---

Following article was written and contributed by Sylvain Douce 'Channie'

image: Official GDC Flickr (Creative Commons 2.0)

Game Developers Conference (GDC) is the largest annual gathering of professional video game developers, focusing on learning, inspiration, and networking.

This is the transcript of the Q&A session from the Level Design in a Day tutorial that occurred at GDC about two years ago. A broad range of topics were discussed; from portfolio and getting your foot at the door, to balancing and branching issues. Enjoy!

Note: Matthias Worch served as the moderator and presented these questions to the panel.

Panel consisted of:

Coray Seifert: Senior Game Designer Arkadium
Ed Byrne: Game Designer Uber Entertainment
Matthias Worch: Lead Designer Lucas Arts
Joel Burgess: Senior Designer Bethesda Game Studios
Neil Alphonso: Lead Designer Splash Damage
Jim Brown: Lead Level Designer Epic Games

Q1) When you're talking to your grand-mother, how the hell do you explain what you do?

Neil: Murder simulators...

Forrest: I just say I'm a designer... Leave it at that, I'm not going to try to explain...

Coray: I go with writer, because people actually know what that is...in general

Jim: A lot of times I just say I'm a software designer, and I think you should never do that because it is unfair to what you actually do.

Q2) Let's talk about resumes and portfolios. In a very general way, if somebody asks you for portfolio advice, what would be your answer?

Coray: Do a very small number of things, very very well.

Joel: Show videos of work. Like ideally shipped work you've done or something you've finished with a team and have callouts with timestamps and explain what you were doing in those specific parts. Because with level design there is often a lot of ambiguity about what this specific role this person was in. So if you say you worked on the level 3 of you know, Call of Duty.

I don't know what the LD culture is in that studio, so I don't know if you're building brushes or if you doing pathfinding, or if you're a scripter. So you want to broadcast this clear message about who you are and what you can do because we have our own definitions of what the job is.

Forrest: I actually think this video thing is super important because you're trying to sell yourself to somebody who is generally overworked and too busy and has no time to look at your stuff, and is swamped in applications constantly, so I'm always looking for something that, in just seconds does tell me "oh my god, I've got to hire this person!"

Jim: If you don't have a lot of professional experience to fall back on, I want to see that you are a) capable of finishing something and b) doing more than one thing. If you show me one deathmatch map that is really good, it better be the best DM map I've ever seen in my life or I'll just toss it. If you show me 10 deathmatch maps, they all ten better be the best 10 maps I've ever seen. But if you show me that you can do RPGs, and FPSs, and a little scripting, and some artwork, and some kind of dabble in other areas, they show a wide breadth of stuff. If you show me a bunch of BSP shells, and a piece of concept art, I'll know that you're good at starting something, but not good at finishing it, and I'm not going to hire you. So, just show me you're capable of doing a lot of different things and doing them well, if you don't have solid experience and examples of professional work to show.

Ed: I place a lot of value on how do you think as a candidate, so I've been telling people as they come up: correlating that finished level to support documentation, sketches that you've done, show me how you got from the initial idea of your level to the finished product and then I'll be able to assess you better as a level designer, I can place you better in regards of the team and your experience, I can tell whether you're visually-based, if your support documentation is all bullet lists, that's not a bad thing, but I'll know you're not a visual guy. For me I really need to have a lot of background about that level, and keeping track of all that support documentation and really putting that into a timeline so I can see your thought process, for me that's going to be the clincher.

Forrest: One of my favorite question is to pound on somebody's quickness. These are retired questions, officially at this point. So say you have a shotgun, an assault rifle and some grenades, and a lobster claw, describe for me now what the level looks like. Because, it's freaking random, and the best part about that for me as an interviewer is the answer is generally entertaining...

Matt: What were some of the best answers if gotten since you've officially retired the question?

Forrest: The worst answer was using the lobster claw as a melee knife. One of the good ones was somebody who wanted to set it up as a proximity mine, that would stun people by grabbing them onto their limbs.

Coray: It's really important outside of your portfolio, to establish relationships wherever possible with the people you want to work with so if you have a local IGDA chapter or you're working on mod projects, anyway you can interact with the people you might be working with, because if you're going to establish that relationship and show them you're an all-star before they're even hiring, when they go out, and start hiring, that relationship is already going to be there and it's just "show me that you can do the job"

Jim: And even if you don't get to know them, the fact you know about them, in the area and the company before you're going to your interview is huge.

Coray: A personal recommendation is extremely weighty. If Ed says "hey this guy is good, you should check him out" then that's night and day from just getting a cold e-mail.

Q3) I mentioned earlier how level designers are becoming so specialized. One of the problems I think people in the industry have had is that they don't know exactly what they're applying for. You guys have some experience and based on what you know about your company, how specific are you when posting a job about level design? Is it just generally a level designer or are people being specific about "We need a scripter, we need a builder", and is there even a difference at your companies?

Forrest: We just put up ads for level designers in general. But you know, our applicants all clearly have strong point. Generally it's a call for anybody that fits into that broad category and then based on individual expertise or something. That's why we kind of "fit them in" one way or another, but everybody does sort of need to be a jack of as many trades as is expected of level designers at this specific point, which is they have some ability with GL (Kaos' internal level editor), they're decent scripters, they know something about storytelling and narrative, but don't necessarily need to be great artists or something like that.

I can imagine a point in a not too distant future where it is divided into greater levels of specificity. And I think there are some narrative designer roles and stuff out there, I don't think we have any of those specifically...

Coray: They have those at Ubisoft.

Joel: Bioware, too.

Jim: At Epic I know we focus on something different in terms of when you see one of our Level Design job ad it's not going to list "You'll need to be able to do X, Y, Z" it's just more about what we're looking for from a person. Because being a level designer, it's hard to say what it is you're going to do and it depends on what project you're on, what phase of the project you're on, the team make-up you're fitting into, I mean there is a gazillion of factors that influence it all. So it's impossible to say what the level designer actually does.

Coray: It's a really good point. When you're hiring, your goal is just to go get rock stars. So if I see someone that is a rock star and I know he's going to be awesome but may not have the skillset I need for a specific project, I'm going to put him on my bench and wait until a fitting project comes up.

Q4) Is it better to work one's portfolio and build one or two things to a triple A level, or just start with a broad base and build up every part of your portfolio and just kind of keep it all going at the same none?

Neil: Neither one's better. It really depends on the studio you're applying to. Splash Damage is probably a bit more like Epic; it's a small studio, and people have to be kind of interdisciplinary because we have to juggle things around a whole lot more. But having to work for, say, Sony, which has much bigger teams, you won't have to use all your other skills, because you're literally not allowed to. The simplest and cheesiest response is to just go with what you enjoy the most and with what you feel you can really show your motivation in, and polish it to the n-th degree.

Joel: From a hiring standpoint, when we're looking at somebody's design test, it's often the case we go like "Alright: this person's layout, or this person's writing... Is this something I could have felt comfortable shipping? Or shipping with tweaks?" And maybe this person is weaker at another area, but I look for people who are smart enough, that they can learn that. And so often times, particularly with more junior level people, we bring somebody in who's just a rock star on one axis of what we need, and we will build them up. And that works fine too because everybody's going to have natural strength and weaknesses, and we have level designers; we don't have scripters, we don't have layout guys. It's a generalist gig; you're still expected to be a super ninja. And we can then give those people individual tasks that cater towards their strengths whilst they're still building up their baseline talent.

Jim: One of the most astonishing stories I've heard the whole day was when Joel told me he just recently hired someone who never built a level before. So what does that tell you about how you specialize?

Ed: There's something intangible about when a resume comes across your desk or you talk with someone; you're just intrigued. It's not necessarily about experience; it's about the potential you see in that candidate.

Coray: Be really passionate. Show you're really passionate, get excited in the interviews, get excited about what you're working on, and let that come through. Don't be crazy though! At the same time, be confident, not cocky. Gauge how the conversation is going, gauge how the email chain is going and then up your energy level or... chill out.

Joel: In this industry, it's assumed that you have passion, and it's assumed you are brilliant. Everybody you're going to work with employs a decent half of those traits, and so, what you'll really need, which I think a lot of us don't have, is the ability to deal with people well. Because it's a collaboration thing, like a vast majority of us have co-workers and need to creatively collaborate with them. And if you don't have that raw passion, then you're going to fade into the background with the other candidates.

Q5) Would you rather see something that's visually polished using somebody's else assets or do you want something that is on a whitebox level but is completely scripted to the n-th degree?

Forrest: Don't care. I mean as long as it's good you know... I think all of us are probably used to understanding what a whitebox looks like and how that comes to being a final thing and they could probably all understand too that's what you need out of a level designer. I'd still want to see sound structures and clear gameplay ideas, but you can show that either way.

Joel: There's something about using existing assets. When we are looking at level design candidates, they can build something with the G.E.C.K, from Fallout 3.

And of course, I'm pretty intimately familiar with all that stuff. And I consistently find that the people that get interest are people who have found some way after I've been involved with and seen hundreds of levels with those assets, to do something new. And for us in particular that's a useful trait. But, one of the best things you can do is to apply specific to the company you want to work for. This can be difficult, because if you're looking for a job, you're probably applying to dozens and dozens of places, but none of those places should feel like you're applying to dozens and dozens and dozens of places.

Forrest: It just means thousands of hours of work!

Coray: It's a great point, though. If I get that like, sort of B.S. spidey sense going off of someone just spamming me with a resume, it gets immediately round filed and I'm actually personally offended. But I would say on the original question that this might be indicative of bad habits both on my part and other hiring managers. When you have a lot of resumes coming across your desk, it's often possible to mistake whitebox levels and portfolio material with sloppy work. Especially for someone from HR, or someone who maybe is not a level designer but hiring for a level design position. If you have those polished assets, it's going to grab someone, it's going to get their attention, and then someone who is a level designer and knows about space and about creating levels and why you made the decisions that you made or at least if the motivations were there. You can at least get to that level.

Forrest: The risk is that if you use existing assets to make that art, then you kind of sabotage yourself. I've certainly seen that happen, like "Oh, it would have been a lot better in a greybox form..."

Joel: Particularly with art that I know, even if it's not from one of our games but at least from one that I've played, and I look at somebody's level they've got in their portfolio, and they clearly made something really beautiful, but when I look at it, I know it's beautiful because a great artist at Epic made that asset. And they haven't really done anything gameplay consequential with it. I won't say that's it's a knock, but it certainly doesn't move the needle up for me at all.

Coray: Do you think it would be worth doing both? Do your level in whitebox, take some pictures of it, then art it up and take some pictures of it, now you have both to present?

Forrest and Jim: Absolutely.

Joel: If I got a resume tomorrow from somebody who had like scripted a toggle while I play through the level to turn off the art and see the whitebox and then turn on to see the final product, I think it would be really educational for me, in terms of looking at how that person thinks.

Jim: The first question I ask to a person that makes it to a phone interview with me is: "Tell me how you made your test. Describe to me the process and why you made the decisions you did" If you can circumvent that process and show me in your resume, that's good.

Q6) It takes a clever person to solve a puzzle and an even more clever person to create it. Any advice on puzzle design and puzzle creation in a level? What constitutes a puzzle in FPS and RPGs games and then how do we go about designing them.

Joel: I think one of the things you see that a lot of designers need to learn is to keep it simple. There's a really strong desire to look at the work that we've done and to improve it. When it comes to puzzle design, a lot of times you put together a simple puzzle and then you'll decide that it needs to be harder or more complex, and the more difficult your puzzle is - I'm not saying the difficulty is not a good thing to aim for the more difficult as a designer is to make the puzzle not frustrate, and block a large number of players. When you think about the purpose of puzzles in a game like ours, it's a pacing device. It's a non-combat beat in the experience. And sometimes that role can be fulfilled by something as simple as a hidden lever. And it's not a puzzle, but it stills sort of tickles that same part of our brain, that would go like "OK, I've got a progress block, I need to figure it out, there are no dudes to kill" And you start thinking differently about the space.

And if that turns into a puzzle now, where you've got a sequence puzzle, or a collection matching puzzle or whatever the case may be; that's great. But sometimes, there's just that brief beat in the pacing, where you do something else, that's really good too. That's just a way of saying when you're making puzzle, you don't have to strive for making the greatest most riveting puzzle you've ever done.

Matthias: One of the most successful puzzles -if you can even call it a puzzle at all- we had in Dead Space 2, was in level 6. You just had to remove 3 batteries which would deplete the oxygen supply in an area which there was a fire. The reason why it took people a while was because one of the batteries was in the ceiling. And it is Zero-G space, and you can fly around.

And we just have a bunch of people walking back and forth on the ground finding the two batteries, understanding what it is they have to do, just not getting it. And finally realizing "wait a second, I can actually fly around!" and when people finally looked up, it actually reminds them that they can fly. People actually don't feel frustrated, because it doesn't take them more than 5 minutes or so, it just reminds you of everything the game has to offer, it reminds you of looking at the space in a different way.

Jim: The guys at Valve are masters at this. If you look at Half-Life 2 and their physics puzzles and the way they're putting them in there. I know for a fact that it's not an easy thing for them to achieve. They do so much usability research and tracking. They're constantly updating even today those maps to figure out where the difficulty spikes are, where people are getting frustrated, where they are dying too often... Because so much of level design is leading players by the hands, and guiding them into their own experience and you should never teach somebody by dying, that's horrible.

That becomes a shelf moment, so you really have to be careful to find the right balance between how difficult a puzzle and how engaging a puzzle is, because they're completely different things.

Ed: I've worked on Harry Potter games, maybe I've got the most experience making insipid puzzles, but one of the things I learned, especially for a kids property, is to put puzzles in areas that are not critical paths, so putting puzzles as a means to get extra stuff, extra ammo, or reach something you discovered, you can get to. I think DOOM did that a lot, with secret switches etc. You could just play right through DOOM all the way to the end and not actually find any of the secret areas, and then go back and realize how many little puzzles that you missed and little pieces of walls and everything. I think that you're right; it can become a shelf moment. Even the most perfectly crafted puzzle, may just seem to be too much for a player who's in it just for shooting guys, or picking up coins. I always try to keep all of the complex puzzles as side missions or things the player can hop to, but don't prevent progress.

Matthias: If anybody is more interested in this -the one talk I always recommend people for puzzle design and just general concept of affordance and broadcasting the various ways that you can use the environment- was by Randy Smith in 2009 and it's called - I actually like the session title, it tells you a lot about what puzzle design is about - "Helping your players feel smart: Puzzles as User Interface". And he actually does a very good analysis of Portal and he talks about concepts like affordance etc. The one thing that really sticks out to me is that very early on he has this chat about "Ok, 98% of all players actually figured out this puzzle, so 98% of people kept playing the game" You gate that several time, so you go from 98% to 98% to 98% and I don't know exactly what the number was, but he starts losing like, hundreds of thousands of players, just because, anywhere in that chain, one of these puzzles could be a shelf moment, and if that happens, of course people stop playing.

Q7) Why the hell do we keep making warehouses and what's up with boxes in levels? Can't we figure out with something else?

Forrest: AI can deal with crates really really well!

Jim: I always wanted to make DM-Cliche that is nothing but a square box, with crates and exploding barrels!

Matthias: Any stories you guys have where you have an idea of doing this awesome object or so that turned to be a box; or was a box the entire time?

Jim: It actually was a big problem for us in Gears, because the way the animation system works, all of our covers are exactly of the same height and all of our edges have to be close to flat so you can lean around and shoot, otherwise you're standing in cover but you're not in cover and you get your ass exposed.

So it has become, by the third game, a real difficult and interesting challenge to come up with 96-units high things that are not boxes. As developers we do like to make fun of it. In Gears 1, you run into a room which is an abandoned warehouse with a bunch of barrels stacked up and they fall over and you die... It was meant to be a joke but I don't think anybody died!

Forrest: The box warehouse is a terrible cliche as are sewers and so on... But they just hit a lot of the constraints; that's why I think we tend to work with them a lot. It's an easy solution, and that's why you see them a lot. Because you need constrained space; you need a point of entry and a point of exit, you need to have the space split up in some way, you need to have something that is clearly "parsable", like when you're making cover, the player needs to understand when he sees the cover object "yes, this is going to provide cover for me", because you don't want them to be getting shot through something that they felt was going to be good. A box warehouse kind of fits all of those needs, when the needs are of a magnitude of a dozen such as "I want to feel cool and experience cool things" - being in a warehouse doesn't hit that...

Neil: The first video you've shown of Killzone, which I'm rather responsible for, is a warehouse with a lot of boxes in it. But actually, I would say it's one of the strongest encounters of the entire game just because of the possibilities it gives you.

Forrest: The two earliest encounters in that game I thought are actually the strongest, because of the large and rich possibility space for the player. Plus there were lots of exploding barrels...everywhere!

Ed: Someone had a question earlier about sacrificing gameplay for aesthetics and I think the box factory is kind of a middle ground. You can bend it to your gameplay needs and you can make it really an excellent encounter place, and then because a box factory in a game is going to look a lot like a box factory in real life, it's also not going to diminish the visual realism of the environment potentially. But if you want to do the back of an airport, eventually you'll have to make those aesthetics sacrifices. So I think those boxy environments can be made to look more authentic because in reality they are bland.

Q8) What is the perfect relationship between designers and artists for you within the context of production; how do you guys split the relationship between these two roles and what does each profession do?

Forrest: For us it's really fluid and it just depends on the strengths and the weaknesses of each one; we pair people often. The line of where their responsibility lies is really variable, because we have really aesthetically-strong designers and very design-strong artists. There's not really a set line in the sand that nobody's allowed to cross or anything like that.

Jim: Same thing. Being a smaller team, the lines are very fuzzy for us. I can think of several guys, but one guy in particular we hired because of his LD work in Alice, which was in my opinion some of the best level design work I've seen ever; which is this mind-blowing, crazy, weird stuff being Alice.

And he's been with us close to 10 years now, but every year or two, he switches his job title; sometimes he's an artist, sometimes he's a level designer, sometimes he does both, and it just depends on what we need at the time: and then he's off doing modeling, or doing multiplayer shells, or working on the single player... I mean he's so versatile; he's skilled at both so we use him where we need him because that helps at keeping our team small.

Q9) What do you think of Stealth Survival Horror games? Do they usually follow the same design scheme as an FPS or do they differ considerably?

Matthias: Having actually worked on a survival horror game, at least one that got the name, I always say if Spielberg was to make a horror game, he would make Dead Space. We're not making Amnesia or something like that. I actually found that the way you are designing the spaces is very similar, what really becomes different is the pacing. If you're making an horror game where it's all about both the atmosphere and creating a curve on actually getting people's expectations and tension up. You're going to end up with much more of a spectrum for atmospheric spaces. If I'm looking at the beginning of one of our levels, we actually took out a lot of combat encounters because we don't want to barrage people with combat after combat after combat...

You actually want to keep building up a lot more atmosphere, which is an interesting problem to bring to production because "Why are we spending money on those areas, why don't we have gameplay in here? The game could be longer!" Once you actually buy into that fact, it's OK. To me that was actually the biggest change. I haven't worked on cover-based shooters, at least not at a professional level, so I think the way you start spacing out those levels -something Forrest already talked about- there's a difference there, right? If I am making a cover shooter or something like that, I am going to look at that space differently because I have to put cover in there, whereas in a game like Dead Space I want to make sure people can just come at me and actually have a way of controlling that space, not using geometry, but the stasis instead, and all the various tools the player has at his disposal.

Forrest: My take on Dead Space and all those contemporary survival horror is : they are very structurally similar to a shooter; it's just the sliders that are all being pushed to one extreme, generally the ammo-scarcity slider way up and obfuscation of space slider being pushed way up. But structurally: very similar.

Jim: Don't get me wrong, I love survival horrors. The Silent Hill series contains some of my favorite games I've ever played.

But it's a sub-genre of a sub-genre, and Gears is a good example of something that started out as a survival horror, and in order to reach a broader audience we had to change it. Because we went for something that we hoped would be more successful. That's not to say that kind of games can't be; obviously Dead Space is a great game and it's doing well. But there are other games that have done just as well, or have been just as good and haven't done that.... I'm trying to find the name of such a game on the top of my head... Zombie Killing?

Joel: Zombie Killing? Ok.

Neil: Resident Evil.

Jim: Resident Evil! Thank you!

Some of those early games were so popular! And then as they've gone later and farther into the series, I think they have become almost a parody of themselves in a lot of ways and people complain because "This one is too much of a shooter!" "This one doesn't have enough shooting!" and it's hard for them to find that balance. It's really difficult to do that because you try to reach so many people with a very, very targeted concept.

Joel: But I think a lot of it comes from pacing. I think Forrest sort of does himself a disservice saying his talk applies to shooters: I don't think that's true. I think all the principles you go over in that talk, they do apply. Like "analyzing" - Dead Space has all the setups! - But they're spaced out differently, right? With certain sliders being cranked way up. This is the same thing with the RPGs. At one point in writing my talk, I'd seen the outline of what Forrest had. It's like: well, I want to talk about open world stuff and the difference between open world and not open world. So I was going to take your talk and point for point break it apart and be like: "But that's not open world!"

Forrest: I had no idea you were going to be like that!

Joel: It was educational to me because I thought there was a talk there. We talk about conventional level design and how to break it down for, say, open world. You need to think differently, but at that fundamental level, all the pacing still applies. So I think with survival horror it's just taking the same basic principles of pacing. I think that's powerful, because it's a tonal shift in the game, this is something that's a statement about how level design is an expression of the general game design.

Q10) If you're working on something where you don't have any game assets or actually isn't even based on any pre-existing game, what would be the best way to approach it and make it designed in a way that it could kind of work with any first person game?

Jim: With Unreal or FarCry or anything, you're going to have a toolset there, that is obviously usable because they've developed the game with it, and a group of assets that they've used, and like Joel was saying, your best bet is try to combine those assets in a way that nobody else has done before and show that you can think your way around that problem, and not just end up creating another generic thing.

And if you can't, then you'll need to go the Forrest route which is make sure you have the basics there, that you understand the systems and set everything up and really delve deep into the process and explain to me why you made those decisions.

Matthias: A great way of thinking about this is just imagine you're designing the DLC for an existing game. Because this is exactly what you do when the game is done and you have to do Downloadable Content and you want people to buy it, you don't want to recreate everything, so you're looking at the game and everything you already have and you're wondering how you're going to "recontextualize" that. How with the least amount of effort do you get something completely new and exciting. And if you do that you actually put yourself up exactly against the problems that the guys working at that company have as well and you can compare and you guys can actually start comparing notes and how you would go about it.

Forrest: Whenever you're looking at a space of an encounter you've got to design or anything like that, I think what you need to do is look at the pieces of it and really ask yourself : "what am I going to do with these very specific pieces in this very specific place to do something unique and memorable and interesting and exciting and cool and"? All these things. But it's really hard just to build a generic level space in a vacuum that's actually meaningful and conveys anything.

Ed: I think Joel mentioned this earlier, like if you're going for a company and you send them a level that is super generic and does not have anything to do with the product they make you're not going to really call out of attention to yourself.

Joel: I mean it sounds like what you're trying to do is to build an environment in Maya. Like you made a pre-made, one-piece asset environment to show and tell "This is what I think about level design". And I think if you're starting there from a visual standpoint that can give you a good portfolio for an environment artist, but not a level designer. By and large, studio to studio culture might be different, but you need to think about gameplay first, that's why a lot of our art assets are modular because we have to make these really large games and we need to be able to make design decisions on the fly, without having a 3-weeks turnaround, on like, requesting a specific art for a set piece battle that we want to have, because it's going to be gameplay first and that actually informs our entire art workflow.

Q11) I was wondering if you think we've ever going to get away from chest-high walls?

Joel: You're talking about the space over-broadcasting what it is, right? There was a slide, Forrest was talking about it earlier, about coming into an encounter space and be like: "Okay well, clearly this is geared a certain way. Four guys are going to come from here and are going to take cover there so they can have an angle; and I'm up here and oh look! There's probably a sniper over there!" And you talked about having this problem in Gears, because it is so based on unit measurement so you can snap in, lean and everything else. It's an inherent problem but that's part of the art and level design relationship, it's being able to come up with environments that play well to the desired scenario of the game without them seeming overly contrived. And that's really hard: every game that does it is going to fail from time to time.

Forrest: Actually what you're saying there is basically an animation limitation more than anything else. It's how you need to build the world to the constraints of both the production of specific animations for specific world objects and the memory to fit all of the animations in as well. I'm guessing in some theoretical future where there could be a more reactive animation system and the AI could react to an object in the world much more like a human could, like respond well to an arbitrary height or traverse over something of an arbitrary height.

Neil: It's not really that far to be honest.

Forrest: No I don't think it is. I mean we're getting closer in many ways, but I think that's the constraint that you're seeing so to answer your question, I think yeah, we'll move away from that stuff probably in the short term.

Jim: I think we have! Gears is a good example. The first thing we do when a map is not working is to cut it in half. We take the top half and we cut it off. And then you're like "oh my god, I can see the entire battlefield now, this whole thing has changed!" But if you look at the map of the likes of Assassin's Creed, the whole game is about climbing way higher than chest-high walls.

You're climbing entire buildings so you can get the overview of the city. It's the game you're building, the camera you have, the animation of this... There are so many different factors that are built into it.

Joel: I'm quite surprised you didn't mention Brink!

Neil: I'm trying not to!

Forrest: Come on, sell it!

Joel: Let me do that for you!

Brink has all of this environmental data that goes onto how the character moves around the world so their designers presumably are more empowered to create things that fit the visual style they want and then the way the AIs and characters move through it is more fluid and natural because they have, with the SMART movement stuff, a lot of different situational animations.

Neil: Yeah, it's why I called it, really, a direction choice like we have consciously not been a cover-based shooter. So from the animation system level, everything is set up to blend together, we don't need fixed heights for anything, we do have height ranges, but that's more a design thing to keep things intuitive for players, that's not a technical restriction. Whereas most games when you see that, that's exactly what it is. Those are canned animations; it's those things that need to be at a certain height to get a line of sight over something. And that's just been the flavor of the month for the last few years; but you know, it's going to change because people are becoming tired of it.

Q12) My question goes to Ed, you talked about "unscaping the goat", and how scapegoating leads to all these problems. But what if the problem is somebody on your team and worse, what if it is you and you just realized it. What would you do?

Ed: One of the hardest things to realize is that you're a problem and the next hardest thing is to remove yourself. I think everyone gets at that point, something we don't talk a lot about, at the GDCs either, is burnout: there's a point of diminishing returns where you are no longer really functioning properly because you're either tired of the franchise, you want to get out of the industry, you're tired of your job, marriage issues, whatever... You have to recognize into the fact that, as a collaborative process, the game depends on a variety of functional entities. If you're in a position of authority, if you're a lead level designer potentially, and you're not working well, that means you're not disseminating the information to your team. Well the first thing to do is realize that you're a problem, work out where you need to go to be less of a problem or where you need to stop it. And if it is burnout, I think a lot of times we have to be honest and say "I have to go."

Jim: There are ways to go around it. Obviously depending on your studio you don't have to say "That's it. I quit." You can fix your behavior if you realize that it is a problem; if it's something that's truly making you unhappy and you don't want to fix it because you're just unhappy, then maybe it's time to leave, I don't know. But first you have to figure out on your own: Is it a problem with you or is it a problem with the company per say, which you can't be a part of. Again, I'm drawing parallels to my own career. The reason I started working on Gears is because I had spent 10 years on Unreal and I didn't want to make another Unreal game.

And so I moved over and switched team. I was lucky to have that opportunity, and we're a small enough team and we communicate well enough that we often ask people "What do you want to do next? Are you getting burnout? Do you want to try something new? Do you want to do something different?" Not everybody has that luxury unfortunately and you have to make that decision yourself, whether you want to stick it out or move on.

Ed: Some companies have skunk work projects or they have smaller teams within the company that are doing smaller experimental work or working on smaller titles and I think this is a really good idea because it can bring people out of here. I mean we're looking at like 5, 6-year cycles sometimes. 6 years of your life making a game that by the time it ships is completely different than maybe what you had at the start. The best companies are the ones that realize that and allow people to transition out that mode for a little bit and do something maybe a little more innovative, you know, do something a little bit more experimental and then after 6 months or a year of "refreshment", go back into the main loop. Every company should be concerned about senior people getting to that point of burnout or the point where they're not functioning properly. Hopefully there is plan and place for that. And the person you'll really need to talk to is your lead or your manager or the person in charge and explain "I'm feeling this, is this something you can help?"

Neil: I can't believe you said "quit first" instead of that one!

Ed: I was un-implying the "first thing you should do is resign"! But you can see the failure of corporations, companies, even empires sometimes by people failing to realize that they have an issue, and then once they realize they do have an issue, not doing anything about it.

Jim: Well you spoke about skunk work projects, and we do a lot of that. We have one guy in particular who, I think was feeling a little burnt out and tired, and so he started what we called the "LD Cabal". And he just started playing other people's games, looking at other people's tools, and then he would just call us all into a room afterhours and he would tell us what he found, like "hey they did this really cool thing that I have never thought of before! And they have this new tool or this new technique that we've never used!" And after he made his presentation, he says "Ok, so here's the challenge: you have 3 days. Come up with a new way to do X. Or show me something you can do with this toolset that you've never even touched before." And it's a really good way to invigorate the team and keep people fresh, even if they're still working on the same franchise or the same game and they're feeling a little burnt out. This is just kind of giving a creative outlet for your brain. We have figure drawing classes that our guys are going to, there are just random movie nights where we get together and we talk about stuff. There are so many different outlets and it's an important part of the job. I think it's really important, our industry being so unique, this is the team that you live with. The team you go into crunch with, you spend long hours with, you spend weekends with. And a lot of times you sacrifice time with your family and it becomes kind of a second family for you. So you have to have something other than work. I mean there's a work-life balance that has to happen for people to be happy.

Q13) Super Meat Boy: How are you feeling about the level design? Good LD or just unfair to the player?

Forrest: I think a lot of the joy of that game is how unfair it is. I think the most brilliant thing about that game is how each failure becomes just a part of that symphony of awesome that you're going to see the moment you actually finish that level. It's so much fun to watch all your meat boys get destroyed as you play it back.

Q14) How valuable is concept art? How much does a level evolve from the original concept?

Matthias: This is going to be very different for each company. I can tell you what I think is a very good choice as of Visceral and EA did it. As level designers we would do initially paper map then just do a whitebox, which is just a bunch of white boxes, I mean we're not even trying to do anything else.

After that, before we even do any sort of concept art, even if there might have been a couple of pieces that just establish the mood, like "We're doing a school" or something like that. We would then do a 3D concept with blocks, where somebody would very quickly go in and just sketch out: "This is where table goes, this and that". Then we would take screenshots of that and actually paint over that. Because that way you're not doing concept arts that goes to the nothingness but instead you're doing actually specific concepts for specific parts of the level. It's super helpful, you can even send that off to the 3D Artist and they have a very good benchmark on what they should be doing.

Q15) Does the short nature of shooters indicates the lack of ability to create new scenarios for people to play longer? Is everything we do getting increasingly complex that we are making them shorter and shorter just to be able to make them in a certain amount of time?

Coray: It's a tradeoff. Personally I would rather have a short game that is chock-full of awesome.

Q16) Checkpoints: At what point do you know you have too many? How do you guys do checkpoint design in general?

Ed: That tends to follow the usability studies. You can watch playtesters go through a level and begin to understand where the tough areas are. If you have a metrics tracking system you can look at the heatmaps and see where people are dying over and over again. There is a subtle art to checkpoint placement I think, in terms of where you can expect them to be needed, whether they're too close or not too close to each other, not having a checkpoint right before some ginormous battle. I think I always felt there is at least one designer in every generation that just place checkpoints and I am just trusting them with that.

Forrest: I think it's just technical. A lot of engines have the limit like you can't save when there's an active AI or something that cannot save its behavior tree or something like that.

Joel: Before I worked at Bethesda, because we do now have a "save anywhere" feature so it's not even part of our process.

But before when I did work on games that did have checkpoints, the only reason not to put checkpoint every five feet was because it was a pain. You had to make sure you accounted for different states people would be in, or that the level was in, and every time you had to do a checkpoint, it took you offline for an afternoon just to script that checkpoint. But I don't think there are too many.

Q17) At what point do you know your level is too hard?

Forrest: As a designer, if I feel something is just right and awesome, then it's too hard! I mean, you internalize your own game so well that what is fun for you as a creator is always hard mode. It's just a simple metric, if it's too hard for you as the creator of the thing, then it's way too hard, it's impossible, it's ridiculous.

Joel: This is the same tendency with puzzle games we talked earlier. I think we all just tend to make our stuff more and more difficult, because for us, I mean, I already played thousands and thousands of hours of Skyrim, and to me, Skyrim is already too easy. But for somebody who is actually playing the game, they don't have that, right? Somebody who's playing your game for 3 hours, 10 hours or even 150 hours, it's still a fraction of your familiarity with it. So yes, by and large, you want to thin down towards the other direction because it's easy enough to do usability testing and then layer in little bits of difficulty. Taking it out is harder.

Forrest: Also, in usability testing, a subtle clue to look for is if a player actually throws the controller... Which I have seen!

Jim: Rage quit is not a good sign...

Matthias: What we do is to design for hard mode. So if it's allright for you, then it will become hard for the others.

Ed: Way back in the day, at a GDC, I don't remember who said this but there is an old tip saying for level designers to play their level backwards, to give them an idea on how difficult it really is and a lot of times when you play it backwards you find yourself dying over and over again. I think this might be called a rookie mistake, I remember that when I was a Junior Designer, or when I see entry-level designers coming in, there is a notion that your job is to challenge the player somehow, like there is a competition between you the designer, and the player when really you need to make the player feel smart and not just punish him for every wrong move.

Forrest: Yes, it's very easy to win as the designer!

Q18) How do you feel about mutually exclusive branching content and at what point you think you should draw the line between useful design tools and potential complete waste of time.

Jim: It depends on your game...

Forrest: ...and the cost of your branch. If you're making a Call of Duty equivalent and you've got a branch in the first ten minutes of the game that will make two completely different games and you'll need to double your team size and that's not very viable.

Jim: When you're looking at a game like Heavy Rain, this is it.

The whole game was based around creating a completely different experience based on the choices you made in the first five minutes. I think ultimately it can be done, but it depends on your production budget and time and what your overlying goals are as a company.

Joel: It really depends on the type of game you're making. For us, we do non-essential content all over the place, but it's because those things do begin and end on their own accord. They don't always tie back in. We do big choices, and the problem with the way we do that right now is we inherently always have to come back and towards it. In the AI Summit last year, the gentleman who worked on Façade had a really smart talk about branching systems and being able to do this sort of things but without ending up doing sort of tree-connect, checkpoint-connect things that we're currently do. So there may be things like that, that are the way forward that are tech-based, but when it comes down to it, right now it's more a matter of allocation of resources, that are the level designers.

Coray: On a higher level, it's also really important that the player knows that he's making a decision. So that they just don't think that they're going down a fixed path. You have to assume your players will only play through your game once and not going to go back and play every branch of it. So just make sure your player knows A) They're making an instinct decision and B) Have some idea of what would happen if they made the decision the other way.

Neil: Nobody does that better than Bioware. Since Knights of the Old Republic, they've been doing those really well, really clear Good / Evil paths. If you make it through with one side, there's a good chance you're going to play through the other just so you can see what kind of experience happens. So yes, you have to sell that from the get go as part of the experience, and again, it's more a director level thing.

Q19) It's rather easy to balance the difficulty game design wise, but how about from a level design standpoint? What kind of feedback do you get and how do you tweak your "values"?

Joel: It's a dangerous thing to take feedback at face value. There is actually an art to interpreting feedback. When you have people come in and test your game, the less familiar with the game they are, the better feedback you're going to get. They are going to give you some types of feedback that don't really correlate with your game. And as a designer, you'll have to figure out what needs that person has, that your game is not meeting; and the actual feedback that person gave may not actually be a feasible route for you, or may not be the best route for you. But if you understand what it is that led them to wanting it, then maybe you can find something that does work in the context of your game that you can do to react to it.

Jim: When you're in preproduction, it's a good place to do that type of experimentation, but beyond that, if you're stringed too far from your core loop of what your game is actually trying to achieve, if you even give players the opportunity to veer off the course that far, it's very dangerous because your game just lost focus.

Related articles:

GDC 2010 Level Design Notes Part 1

GDC 2010 Level Design Notes Part 2

GDC 2010 Level Design Notes Part 3

GDC 2010 Level Design Notes Part 4

GDC 2010 Level Design Notes Part 5

How to Use Cover Object Placement for Level Design

Get the Book »