Penultimate Level Design

Slide from Scott Roger's talk on Level Design and Disneyland

Slide from Scott Roger's talk on Level Design and Disneyland

As I mentioned in my other GDC envy post, there’s been a bit more about Level Design this year – at least it seems that way.  Another of the exceptionally popular talks was by Mr (Boss’) Scott Rogers and how he learned a great deal from the tactics that Walt Disney and his Imagineers used in creating Disneyland.

You can find the complete slides on his blog here (and they’re a good read – curse my non-GDC’ness!), but what I find interesting is that this is not the first time that I’ve read of people taking level design inspiration from Disneyland.  A few years ago, a Mr Don Carson did a 2 (and eventually 3) part Gamasutra feature on level design and environmental storytelling (subjects very close to my heart).  You can find parts 12 and 3 linked.

Both of these people looked to Disneyland to teach them about using environment to draw the eye, tempt the senses, tell stories and reward the curious – for all that’s not so peachy about the Disney monster, these are lessons that many level designers either still haven’t learned, or know only on a vague instinctive level.  So many of us still see a corridor as a corridor instead of an opportunity to guide the players eye, tease them with anticipation, delight them with newly discovered secrets or even to give them a breather between visually intense nodes.

These are old lessons, taught by a man long dead who grasped these long before others even realised what he was doing – we should have these things down to a fine art by this time!


So, every year I wish I’d gone to GDC, and every year it turns out I’m not there for various reasons.  This year I’m particularly sad as it seems that level design is getting a bit more attention these days (and the indie stuff sounds wonderfully exciting).

Interestingly, Bioware decided to show off a little of their iterative level design system that they’re using for Mass Effect 2.  With their newfound focus on more-shooty-less-talky the level design for their game has become even more important.

Looking at the video above, you can see that there are 5 playable iteration stages that a level appears to go through:

  1. Narrative Playable – Looks like Unreal BSP, the old level designer standby – I’d know that texture anywhere.
  2. White Box – First pass imported geometry, I’d guess.  No textures, plain firstpass lighting, placeholder cover.
  3. Orange Box – Cover hints added to the geometry, some lighting, limited scripting.  Better idea of the combat.
  4. Hardening – Textures and scripting.  The AI is behaving far better.
  5. ?  – Not mentioned, but I’d assume this stage is ‘Polish and Ship It!’

Now, I’m of course working off 2nd or 3rd hand information here – only having seen the video that someone sneaked out, but I like the names they give these stages.  In other situations I’ve seen people going through these stages in various names and forms, but seldom are they defined so clearly.  (I’ve also heard people call White-Box, Vanilla, or Blue-Box.)

I have to say though – having used BSP in the past, it lends itself to a particular style of level design – axis orientated and flat floor’ed, and that brief, unfair video above doesn’t suggest that they’ve made a push to get away from those easy quirks.  If they’re focussing on the shooter side of their game, then I’m hoping they’re going to investigate ways to knowingly play with the geometry more, surprise the player and use some more quirky angles and vertical interest.

Potentially, Conker Mill - Environment Profile Fodder

Potentially, Conker Mill - Environment Profile Fodder

When you’re designing a level, unless you’re part of a one man or woman team, it’s certain that someone else is going to have some input into how your level is eventually shown to the player. There’s a whole chain, in fact, from initial concept to rough layouts (and vanilla geometry) to artistic concept art and tone-setting artwork, to final graphics creation, object creation, lighting and effects passes.

Making sure that the final level presented to the player is a unified whole is frequently left in the hands of the concept artists and art director – and perhaps that’s not where all the blame/responsibility should lay.

In fact, in the past I (as level designer for a particular environment) have taken it upon myself to create an Environment Profile – a short document documenting the intrinsic nature of the place. This will, if written well, help out the concept artists to get the feel for the place, help the level dressers make the place feel lived in (if it’s supposed to be) and even help the character artists come up with how the people there will look/act.

How To Write An Environment Profile

  1. Keep it short – Anything over 2 pages of well spaced A4 (Letter-sized or so) won’t get read by the frequently word-averse art types. Break it up into smaller segments if you have to (give different areas different documents) but a short novel of the history of the place won’t get read.
  2. Who built it? – The initial reason for the space to be created. Was it a home? An office? A factory? Who was going to live/work there? Why was it built the way it was?
  3. When did they build it? – If they built it in the 1890’s, it’ll have a different feel and age to somewhere built in the 1970’s. Things age, break, get improved upon, get expanded on.
  4. What is it used for now? – Is it abandoned and frequented by gangs of kids? Is it refurbished and used as trendy offices? The people there – what do they use it for?
  5. Who are these people? – The people who have used it – builders, owners, users, workers, homeowners – these people need a brief description to give a stamp of their presence on the world. Did the kids growing up there shoot arrows into the loft? Were the workers in the factory superstisious about the mooses head in the staff room? Did the people who built it use cheap materials because of the Great Depression?

So, short version: Who built it and when? Who uses it now and why?

Here’s an example one, for something fairly mundane: Conker Mill (Studios)
History – Built in the late 1700’s, Conker Mill was built and run by Mr and Mrs Miller to grind grain for the nearby ‘big house’. Using the most modern equipment available at the time, and of sturdy stone construction, it survived 3 of the Miller’s children growing up in it (2 childrens bedrooms, one for their daughter and the other for their two sons) and passed to the eldest son.
It was abandoned during WWII with no-one to care for it, and lay empty and rotting until the 1980’s when it was bought up and converted into a set of art studio units, Conker Mill Studios. It now houses up to four artists, their workshops and gallary spaces. One is an artisnal blacksmith, one does ‘edgy’ taxidermy, one creates musical instruments from found objects and the other paints rocks.

Nothing too cutting edge there, but much better than a brief of “Its a set of offices.” It’s something for a level designer to get their teeth into – something for concept artists to latch on to and something for level dressers to have a bit of fun with! That should result in a level that feels like a lived in place, not a random collection of corridors and rooms.

Post apocalyptic Lego scene by Legohaulic

Post apocalyptic Lego scene by Legohaulic

So, I was thinking…* Level design is not only a funny beast, but one that is done differently in every company, as far as I can see. It’s interesting – some companies don’t need any level designers, some have artists who fulfil the role, some call them Level Builders and have them create the final art while they’re doing it, while others are Level Scripters who don’t do any of the art, and instead finely tune the programming-language-like scripting.

I am most familiar with Level Designers who are not responsible for the final graphics, but who are responsible for the rough layout of the level/environment, and the eventual scripting of the gameplay within that (arted up) level. The early phases of this process vary dramatically from studio to studio – the roughing out of the layout. Some historically have used Lego, others use sandpits, some use graph paper, yet more go straight into a 3d package, and some use Visio. This, is what I’m interested in.

Things like Lego, sand and paper models – my worry is that they initially seem like an excellent idea, but might ultimately be more limiting than enabling. Imagine, for example that we’re making our level layout in Lego:

  • Easy to use – Unfortunate indeed is the level designer who’s unfamiliar with Lego and its easy to use bobbly blocks. Even without ‘custom pieces’ you can build vehicles, buildings and basic shapes very quickly.
  • Available – Need more? Go buy some or start asking around the office to see how many others have secret stashes in their lofts from their childhoods. But it in bulk, buy it online, buy it by the bucket, buy it piece by piece – there’s so much of it! You could probably buy lots off Ebay – only lightly chewed!
  • Thousands of pieces – Custom pieces range from sharks and daiseys, to lightsabers and space buggy wheels. Moon scape panels, road shapes, grassy panels – hundreds(?) of themes from the everlasting and reliable ‘City’ range, to the classic Space series, the sorely missed Ninja theme and the old favourite Knight theme.
  • Permanence – Lego is tough for its size. If you build it, it can stay up for weeks (barring small children attacks) and when you take it down it’s completely reusable.
  • Modular – You can easily duplicate Lego creations if you have the parts, and even if you only have one of each ‘lump’ of level, you can move them around without them falling to bits, usually.


  • Expensive – Lego is costly! If you’ve not bought a Lego kit for a while, then you might not realise, but even a small kit will set you back a tenner or so! If you’re going to be building large level spaces, you’re going to need a LOT of Lego.
  • Angular – This is where it gets important. Lego is built on 90 degree angles, and trying to deviate from that is likely to involve custom pieces and headaches. This may influence your designs and cause you to lean towards right-angled levels, even when it’s not appropriate!
  • Small – Lego has a very particular scale, if you intend to use their custom pieces, and that scale is rather small. Nothing is smaller than a single ‘stud’ in size, and on the other end of the scale if you want to make large things, then no Lego block (as opposed to panel) is much larger than half a CD box. I suppose fidelity is a better word – or granularity, for extra pretention points.

Now, of course you can get around some of these problems, but the 90 degree problem is the most important one I see here. Level design is a very squidgy thing, and the slightest influence in the earliest days can carry through to the final game – so even if you try very hard to resist he 90 degree problem, it’s likely that your end result will still bare the scars.

Other mediums have their pros and cons, of course, and I might go into those in subsequent posts, but Lego as a level design tool isn’t going to be my first choice. The 90 degree problem has mostly scared me off, as I am personally very averse to overly angular levels. I can see, however, that for other game styles or genres, Lego would be excellent – creating Sci-Fi or urban levels set in modernist architecture, for example, would be super-quick to mock up in Lego. I don’t think it’s my thing.

*This is a phrase I use a lot, usually followed by my latest bonkers thought.

So, I’m in the industry – have been for just over seven years now(!) – and I keep learning new things every day. It’s not enough! I want to improve at what I do, and while there’s definitely something to be said by learning by doing, and I do learn by making more levels, I think that there’s more I can be doing:

Current learning avenues:

  • Life Drawing – Always good to get a bit better at drawing, what with images being worth a thousand words and all. A simple sketch of a bit of level layout, or a stick figure in a storyboard, can save a hell of a lot of explaining and misunderstanding. I’ve not done much drawing since school (A ‘B’ grade in my Scottish Higher) so this is flexing long rusty muscles!
  • C# – Knowing a coding language helps a hell of a lot, despite the old adage “A little knowledge is a dangerous thing” and C# is the cool kid on the block. I’m a bit of a whiz with DarkBasic, but there are some things it just doesn’t like doing. Also, C# opens the doors for XNA fun too (and has Wiimote libraries).
  • Board Games – Playing a lot more of these now, and learning a lot of the different kinds of play mechanics you can have. What makes a non-computerised game slow or interesting, and it’s pretty amazing (really) that with some card, dice and some pencils, you can make a multiplayer game! Yeah, not quite the same, I realise, but paper prototyping has its place, I’m sure.

So, I’m not sure what else to do – I admit that my C# knowledge could do with a lot more work, and well, making levels for other genres is something on my list (to keep my skills flexible), but what else would be useful?

Some random ideas: make my own boardgames, write a comic, learn Flash, make an XNA game, write more (short fiction), run some RPGs, paper prototype some game ideas… What others?

Damnnit! I do love games! I absolutely adore them! I love the little slice of alternate world they present me with, I love the carefully crafted environment for learning and excelling in new skills they offer me, I love the puzzles – whether that’s a Professor Leyton style match puzzle, or the puzzle of how to get past that entrenched team of enemies without a grenade.

The reason for this opening outburst? Well, I’d read a piece of recruitment waffle that said they were looking for people like that – people who loved games, their worlds and their challenges. It spoke to me, and I shone with the glow that someone has when they think “Someone else knows!” What happened next was a well deserved, but ultimately saddening plummet back to Earth. There aren’t companies who feel like that – not any with more than 4 people in them, anyway. They may say they do, and they may even contain a certain amount of people who do – but a whole, large company? Nah – I was deluding myself, and the recruitment waffle was preying on that.

I don’t want it to be that way, though. I still love games and the worlds and challenges they present me. I revel in the chance to make or improve those worlds myself – to carefully craft an environment, a backstory, a mission or a puzzle, and by Jove I intend to keep doing just that, and improve at every opportunity. So what if there’s a lot of cynisim in the industry, if recruiters prey on the enthusiastic (mmm, crunch-fodder), if market forces and limitations mean games frequently fall drastically short of their potential – if we don’t keep trying, if we don’t keep our love of games burning with the intensity of a white dwarf star, then the industry will have one more casualty. Me. And I don’t want that to happen.

et cetera