Interview: Chris Kay, Level Designer at Crytek
(Published on: August 10th, 2011)
Chris Kay, Level Designer at Crytek will tell us how levels were prepared in Crysis 2®. This article is the second part of interviews with people responsible for the look and feel of game environments.
Mateusz: Hi, tell us something about you.
Chris: I was named Chris Kay, 26 years ago one cold night in South England. These days I keep myself entertained at Crytek working as a Senior Level Designer.
M: Why level design?
C: Interestingly enough I didn’t have a master plan for becoming a professional level designer – it kind of fell into place. I had many possibilities for a career in my teens, a drum teacher, tennis coach, DJ and even a fireman! While all these activities are rewarding in their own unique way, I was missing the sense of creativity that I enjoyed so much in my other hobby… level design!
As early as I can remember I was building things. LEGO, bases in the local forests, booby traps with tripwires and eggs! So I suppose it was only natural for me to take this creativity into the digital world of games. Being a creative type of person it should come as no surprise that I enjoy and often study other forms of art, the history of past events and science of the future. If I didn’t enjoy making games so much I’d probably of ended up being a mad scientist!
M: How did you stared with level design?
C: Like many others I didn’t just open up a level editor and get cracking, but rather I spent years playing games usually sitting there wondering how on earth it was all done. Back in the days of the Nintendo Game Boy and Sim City it was all quite secret and rarely was there an opportunity to get inside an editing tool to discover things.
My first real experience in level design was with StarCraft, Blizzard made a point to make their editing tool extremely visible for everyone, both in the games menu and in the windows program bar, and for that I’m forever grateful. My best friend really enjoyed StarCraft too, in fact at the time I didn’t have my own PC so I went around his house after school to play it. We discovered the editor and built our own little missions including briefing screens with dialog text and some kind of map that probably didn’t fit at all. At the time it was most likely a showing off type of thing… Wow, how did you do THAT?! Nevertheless it was a pure discovery phase and opened up a world of possibilities. After getting the hang of StarCraft’s editor and playing around with Age of Empires tools my friend once again prompted me to check out another game, Unreal.
Working with an RTS editor is allot of fun and has vast design challenges, but I just couldn’t shake the feeling that it was limiting what I wanted to do. I wanted to MAKE something – not just place it. It’s safe to say that early on I was driven mostly by artistic and technical goals, rather than being focused on design. It took me a long time to appreciate the importance of design.
Traditionally you’re encouraged to focus on the design side early on, but my experience was heavily art focused and it paid off. Within a few years I could build almost anything I wanted. I was driven by visual ideas rather than paper sketches and game theory. I spent ages practicing architecture, geometry, scale, lighting and texturing. Of course the levels where extremely rough around the edges but at least I was motivated to finish them!
After a number of years spent doing things pretty much solo, I started getting involved with forum communities and things really took off. Noticing the huge interest in custom maps I quickly started to build my own and get feedback from others. It wasn’t long before I was churning out new maps every few months. Working on de_clan3_heat was extremely beneficial for me, up to that point it was more visually focused but the clan league map needed a very careful and balanced design. It was the first map I made that was tested over and over to get just right. From then on I’d consider the layout, scale and pacing to be one of (if not the) most important part of level design.
The next major change for me was getting involved with modding for Half-life 1 & 2, so rather than working on a finished game you’re actually involved in the development and discovering what the mod is all about. Although I worked on levels for many modifications such as Digital Paintball, The Specialists and Tour of Duty my most successful level was for Insurgency. I’d done enough to get my first professional job at a game studio near where I lived, so that’s the lengthy story as to how I got into level design!
M: What’s your responsibilities at Crytek?
C: At a basic level I’m responsible for proving gameplay ideas and the general look and feel of several levels of a game. As a design team you’ll learn things that work well (and don’t work well) so the key is to communicate it with others so you can ramp up improvements as fast as possible.
I specialized a little more throughout the development of Crysis 2® and at different stages of development my focus would shift from one thing to another:
Early on we knew that verticality would play a much bigger role in Crysis 2, with that decision made it was clear that we’d need to use metric rules to allow the player and AI to navigate the level spaces with ease. By the end of pre-production we had a set of defined metric scales for everything from doors, cover, elevations, ledge grab and horizontal jumps. To get to these values is no easy task, art, animation and AI teams need to be happy with the results and stick to them almost religiously. The metric rules where so successful that both AI and the player shared the same rules, meaning anywhere the player can go the AI should be able to also.
After the whitebox is approved you start the long process of integrating environment art, full level scripting including AI, story dialog, cinematics, objectives, save checkpoints and a ton of other small things. Production can be a hard slog for a level designer and you’ll struggle to keep up with all the different departments making changes.
Nevertheless there’s always something to do at this time, depending on how things are progressing it’s a good time to start some play testing and iterate to fix any particular issues. You might find people are getting lost and not sure what to do, save checkpoints are too far apart or there’s absolutely no chance of it performing to target frame rate and it’s down to you and your team to solve these before it’s too late.
No matter how efficient a sign off process is changes are inevitable, perhaps a major story element changes or a part of the game is cut for production reasons. Sometimes you have to simply bite the bullet and do your best to patch up the mess. Experience and good early planning can definitely reduce the amount of changes but it won’t stop them from happening completely!
Alpha – Shipping:
When the level is considered alpha sign off you’ll start general balancing and consistency passes, at this point all the games mechanics are working so it’s time to really use observation and feedback. You’ll be adjusting weapon load outs, ammo locations, AI scripting and player leading.
Consistency pass will include localization, making sure all printed messages are referenced for translation, all door switches, buttons and doors act as expected. Art will be adding the final polishing touches to lighting, effects and material tweaks and essentially doing a consistency pass of their own.
On top of all the balancing you’ll be fixing bugs like crazy right up to shipping, since you are the owner of the mission you’ll be helping to assign bugs to the right people if you can’t fix it yourself. QA will literally rip your logic apart and try everything they can to break your mission; often the things they find will astonish you.
Meanwhile, it’s important to keep an eye on performance and although you’re not expected to be an expert in this field it certainly helps. At the very least you’ll be asked to keep the level within predefined budgets and investigate should something drastically change.
“Second Chance” Crysis 2®
M: How you’ve been involved into pre-production of the game?
C: Pre-production is one of my favorite periods when developing a game, it gives you a chance to try out tons of cool ideas as quickly as you can. You don’t take any single idea and work off it for months but instead keep trying new things and see what other people think … is it fun? Does it fit at all with the games design? Could it really be made in the time and scope of the project? These are some of many questions you should be asking yourself and if the answers are positive you’re heading in the right direction.
At the start of a project you get some kind of short document with some basic information, concept art and plan of the games scope. For me it’s been a little easier in this regard because working on a sequel gives you a firm foundation to build upon. Nevertheless is should come as no surprise that the game your making will have similarities to other titles and you should have no problem picturing the basic design principles in your head.
All you really need is a plan of the gameplay basics, theme of the environments (to some extent) and a discovery phase. Now is a good time for lots of brainstorming sessions where literally hundreds of locations, gameplay and cinematic ideas are written down. It’s important to narrow down these ideas by voting as a team until you end up with some really strong ones. This is a great place to gather more detailed reference, story beats and doing any other preparations you might be asked to do before getting stuck in and actually building it!
The next stage in preproduction is to build extremely rough prototypes of your level, a block layout that can be adjusted easily. The key here is to get as much feedback about the pacing, scale, potentially bad performance hotspots and artistic input – artists should be involved at the earliest stage possible, not directly working on the block layout but giving their feedback regarding the theme and scale.
If you keep the block mesh simple and iterate using the feedback in no time at all you should have the foundations of a level complete. It’s important to resist the urge to add small details at this early stage – it’s not always that simple and sometimes situations will arise that require you to go into more detail.
“Times Square” Crysis 2®
M: How looks Level Designer’s typical workflow?
C: For a level designer the most important stage has to be the part when you can easily make BIG changes to vastly improve your overall design. The closer a level gets to completion the more costly both in time and money it becomes to make a change. Up to the point where production cannot allow you to do so anymore. It’s true that you’ll never consider your level “finished” and “perfected” but you’ll slap yourself for ignoring a nagging feeling that something was wrong and you ignored it – when all you had to do was move around a few blocks several months ago!
There’s a ton of stages a level goes through to get from paper to the shelf, all of them have their own skillset and specialties. For a modern single player shooter a typical workflow could be something like this:
- Figure out the theme of the mission, rough story/plot and estimated play time goals
- Build a very rough 3D scale prototype of the level, evaluate pacing, scale, potential performance issues and maybe drop in some basic AI
Early Main Production
- Take the level gradually to the next step; start to refine the combat spaces and scripting
- Meanwhile artists are replacing your rough geometry with something close to the final art, often not textured at this point
Middle Main Production
- Play test sessions are bringing back very useful results, work to resolve any issues with the levels pacing, leading and “fun” factor
- The level now should have no designer art, perfect time to start the long optimization task
- Level goes into a polishing phase, any kinks are ironed out and bugs need to get fixed
- Final art pass is completed and performance is raised to target frame rate
- All departments frantically fix bugs
“Dark Heart” Crysis 2®
M: Crysis 2® hero can use Special Forces to get advantage of enemies. How level was designed to work with all that abilities?
C: Crysis® has a vast set of tools at the player’s disposal; we extended those tools for Crysis 2® and added verticality into the level design. The basic principle of a Crysis® level is all about observing from a distance, making a plan of attack and then executing it. The verticality comes naturally in the sense that a good vantage point is usually elevated – so the challenge then is how you give the player multiple paths and options to achieve their goal.
The levels are carefully designed to allow for a variety of different play styles, elevated routes and jumps are provided for the maneuvering player, underground flanks are added for the stealthy player and primary combat zones are littered with cover for the run and gun type. The challenge really becomes working out how to best combine all these mechanics into a single area and have them be readable by the player.
M: So it’s important to work with Level Arist and 3D Artists as close as possible?
C: As a level designer you’ll be working closely to allot of different people, but it’s safe to say that you’ll be spending allot of time working with artists.
The key to successfully working with artists is to give them creative freedom, this doesn’t mean you have no input for the artistic look of the level (actually it’s quite the opposite) but it means you need to carefully balance time spent on artistic tasks. When you have people dedicated to making art and placing it in the levels, the last thing you should do is get in their way all the time.
The hardest thing to deal with is miss communications between art and design, it’s vital for the level designer to explain the game play and environment space in detail so that the artists are on the same page. The artist should also feel encouraged to provide work in progress assets to make sure everything is okay before finalizing the model.
Performance concerns can be tricky too, environment art will take up a massive chunk of the GPU budget for the game and one thing to remember is that artists are never ever fully satisfied with the detail and look of an area. It’s a part of your job as a level designer to remind them of the budgets from time to time but also do anything you can to give them even more to play with. Clever layout and streaming setups will make an artist your best friend ;)
M: You think that everyone can design levels?
C: Level designers get very excited about the long term goals that they want to achieve and seem to enjoy the really long process to get there. Not everyone has the patience or commitment to work on a single level for a very long time. Quite often level designers are compared to that of a movie director in the sense they bring all the pieces together and work on fewer but bigger long term goals – maintaining and executing their vision throughout the entire production.
If you get a kick out of long term planning, aiming to improve a design over a long period of time and being the “go to guy” for chunks of the game then you’re probably cut out to be a level designer.
Personally I’m quite envious of artists, every few weeks or so they get to work on something different, so for people who prefer shorter term goals there are other options – just as rewarding in the end but a much faster turnaround.
M: So what beginner level designer should know to get his first job?
C: This will sound very obvious but it really is the bottom line … do something that makes you worth hiring. That means putting something together that is close to that of retail quality and explains your thought processes. Remember that whatever you add to your portfolio will attract developers of similar games. If you make RTS maps I doubt Crytek would come knocking on your door. I’ve always shown my FPS levels and that’s what I got hired for.
There’s no doubt in my mind that having some great looking levels in your portfolio will be very beneficial. If all you have is a paper design and ugly looking placeholder art to show just remember that there are people who have all that and the finished art. Even though you’re not being hired as an artist remember that first impressions mean allot and even the most diehard designers like nice scene composition!
Times have moved on and only 10 years ago it was very possible for level designers to make their own environment art. These days it’s simply not an efficient way of doing things so I would avoid it if your primary goal is employment. So… use artwork that’s available to you already.
It’s important to remember that your first job will most likely NOT be what you expected – it usually takes many years in the profession to start working on your dream games. In other words your career should get better as you gain reputation.
Join mods, make some levels for your favorite games and get tons of crits and feedback from mapping communities, the connections you’ll make there will be invaluable in the future. Finally DON’T try and rush into the industry, you should be able to make an informed decision when you’re getting enough positive feedback or perhaps even be approached by a developer.
M: What game titles you enjoyed last time and what you can recommend?
C: Uncharted 2 – Not long ago I played Uncharted 2 for the first time, this was a real game changer for me as I’d never enjoyed console platformer games in the past. Uncharted 2 is about as perfect a game can get for me, every aspect is masterfully executed. I’ve never played a game with such a variety of environments and gameplay ideas, every mission feels unique and is balanced to perfection.
It left me with a better understanding of platforming, masterful pacing and cinematic storytelling. Make sure to check out the Art Direction of Uncharted 2 from a previous GDC it’s nothing short of awesome.
Mass Effect 1 & 2 – In my opinion the greatest single player experiences ever created, I’ve never been so engrossed with a storyline and characters it compares to the feeling I got watching Star Wars trilogy for the first time – epic is an understatement.
After playing Mass Effect some 7 times, I’ve taken away the belief that you can be as emotionally connected to a games character as to that of a movies. That a complex story spanning the universe can be told in a perfect balance of non-linearity and structure all while providing tons of RPG elements that keep the experience driving forward. Creating a game of this scope is a long term ambition of mine.
Everquest 2 – They say your first MMO will be your favorite and in my case it’s absolutely true. Diving into a world so vast and seemingly free to go and do anything was an eye opening experience to say the least. The mature writing style, dark and twisted fantasy setting and soundtrack are all inspirations to me.
I am awe struck at all the locations and themes in the game, lava volcanos, floating rock mountains, areas made entirely out of Ice, pirates bays, Asian villages, haunted castles, swamps, forests, ancient ruins, catacombs, jungles with aztec style temples, giant tree house cities, graveyards, sprawling desert lands, mechanical steam powered cities… just to name a few. In terms of themes and location design Everquest 2 sets the bar to the highest point. Some 753 zones are logged on the Wiki… now that’s allot of levels and an insane amount of creativity.
Although the MMO genre has typically shifted towards a more casual and streamlined approach, I’ve yet to experience anything close to the 5 year long adventure that I shared with so many others while playing Everquest 2 and it’s sad to say that I don’t ever see it happening again due to current trends in the industry.
M: OK, thanks for the talk Chris! Where readers can find you?
C: You can check out my website at www.2d-chris.com