We had the chance to sit down with Serellan owner and development lead, Christian Allen, and talk up his bold new tactical shooter, Epsilon, at this year’s E3. This interview was conducted by Brandon Orselli and Cody Long, and we’re very thankful to have been given the time to chat with Christian about his game.

Without further ado, you can find our interview below, alongside the debut screenshots (above as well as below) for the game, courtesy of Serellan:

Brandon Orselli: So I’m here with Christian Allen, and we’re going to be talking about his new game, Epsilon–which runs in the Unreal Engine.

Christian Allen: Yeah! So, the team back home right now is working with the newly released Unreal 4.8. Epsilon is a squad-based game, so we’ve been holding off on doing AI work until the new version launched, which it did just this Thursday I think? The reason we held off on working on AI is that 4.8 comes with a new perception system for the AI, and honestly the whole update has been really, really good for indie developers.

It’s funny, but doors actually present a hurdle when it comes to squad-based tactical shooters. There are some games that did away with them altogether. The reason is that doors are a giant fucking pain in the ass for game developers. There are a host of problems that arise, whether it be breaching, or telling a squad mate to open one.

What we did to avoid this problem was: we built all the functionality into the doors themselves. The AI doesn’t actually do anything, the door waits for a command and reaches out to the AI. It’s like, “Okay, you’re gonna breach me? Well, you gotta go here.”

Cody Long: Oh, yeah. I heard Sony was putting out their new Smart Door this year. (Bad joke)

Christian: Haha, yeah. I think I might actually sell my doors on the marketplace, because they’re pretty fuckin’ impressive. But anyway, about Epsilon–We’ve been in development for about six months. We’ve got a small team, about five people working on it full time, and a couple of my industry buddies helping out here and there.

Craig Gilmore, who’s doing our story art for us, is going to be handling our cinematics. Cinematics will play out in sort of a Max Payne kinda way, with that graphic novel style. He worked with me as a concept artist on Ghost Recon, and now he does film and TV. He was a storyboard artist for The Walking Dead, and did a bunch of their scenes. Like, the scene where they got trapped in the revolving door, and the dude’s face got ripped off?

Cody: That was pretty metal.

Christian: Yeah, it was pretty metal. I actually had to turn away from that! I love The Walking Dead, but gore where people are getting their guts eaten and stuff, I gotta turn away. Anyway, so he’s done some really cool storyboards, and helped us develop our storyline. The name, “Epsilon,” references the team you’re playing as.

It’s a secret division of Interpol–ahem–I mean, a secret division of Unipol, because Interpol is trademarked. (Laughs) I did not know Interpol was a private trademarked organization. Can’t use their logo. I also found out the Red Cross is the same kind of deal when working on Ghost Recon. Anyway, you’re playing as this international team working for Unipol.

Over the course of the story, each mission will have a chapter of the animated story, and we’ve got a whole thing planned out for that to build to a full graphic novel. Since Epsilon is sort of my brainchild, and you know me, you find out that you’re basically doing everything wrong. You think you’re taking out the bad guys and saving the world, but really you’re just a tool for the man. We are excited to explore themes like the rise of Neo-Fascism in Eastern Europe, and the global oligarchy system, like the Bilderberg group. Lots of cool story potential.

Cody: Sort of like Metal Gear Solid 3, right? Like, the whole game you’re thinking, “Wow, I’m the good guy,” but then you get to the end and you realize you’re just a huge piece of crap, and you killed your best friend/mentor.

Christian: Oh, yeah. I like to do shit like that. One of my favorite things in Halo Reach is that, well, we kill you. You die, and there’s nothing you can do about it. You die, everyone you know dies, and that’s how the game’s gonna fucking end. There was a lot of people working on the project with me who didn’t want to kill off the main character, believe it or not.

I really like that idea, though, where you basically fight as long as you can until you get overwhelmed. There were other ideas, like, “Oh, you escape with Master Chief,” or you just go off and the player never learns what happened. But a couple guys who were working on it—, I think it was mainly Chris Opdahl, really pushed for that cool ending sequence. I think it was fucking brilliant.

Cody: It’s really powerful when you have an ending like that. For example, Red Dead Redemption ended in a similar fashion. Spoiler time, in the end you get gunned down in a hail of bullets. There’s a point where you realize, “Oh, I can’t shoot all these guys.” As a narrative tool, it’s really effective, because despite all the tools you have at your disposal, you still can’t win.

Christian: Yeah, I think that reality is something a lot of people try to shy away from. When I was working on Reach, there were a lot of people saying, “Who’s gonna want to play a game you can’t win?” Well, we’re just trying to tell a story, here. It’s about the sacrifice of the Spartans, all these men and women who fought to defend the world. Well, except Jun, Jun lives. If he died it would have fucked up the whole time continuum.

The team at Serellan.

Getting back on track, yes, Epsilon has been in development for six months. We’re working toward rapid prototypes. Deane is my lead artist, and she’s fucking awesome. She worked for Zipper on most of the SOCOM games, MAG; she was one of the level artists on Killzone 2. And she’s just fucking brilliant.

She never does what I say. (Laughs) The first level on Takedown, I wanted to do this sort of hacienda, with a central building and a courtyard. You can come in from any angle, and there are all these different entrances. I thought it would be really cool! And she was like, “Okay.” And she comes back two days later with the grey box.

She was like, “Well, I built your layout, but it’s not gonna be a hacienda. It’s gonna be this modern office building.” And I was just like, “O-okay, cool!” And then she made it, and it was the best looking level in the game. She kinda did the same thing in Epsilon, and you’ll see it in the video (Editor’s note: We’ll have that later). And it’s just beautiful–she does modern art on the side, too, so she’s a true artist. And she’s just like fire and forget from a development standpoint.

We have a lot of old industry folks working on the project, but we’ve brought in some new people too, and they’re all super talented. We had probably 100 applicants for the two positions we had open, and we chose two. Jorge programs the shit out of everything, and he doesn’t speak the best English, but he writes awesome code. Seth is our Tech Animator, and he calls himself an animator, but he’s basically a designer and scripter like me.

We’re really focusing on, y’know, modern, hi-tech visuals.

Brandon: So it’s a near-future kind of style, right?

Christian: Yeah. In Takedown, we decided to go super hardcore. We didn’t give you any information, and it was pretty hard to figure out what to do if you weren’t a fan of tactical shooters. Which is great if you’re a hardcore fan, but it wasn’t great for a lot of other people.

Cody: So it wasn’t very accessible?

Christian: Yeah. People didn’t understand, we didn’t do a good enough job of explaining why things were the way they were. Like, “I don’t see my reticle on screen! I don’t understand.” It’s like, “Dude, there’s no reticle.” “What do you mean? How do I aim without a reticle?” But you don’t have a reticle in real life.

Anyway, so we went with the near future style, with a HUD that comes down. You can turn it on or off, and it has different visual modes, so you can see in night vision and thermal. We’ve also got a movement-based detection system. We’re just really trying to focus on intelligence, and so when you start the game, you’re presented with a map of the level. You’ve basically hacked into the security system to get a layout of the place.

An in-game cutscene.

So you have access to all these cameras, and if you see a guy, you can click on him and you’ll get an icon on the map that shows where he is. But that’s only if he’s in view of the security cameras, so it’s not this perfect strategy. You’ll also have stuff like the opti-cam, that lets you see underneath doors. There’s also a cool device that lets you see through walls. The problem with that is—if someone has a jammer, they could literally be standing right next to the door holding a shotgun and you wouldn’t see them until you walked through.

We’re really just thinking on the tactical side, and we’re less about being super hardcore and kicking the player’s ass over and over, and more about giving them lots of tools and information about the game. I like to call Epsilon a kind of hybrid stealth-action game. Not stealth in the, “Oh if you get seen by the camera, you lose,” or, “the only way to get through this section is to hide in the corner,” kinda way, though.

It’s gonna focus more on thorough planning. There’s going to be a dedicated planning phase, and co-op partners can share their plans, and revise them on the fly. This is a great feature because in other games with this option, you run into the problem where you have a plan and then it fails. Instead of restarting and completely revising the plan you had originally, in Epsilon, you can change it as you go.

What we want to do, ideally, is release a polished core game, and then expand upon it. So instead of releasing a half-finished experience, we’re releasing what is essentially the full game, and then we’ll add stuff like gadgets, maps, and weapons as we develop them. We’re also trying to make a game where going non-lethal is a legitimate strategy. We’ve already got tasers, smoke and gas grenades, EM chaff grenades, etc. It’s hard to balance because not a lot of people are going to play that way, but I still want to include the option.

I think map editing is important, too, and we’d like to include that in the release as well. We tried to do that sort of thing with Takedown, but there were some issues with Unreal 3, and it was just a big mess overall. But Unreal 4 being open as it is, all you’d have to do would be sync to the same version we release on, and boom. We’d also ideally like to have contests and stuff where you can show off your maps.

I also feel like it’s important to get some of the less hardcore people involved with playtesting too. It’s great to listen to your ten to fifteen thousand hardcore fans, but sometimes they don’t give you the constructive feedback that you need. Especially on the technical side—they know how to open up NAT settings, they know how to use Ventrilo.

When we launched Takedown, we were like, “Well, Unreal Chat sucks, Steam Chat sucks, do you want use to implement a chat feature, or should we just work on other stuff?” And our hardcore fans were just like, “Ahh, fuck it, we just use Ventrilo, no big deal.” That might’ve been the case for the first ten thousand or so people that bought the game, but the next fifty thousand people were fuckin’ pissed there was no chat function in a squad-based game. (Laughs)

In-game screenshot of what environments look like, in Unreal Engine 4.

Cody: Awesome. Now, I was a big fan of SWAT 4, it being the first one in the series I actually got around to playing. SWAT 4’s AI made me want to break my keyboard in half. You mentioned some AI hurdles you had to overcome earlier—do you think Epsilon will have good AI overall?

Christian: We’re working on it right now. Basically, the best way to make the AI work is to give the player as much control of them as possible. They’re not human, so they’re always going to frustrate the player, that’s almost universal. In Epsilon, at any point you can override your squadmates’ commands and have them go somewhere, or fall back on you.

It’s a real challenge because a lot of it comes down to perception. When you send your guys into a room and they just get murdered, they’re stupid, even though you ordered them to do it. When your guys just murder people with impunity, they seem like robots, but if they don’t and they get killed, they’re stupid.

I’ve never had anyone say the AI in my games was great. The fact is that a robot is never going to be able to do everything a human can, in regards to gameplay. The only thing you can do to alleviate this fact is to give the player the maximum amount of control over his AI companions as you can. With Unreal’s new AI systems, we’re able to do a lot, but it will never be perfect.

Brandon: Alright. One last question, and then we’ll wrap things up. Do you think your long-time fans will be able to look forward to Epsilon? And does it have all the hallmarks of your previous tactical shooters?

Christian: Yeah, I think so. We definitely want to appeal to that audience, but we want to make the game less how good you are with a mouse, and more how good you are at thinking about your surroundings and reacting accordingly–as well as how good you are at using the tools at your disposal.

We won’t ever hold the player’s hand or make linear cinematic button pushing bullshit, but there’s a point where a game withholds too much information from you. We’ve made that mistake in the past, and with Epsilon we’re trying to find a happy medium.

Brandon: Great. That’s all the time we have, but thanks for chatting with us. Looking forward to hearing more about Epsilon as it is developed.

We’d like to thank Christian Allen and the rest of the team at Serellan again for spending time with us, so that we could talk about and showcase their game.

Niche Gamer regularly interviews developers on a variety of subjects—if you’re a developer and want to chat with us, please contact us!