch. After all, sound waves physically impact us (especially if we have a killer sound system). In a visceral horror game, then, sound is paramount.
IMO, Siren 2 is still the finest horror game to ever be released. It took the basic framework of Siren 1, removed a lot of the clunkiness and added some cool new ideas of its own (like the blind guy stages). However, Siren 2 came out 6 years ago. It's high time for someone to surpass it.
Just a thought.
http://research.microsoft.com/en-us/projects/homeos/
I agree Kevin, I think the most effective parts of the The Thing were dealing with the uncertainties of the other humans. The action was okay, but everything fell apart when things went from non-action to action. It was a shame seeing how scripted things were (cross this line, and no matter how much testing you do on a human, he will turn into the Thing because he's scripted to do that). But overall the human parts were unique enough to make the Xbox The Thing still memorable for me.
From the list: #5 I agree with completely. If you just have the player fighting mindless enemies one after another, then it just becomes a slow paced action title. Enemies should be able to hunt the player and mess with them. With so many horror games that have action in them, why haven't we've seen a game where the enemies can knock the weapons out of the player's hands or just simple grab them and use them on the player?
With weapon design I think a mix of weak and power can work to elevate the fear. Let the player bash some minor enemies around without much trouble. Then introduce something that can shrug off a knife to the head and throw the player though a wall. There should be a mix of combat and stealth, to force the player to decide between fighting or hiding.
Personally I think the key to future horror design is to take a page from the rogue-like genre. Introduce scenarios that will not play the same way each time, with monsters that are constantly reacting to the player's actions. The designer should mess with the player's perception every chance they get, set up rules then break them when the time is right. Repetition should be avoided no matter what, as that is one of the main ways to kill horror in a title. Give the player freedom to explore a wide environment, instead of leading them from one room to the other.
The design doc I'm writing is basically taking all my thoughts on horror design and trying to create the ultimate horror title. As I want to make a game that would scare even me, the guy who came up with the idea from scratch :)
Also, what's great about this game is that there are tons of disturbing things that have the player's imagination go wild, in addition to pure horror elements.
I don't think the cutscenes take away from the game. They can be well used. Actually when you have a scary or gory cutscene and you don't know when you're gonna be given the controls back, the effect can be quite horrifying.
On a side note: it's a problematic point I know, but the clunky controls of Silent Hill 2 contribute quite much to the game's scariness (especially the inability to quickly turn 180°).
Piramid-head, Tyrant, T-101 and Birkin are all very slow foes. Slower than the player, and with very delayed reaction times. The point is not making the player weak, but keeping pace of combat inside the suspense territory, for the player to be able to concentrate not solely on combat, but have just enough time to think about it, what should the player do, what can go wrong.
Horror and suspense should not end just because a combat started, without that it just becomes an action game. If player and enemies where twice as fast and mobile, there would be not enough time for the player's mind to dig deep into the situation psychological details.
It doesn't matter how disturbing a monster is in Dead Space, how much threat it presents, you're just searching for their limbs, there's not enough time to "get it".
Of course that's because those bosses are inspired by Friday 13 and Halloween, but still works, and there are other ways to elate the same emotions of imminent unavoidable danger/evil.
Lone Survivor does some interesting things around your third point about Doubt in games. I have yet to complete the game but it is doing a very good job of blurring the line between what is real and what is not. It also ties in nicely with point #1 about Normality: the moments of doubt are when this post-apocalyptic world suddenly shifts into seeming normalcy. There is a scene early on in the game where you stumble upon a party where people are casually listening to music and drinking cocktails while you have had to evade and fight unimaginable monsters to get there. It's a great scene and is more disturbing on some level than the more traditional horrors that precede it. 'Lynchian' get thrown around a lot, but in this case it applies quite nicely.