Q&A: A Legendary Talk With Spark's Garcia-Shelton

Independent Southern California-based Spark Unlimited was born out of the Medal of Honor developer diaspora, making its debut in 2004 with the release of Call of Duty: Finest Hour. The studio recently completed its first original game, the alternate history WWII shooter Turning Point: Fall of Liberty for Codemasters, and is in the midst of developing its next original title, Legendary, which has since shed its subtitle The Box. The first-person shooter takes the legend of Pandora's Box into modern-day New York, complete with mythical creatures bombarding The Big Apple, and will be published by indie-minded Gamecock Media Group later this year for PlayStation 3, Xbox 360 and PC. In this in-depth chat, the game's producer, John Garcia-Shelton, sat down with Gamasutra to discuss the team's use of Unreal Engine 3, the balance between linear and dynamic elements in game design, and the challenge in succumbing to what's expected for shooters. The Tech First, are you guys using your own engine or are you licensing one? JGS: We’re using the Unreal 3 engine for the tech. I’ve been asking a lot of people recently how they feel about proprietary technology versus the licensed stuff. What’s your take on it right now? JGS: Well the thing I like about licensed technology, especially when it’s got a lot of tools in it, is that it allows you to invest your resources in what’s unique about your game. The catch is that when you’re spending so much time on making what’s generic – a renderer, collision systems, the physics, all these things – you’re not actually concentrating on what people want when they buy a game, which is to play something different – something that’s just unique. I like to concentrate as much of our resources on those things that I’ve identified as things that are going to be fun and different. So if I put everybody on that, I know I can take it to a place that other people haven’t taken a game to. And that’s what I’m after – something people play that’s a little different, that they feel is different than their other games. It seems really difficult – especially now – to differentiate yourselves as a first person shooter because there are a million around. JGS: It’s a challenge. The challenge with differentiating yourself as a first person shooter is that it’s not safe. Part of the problem is that people are used to what’s already sold and they want to be safe and make something close to that. And, what’s identified as quality and what’s good is such a limited scope right now: how are your textures, how’s your shading, how’s your lighting? It’s all these things that really force you into making one type of game: it’s corridor, small spaces, not many enemies – limited scope, which also ties up how many things you can do with that space. You can do a lot of different ways and mess with that and change it up, but at its core, you’re still creating limiting technology. What we’ve tried to do is say, uh-uh. That we’re going to go somewhere a little different – we’re going to go really big on the creature padding and make sure that it’s different. But it’s not so much about doing what other people have done, because I think you’re going to have a hard time winning and standing out with that. Win or lose, we want to be going after something that’s different; I want us to be taking on a challenge to make something that people who play it feel that this game isn’t like any other game. Every first person shooter’s going to have similar aspects: in the end, you interact with things with your weapon, that’s the interaction of a first person shooter with its world, so in its heart you need to deal with that mechanic, but outside of that there’s still so many different things you can take to make it a different experience. I think a lot of times people go with what’s safe because that’s the easiest thing to get through and the easiest thing for people to understand, but we need people to take stabs at different things. An example is Grand Theft Auto: it wasn’t the best looking game, but, man, it was awesome. And that changed everything. It had to be really difficult to get that through. I bet they had a really tough time because people were probably going, it doesn’t look as good as every other game – people aren’t going to be used to this. And they were probably constantly trying to move it towards other things that they were comfortable with. But you hit with something like that and you can hit so big and people can be so happy that you put out a game like that, so we’re really just hoping that we can make something that will touch enough people, that we can be proud of what we worked on. The Mechanics One of the things that seems interesting is that you have to grab essence from these creatures that you’ve killed manually, instead of just running over it. What’s the rationale there? It just seems like an extra step? JGS: It’s actually gone through various iterations at different times to actually tweak that mechanic. The idea is that we don’t want it so run-and-gun that that becomes such an easy process that it lessens the value of it and lessens the value of the fighting. There are enough mechanics in there for the player to do that in a controlled way and still succeed and get ahead. It’s a good way to get players to think about their health and think about their healing and how they’re using it, and a way that doesn’t lessen the value of those mechanics. You need to establish things and then ramp people up on it so they understand the goals and understand the obstacles, but still then that they can accomplish things. It fits into that world that we have things a little more controlled in that regard. But that kind of thing still gets play-tested and maybe we’ll tweak that and open the range on it – it used to be that you had to look at it and now you can absorb it anywhere around you, so we continue to tweak those things. And actually it’s going to get a little better because we’re going to move the text prompt off of the screen and make it part of the HUD, as far as in absorb health. So you’ll get this thing saying if you go in an area where you can absorb it, you can. The text prompt’s actually a little misleading because you can absorb animus without looking at it – but you do still have to choose to absorb it. It seems like it’s one of a few design elements that’s very deliberate. The way you choose the weapons is with the d-pad, even with the melee. You have to mean to choose something – it’s very intentional. JGS: There are some things which are going to deal with player fatigue, such as when you’re running out of ammo, switching from one weapon to the next, but, yes, to choose those things is important, because the weapons pay off so differently against different elements. If you take out the thinking you end up lessening the value of the experience, so I want players to think about those encounters so they have to decide what is the best way to deal with that challenge. And what it can lead to is things like – an example is you’re low on health and you’re fighting some Black Order soldiers, and a creature who a minute ago was an obstacle, something that was a challenge that you weren’t looking forward to, is now an opportunity. You start to be happy to see that creature come out because you can kill it and take its life force. We can really kind of flip by making some things more deliberate. We can make that payoff when some things swing your way that much more joyful, and bring your energy level and your experience of enjoying it so much better. Dynamic vs. Linear How do you design with all of this stuff in mind? I know you’re going for multiple ways for you to solve a scenario. JGS: You design by iteration. The thing about games is you can always say, this will be good, just set things up like that. But in the end, because of the 3D nature and because the player passes through it, it’s always got to be that path experience that defines if it’s good yet, if it’s fun yet, or if it’s not. So we take our stabs on how those encounters play out and then we do a lot of play-testing on it. Is this playing right? Maybe there’s too many soldiers, maybe there’s too few enemies. Maybe we need to swap out some of the soldier types so instead of just guys with rifles and assault rifles, we put in a shotgun guy with maybe a shorter range. You just try to get enough playthroughs and get some good focus-testing – because you’ve been playing the game so long you start to think, “Oh, that’s great,” and then you get some people in and they don’t think that - no, they’re not getting that. So it’s important to tie in all those aspects and just keep on doing new passes on it so you can really fine-tune it, because what we thought was great is mostly great, but it can always be better and we can learn from watching people play it how to make it really a good playthrough experience that has very little player fatigue. Player fatigue is when you may play something for five minutes and like it, but you get into the half-hour, forty-five minutes, you’re like, “Okay, I need a break from this, I need something different.” And you try to break those things up so the player doesn’t realise they’ve been there for two or three hours, playing the game in one sitting, so it’s important to really try to grind those things out and polish it up so that experience doesn’t happen. It seems to be straddling a line between gated linearity and multiple outcomes. I think Resident Evil 4 did a really good job of that, but where’s the balance for you? JGS: The balance is that in pure dynamic games, you get a game of eights. All the intensity hits eight at its best, because everything’s so alive and changing and it can go through so many different permutations that it doesn’t necessarily hit a high point. Linearity allows you to focus people into a spot where you can really control the moment and create these high punch moments, these dramatic entrances or dramatic deaths, these energy-changing events. That’s what that linearity buys you, so it’s important to mix both if you want to have that rollercoaster or amusement park high action game. But still, you need to have the meat of something that’s live, because we’ve all played games where they have really great production values and everything’s dramatic and big, and yet you really start to recognise that you’re just going to the next scripted series, the next scripted series, the next scripted series, and you can play that game again and it’s going to play out the exact same way. We wanted to break that through while still having those nodal points where it’s still going to happen the same way, but that’s because we’re controlling the story and we’re trying to get you to the narrative completion. But for the most part we want to keep it live so those creatures are going to be behaving differently every time, because you’re in a different space, they target a different target every time. It’s a way we can get a lot of replayability, for fun – it will always be fun each time you play through it. With the characters targeting different people and moving in different ways – is it randomized paths? Do they select different targets or do they have a certain kind of pathfinding that changes depending on things? JGS: As an undercarriage for our AI we use Kynapse, which uses a nav-mesh, which basically finds a bunch of different paths so characters can take pre-determined paths which are then weighted depending on destructible objects – if the creatures throw things in the paths, that will weight things differently as well as if there are already targets in these paths. So there’s all these weights and measures that are put on those different paths and targets that will change every moment, and will allow the player to decide who’s their next target, their best target of opportunity. The creatures and NPCs go through value judgements – what’s my best target, my next target – but because they’ve moved, and the other creatures have moved, you’re constantly getting those things updated, so it changes dynamically. Will they generally have a preference for going after the enemy characters or you? JGS: Their first preference is saving their lives, so if they’re targeting something, and something else is coming at them, they’re going to switch to that target. After that it’s really going to depend on either the best target in my cover, because if I stay in my cover it’s a better position for me, or, since they’re following you and you’re moving forward, then they’re going to be moving forward, and they’re going to be basing it on that spot. But as they’re moving forward that’s getting updated. So they really have no preference other than what’s the best target to shoot at, because all things are going to try to kill them. So in many ways it’s not that different from if you or I were out there, because all these people are trying to kill you and all these things are trying to kill you – you’re just going to try and decide what’s trying to kill you a little more at the moment and say, okay, you’re the most important thing. It can create great moments, because there’s times where it can feel like I’m working with the Black Order, and then all of a sudden they start shooting at me, and it takes me a second to realise that, oh wait, we’re not on the same side, we just both have the same goal. And then it changes, and you have to deal with that. When I played it, there’s music throughout. A lot of games recently have been going without music and then bringing it in when there’s an event happening. Why did you choose to keep the music? JGS: We actually mix and match it and you’ll actually experience a lot of times when the music falls down. In this area, because it’s high action and the intensity for the demo, it does have more rock and roll stuff, but there’s times where we’ll pull out the drums and maybe just go – in some of the slower episodes, like episode four, which is much more creepy and it’s night time – we throw out a more ambient music. So the music’s going to change constantly. It won’t just be rock and roll-driven music sometimes, but it still comes from a basis in rock and roll – it’s the guitars, the distortion, and we use those things to play. But there’ll be times, and plenty of times, where it’s not going to be just that rock and roll, it’ll be different music – or no music at all, just sounds that, like I said, are driven from that music. The Team How big is the team now? JGS: Jeez. We’ve never been a huge team, I think our biggest we’ve ever been is, like, sixty, and right now we’re about that size. On average, for the couple of years we’ve been working on it, we’re anywhere from thirty to fifty. I’m pretty impressed we’ve been able to do it without – you look at some credit lists and they’re just huge – we’ve been able to do it with a pretty good senior group of guys that have made other games. We’ve made games, we know the mistakes we’ve made before, we just want to really focus, and everyone owns it a lot, which makes it a lot easier to make a game. They have to yell at me when I make certain choices: “What do you mean?! We’ve got to do it that way!” It’s just like any happy family – you’ve got to argue a little bit. Are you continuing going forward with the larger scale games? Are you looking at downloadable or handheld stuff? JGS: I think everybody’s going to go the way towards more downloadable games. My experience has been with console games, with Medal of Honor, and the first console Call of Duty. I’m probably going to be sticking in big games, and most of Spark will be sticking in big games. I think they will turn more episodic as time goes on, it just tends to be the way. There’s a reason we call these episodes instead of levels, and we try to condense them into moments. We’re all sort of making our way slowly towards that time where you’re not paying sixty dollars for a full game, you’re getting a couple of episodes for a certain amount – whatever people say is the right amount of money. I think eventually that’s where things are going. Handhelds? I think Spark has a lot of ideas, but the key thing is to be sure that what you’re doing you can do really good, and with this game we just want to be sure that we make a great game. Then, if there are other challenges, we can figure how to be great at those -- but right now, this is the one to make sure we’re great at.