SOMA didn't scare the scuba suit off me, but I did find a creeping sort of potential in its soaked-to-the-bone corridors. Amnesia: The Dark Descent 2 this ain't. Or at least, it's not aiming to be. Currently, it still feels a lot like a slower-paced, less-monster-packed Amnesia in a different (though still very traditionally survival-horror-y) setting, but Frictional creative director Thomas Grip has big plans. I spoke with him about how he hopes to evolve the game, inevitable comparisons to the Big Daddy of gaming's small undersea pond, BioShock, why simple monster AI is better than more sophisticated options, the mundanity of death, and how SOMA's been pretty profoundly influenced by indie mega-hits like Dear Esther and Gone Home.
RPS: So let's get the zombie skeleton ghost elephant whale in the room out of the way first. Why an underwater setting?
Grip: Actually, it was a pretty quick decision. Me and Jens, who’s the other co-founder of Frictional Games, we used to live in the same city, and we’d meet up every other month in a park. Just after we’d released Amnesia, we were were talking to each other – what should the next game be about? He says we should set it underwater, and I said, oh, I was thinking about that too. It was that sort of decision. We’d both been thinking about it a lot.
It’s a very cool setting, because normally, you don’t really think about how weird the ocean is, with the big pressures around you deep below and all the weird creatures that live there and so forth. It’s a very hostile environment. It’s hostile in a different way from space – even more hostile in a way, because you have this enormous pressure difference just outside you. That builds this claustrophobic feeling. It just felt right.
It was a great decision in hindsight, because we felt like if we were going to set it in space, there would be tons of other games in space, like Alien: Isolation being the big one. It was a good way to differentiate ourselves, and we just liked the setting.
RPS: The second you revealed your setting people started bringing up BioShock, which was set underwater and had some horror overtones, even if it wasn’t actually a horror game. Even if your game is pretty hugely different, are you worried about inevitable comparisons and callbacks?
Grip: I don’t know. We’ll have to see. I think that one of the big differences is that we have a sort of Aliens-like atmosphere in our game, the overall aesthetic. BioShock has their art deco style. It’s almost a fantasy setting in BioShock. They don’t really care about the ocean as such, other than it just has water in it. We’ve tried at length to make sure that we think about pressure differences and what sorts of things you’d have to do to build something underwater. That’s an integral part of the game.
The whole aesthetic is different. In BioShock, at least in the first one, you’re not much actually in the water. Soma, about 50 percent of the gameplay takes place in the water, in sunken stuff or on the ocean bottom. It’s a lot more focused on being in the water than BioShock is. But it’s hard to see, from this gut reaction, how this is going to shake out. I hope, at least, that not everyone will directly compare it to BioShock. I don’t think so.
RPS: You were mentioning a lot of stuff about pressure differences and whatnot. How does that get involved in gameplay?
Grip: The thing that we’re focusing on with our games now, and it’s a big design cornerstone, is that we don’t want to try and gamify too much of it. One thing we could have done with pressure is add puzzles around it – oh, before I go into this airlock I need to think about that sort of thing. Then it ends up in this abstract system. You’re not emotionally involved in it. Instead, what we’re trying to do – and it’s a bit more difficult – we’re trying to create this emotional background. You’re thinking about it in a more realistic way than you might have done if it was a sort of puzzle system.
There are events and so forth where it plays a big part, like in the sequence that you saw at the end of the first part of the demo, when the water comes crashing down and so forth. But otherwise it’s sort of like a lingering fear. You’re not sure when it’s going to happen. That ties into a lot of things we have in our game.
It’s sort of like what we did in Amnesia with the enemies. They’re pretty simplistic in their behavior. There are a few behaviors you can work around, like hiding behind doors and stuff. But half of the gameplay comes from just hearing the sound effects and trying to guess what’s happening next, what’s going to be around that next corner. That’s really interesting stuff. We’re trying to use that sort of feel for other elements of the game, and not just monster encounters.
RPS: When you say you’re trying to apply that feel to elements beyond just monsters, can you give any examples?
Grip: I’ll say what I can without spoiling anything. For instance, we have our consciousness stuff. That’s a big thematic. I’m not sure how much you got that because it depends on how much you connect into it. It’s something you might have noticed when playing. It’s an aspect of the game at the start, the story, because you’re missing a bit of backstory and you’re dropped into it directly.
For instance, you have you have a robot at the start of the game, and you can find out that it’s the same person as another character that you met with in the game. And that, then, has relevance to stuff that the player is going to find out later in the game. You can use that experience to say, oh, shit! Now I get why that was happening there. You can use that in your own experience of certain events that happen later on. We use it as a sort of reflection on that sort of thing.
Then, what happens is that, because the player has been aware of these sorts of things happening in the narrative sense, when they’re close to something similar happening to themselves, then they can project from what they have learned before – oh, this and this happened in the story – and project that onto things that might happen to themselves. Okay, how am I going to deal with this in order to stop this sort of thing? It’s difficult to talk about, because there’s a couple of mysteries that I really want people to figure out for themselves.
But it’s that sort of thing that we’re aiming for, just like with the monsters. You have this narrative backstory – you know they’re sneaking around – and although the monsters are pretty simplistic, with a few basic characteristics… When you think that the monster is near, you start thinking about those things in the backstory. If you think that something has to do with a situation that’s similar to the one that you experienced before, when you’re close to something like that happening, you start thinking about that. Just as in Amnesia, there might be ways for you to avoid these things happening, or at least sort of dampen the blow. You can start to project and think about, how am I going to advance from here on? It’s that sort of thinking. It’s a bit vague, I know, but the monster thing… That’s easy to do. The other idea requires a bit more buildup.
RPS: Are there multiple story paths, then?
Grip: No. It’s going to be more like The Walking Dead. There are going to be specific events. It’s pretty linear in a sense, but there are always [bits of story]. For instance, in the demo, did you find the girl who committed suicide? There are certain things you can notice in the environment that lead you to finding her, and finding her is going to reveal a bit more about what happened. It’s more in the sense that you can find side paths and things to figure out. That might change how you view certain things.
But it’s not going to be some sort of Heavy Rain thing, where everything branches out a lot and different characters could die depending on the choices you make. It’s going to be pretty linear, with a certain path that you walk on, but there are different choices. In the end players might have very different experiences, even though they went through, in broad terms, the same path.
RPS: You were talking about the fact that Amnesia had fairly simplistic monster AI overall. Are the monsters in Soma going to be any more sophisticated?
Grip: No, they’re not. What we’re trying to do, though, is that they’re going to have different ways of interacting with the player. We’re not overly reliant on things like really sophisticated AI models. I think it works well when you have simplistic models that can work in many different situations. Having some AI that works really well, but if the player screws up a little bit, then it goes really wrong.
We’re going to have something pretty much like in Amnesia, but the enemies are going to be a lot more varied. The monster encounters are really short in the demo, but they have some special abilities that ends up making life for the player a bit more worrisome, more than the Amnesia monsters.
We also have some enemies that aren’t really outright hostile to the player, unless you do certain things. That brings back how I wanted the player to imagine things. For instance, these enemy creatures that aren’t hostile, or are hostile under certain conditions, it’s sort of vague, exactly what those conditions are. The player is going to have to make up their own mind, and depending on story stuff that happens before, what actions it might be that will make these creatures aggressive. It’s that sort of stuff, where you have different things happening to the player depending on what they do. It’s on that sort of level.
RPS: Briefly, I wanted to circle back around to the AI machine arm thing that was at the start of the demo. The one that's clearly dying, that mentions how scared it is and that it's happy to see somebody named Amy (who is clearly not me). What’s the context of that? What’s happening in the world, in the way that the game world functions, from a story standpoint?
Grip: What happens is, the player in the game is not an Amnesia character. He has a well-defined past. When you get to PATHOS-2, which is the research station where this all takes place, you’re not entirely sure how you got there. There is a mystery in that.
The interesting part, especially for players who play the game from the start, is that they all have very different explanations, going into the game, for why they ended up there. You can understand the reason from the very start, but I think that some people won’t really think about it. That’s very much related to the thematics that we’re working with. From the start, people start thinking about your background.
The robot that you see has connections to that thing. You’re seeing a robot that’s just talking in gibberish. It seems to think that you’re someone else. What we also have, another thing that adds into this… When you interact with the corpses that you find, you get a little snippet of dialogue – the last 10 or 20 seconds of the dead person or machine’s - their last 20 seconds in life. What we also have, and you can find hints about this on the map, are certain black box devices, that all the personnel and machines have installed. You can go around and figure things out.
RPS: That’s actually really fascinating. Most games tend to deal with two states for enemies, or any type of character – alive and dead. It sounds like you’re hitting a last little in-between state. That last twinkle of life that most games (or works in other mediums, really) don't even consider.
Grip: Yeah. When you have audio logs in a game - we’ve done this in Amnesia, too - there's a worst-case example of adding diary entries where it just makes no sense at all. You just have to go along with it. A lot of games do that. In, like, BioShock you can find the personal thoughts of some character, and you have no idea why they would be sitting there recording it. Why is any of that here? So we’re trying to make sure that all the audio logs and notes that you find really feel like they belong in the world.
As a design challenge to ourselves, we tried to make it so just noting where the audio log is, you can figure that out. You don’t have to wonder, why is this audio log here? And then you find out by discovering some extra story. The corpses are a good way of doing that. This sort of device also ties neatly into the story in several ways. It’s one of those things where we started from two different directions and it all came together in a nice way. It is a bit morbid. But all of our games are morbid. I think that’s to be expected of us, I’m afraid.
RPS: That notion is also kind of powerful to me. The moments before someone’s death can be extremely profound.
Grip: Yeah, but they can also be fairly trivial. Just slipping on something, maybe. You might think there’s always going to be some kind of elaborate setup, but some of the audio files you’re going to find, you’ll just find them saying, “Uh?” Or, “Oh, I forgot to…” I read something recently, I forget what it was called, but a sort of true crime story. This person’s last words were something really trivial like, “Oh, don’t forget to feed the cat,” or something like that. It’s weird what people think about.
It’s also interesting… There are two points to this. First off, we always consider ourselves the heroes of our own personal story in life. It’s just like in the movies. The main character never dies. We don’t expect ourselves to die like that, even if it’s really obvious that we’re going to. Another thing is, myself and the whole team, we’ve been very careful that we’re not just making a gratuitous torture-porn fest. We want to make sure that this has meaning. This isn’t just stuff that happens on a daily basis in this world. It’s not here just to be used for fun. We tried to be very aware of that.
RPS: On the subject of creatures that are not necessarily aggressive, I feel like that’s potentially interesting territory for a horror game. That could create instances where the primary source of what makes you feel safe at one moment becomes a thing you’re terrified of the next. Possibly more terrified than you might otherwise be, even. Like, "Oh no! Monsterbro, whyyyyyyy?"
Grip: Have you played Outlast? I think they could have gone a bit further with it, but they have characters you meet in the game who aren’t hostile. They just stand there and look very creepy. Those are the most interesting encounters, to me. They’re very sort of scripted. Nothing happens if you try and push the limits of it. But it’s interesting, because your mind races as you just see someone staring out the window.
It goes back to simplistic AI. Once you have a system that the player needs to gamify… Maybe he only takes cover behind barrels, or something like that, and then tries to attack me from behind. Even if that’s really sophisticated AI, it’s also a fairly simple system to take… You don’t have to take into account what sort of person this is. You just know that he likes to hide behind barrels. That’s what this person becomes. He’s the barrel-hider. That’s all he does.
But in Outlast, you have this person just staring out a window, and you’re like, shit! Is he going to do anything? Why is he standing there? It’s much harder to get a grip on it. Even if it turns out that he can’t do anything, in that moment when you see him you don’t know that.
What happens to thinking in games, it’s the opposite of reality. The more you interact with a character, the less real and complex they become. They become more and more mechanistic, an automaton. In real life, of course, the more you interact with someone, the more you understand them. But in games it’s the opposite. You have to be very careful about the interactions you’re going to do. That’s why I think it’s interesting to cut back on the actions instead of having this really sophisticated behavior. That sophisticated behavior tends to reveal itself a bit more. If you can tone down the things that the player can do with different creatures, the possibilities in their mind open up a bit.
RPS: When I was playing the demo, because the interactions were so limited, I kind of found a space of safe predictability. There was obviously a lot of very heavy atmosphere, and it was meant to be very unsettling. But once I got into the groove of things, doing the same thing repeatedly, I kind of stopped being quite so afraid.
Grip: I think, to defend the demo a bit, there’s some buildup that’s missing from the demo. We have a bit more that we need to do to fix that. I’ve also seen one of the players that’s played come out to be really upset by this. "Shit! Is something going to happen soon?" He’s starting to panic more and more heavily. For some people this sort of works. But as you say, we have to be careful exactly how we do this. For instance, in the very open-space level you presumably found, where you’re walking on the ocean floor, that’s probably where you found the least interactions allowed. That’s where you sort of got into the groove, right?
RPS: I sprinted through that part [laughs]. By that point, I wasn't scared at all. I knew nothing was actually going to hurt me. On top of that, the story and the world hadn't really grabbed me. Nothing stopped me from thinking, "This is just a videogame, just an elaborate funhouse."
Grip: Ah, oh, you sprinted! You’re playing it wrong! What happens there, actually, is that there are several encounters, but in the demo we misjudged a few things. Some of the players who played, they got all those interactions, the three or four encounters that are supposed to keep you active. If you go off-path too much, though, to go investigate, it can become very deadly. Some of the playtesters did just that, and then you get a much more intense experience. You know, oh, shit, if I go along the wrong path here, something is going to get me.
Whereas if you just go forward and don’t get into it, that’s not going to work as well. It’s a very thin line, because if we constantly bombard you with experiences, that’s just going to get repetitive – oh, I know what’s going to happen. You have to follow that thin thread. That’s something we’ll try to improve more as we work on the game and have more people play it. We’ll see what we can do to make sure everyone winds up in the right space, or as many people as possible.
RPS: I have heard from other Frictionalites that you were pretty influenced by games like Gone Home this time around.
Grip: I wouldn’t say that we were heavily inspired by them, but I had the entire team play it. Everyone’s played it. They used that sort of environmental storytelling – the variations on the different notes you can pick up, the places you can find – in a very good way. It’s a good inspiration from that standpoint.
But it’s also that Gone Home builds on a lot of things that we were inspired by from the past, like the BioShock games and the System Shock series and all those sorts of games. There’s a sort of System Shock heritage there in games that we’re all inspired by. Gone Home is one of those, of course, and it was very interesting to play that game. So yeah, it’s an inspiration, but it’s not like we saw Gone Home and said, yeah, we’re gonna make that! But it’s a very good game. It’s something that has a lot of design lessons to learn from.
RPS: Gone Home’s lineage is very interesting in that respect, where it’s rooted in a lot of triple-A stuff, but it expanded on triple-A design ideas in a much more focused way. I feel like that’s happening a lot with games in general, especially in horror. A lot of horror is moving toward more of the indie side of things. Are there any other indie games that have inspired you?
Grip: Gosh, let me think here. There are tons of indie games that I’ve found inspiration from. Dear Esther is one of those big inspirations. The interesting thing, I felt, with both Gone Home and Dear Esther is that they cut off so much of the fat from a game. You normally think, oh, walking through an environment, that’s not going to be very interesting. I need to have stuff happening all the time. I need to put enemies in here.
That’s what they said about Gone Home. You’re just reading notes for all this time? That’ll get boring pretty quickly. But with those games, you understand that [you can do more with less].These elements are more interesting on their own.
It’s something you can think about when designing games. Okay, now we can have a Dear Esther section, where it’s going to just be about the environment for a little while. Or now it’s going to be a Gone Home section, where you’ll just read notes, and we don’t need to have enemy encounters during that time. It’s a very interesting thing to have in your toolkit.
Other games that are very interesting… A lot of Brendon Chung’s games, like 30 Flights of Loving and Gravity Bone. Those are very interesting, too. They also have that same way of playing, and they’re super interesting. 30 Flights of Loving especially was great in many regards. That’s been an inspirational source. There are so many, I need to poke my brain and see what pops out. You have Proteus where you just walk around in an environment. I find it interesting, just from a design standpoint. It’s sort of like when you’re making experiments. One thing that you want to do is limit the parameters for the experiment, because if you have too many knobs to turn, it’s hard to understand what has and hasn’t been affecting it.
What these games all do - and as you said this is happening a lot in the horror genre - they cut away a lot of things. As a game designer, it’s very interesting to just walk through and feel that - shit, this is working all on its own. There’s a whole game built up from this. You can learn a lot of design lessons from that. Another game that also does that, and has tons of problems along with it as well, is Slender, which I also thought was really interesting. There was another game before that that was called Tide, which has the same sort of premise, where you just walk around picking up notes. It’s also very interesting, a very condensed sort of horror experience. Those have been interesting to play around with.
Check back soon for part two, in which we discuss horror gaming's massive screamity-scream-scream-about-how-scared-I-am YouTube culture, why Frictional's choosing to *not* embrace it, procedural generation, and the future of horror gaming.