Game Architecture

Miegakure is a four-dimensional puzzle platformer. And when I say four dimensional, I don’t mean to use the 4th to refer to time or something equally lame. No, another actual spacial dimension. As far as I can tell the in-development game is played in three dimensions and at the push of a button one of the dimensions can be exchanged with another one. An intriguing concept and worth to watch for the mindbending spaces it is bound to create.

I won’t pretend to know how this is supposed to work but it certainly is interesting. Here’s a video that might help

Sorry for the delay but while I was able to write up the review for the first day of GDC at 4 o’ clock in the morning due to the jetlag, the rest of the days and nights I was way too busy to find the time to do so. However I still want to post the rest of the days but I’ll do so incrementally. That means I’ll start the post for one day, writing up one session, and then adding to that over time until everything is written.

That said, let’s start.

Level Design in a Day: Best Practices from the Best in the Business
Tuesday I decided to check out this full day tutorial. I was expecting a hands-on experience but because the attendance was so massive it was more of a lecture. Actually a couple lectures, each one on a different topic and held by a different Level Designer.

Level Design in a Day: Pre-Production
In this segment Ed Byrne of Zipper Interactive detailed the level design pre-production workflow as it exists at Zipper. It basically boils down to:

• Level Brainstorming
  A session where the team openly brainstorms for interesting locations and environments
• Abstracts
• Encounter Ideas
  A brainstorming for ideas for high points in maps
• Cell Diagram
  Arranging encounters into a sequence and context
• Encounter Models
  Prototyping of encounters, not neccessarily within the engine
• Walkthrough
  Writing a detailed walkthrough of the player experience
• Paper Design
  Creating a detailed design of the individual levels

There was some discussion on this process among the panel, especially on the last part of Paper Design. A few designers mentioned that they stopped doing 2d paper designs and instead work directly within the engine, whiteboxing the level. The argument against 2d layout plans was that the dimension of height is often underused if a level is planned on a flat piece of paper.

Someone from the audience also made a good suggestion for an alternate Paper Design tool: Google Sketchup and/or Layout. While I haven’t really worked much with either program, I’ve heard good things and might give them a spin sometime.

Level Design in a Day: Core Space Creation
The second lecture was from Matthias Worch of Visceral Games. He talked about “Digital Ditch Digging”, the meat and potatoes of Level Design. Using Bioshock as an example, he talked about how the experience of the player is shaped by the “physical properties” and “ecology” of a level.

With physical properties Matthias means the walls and boundaries of the environment while ecology refers to the placement of items, pickups and other resources. These two elements combined can create “weighted spaces”, making certain locations more desireable or frequented, creating hotspots and choke points.

Something I really liked was his example of the simple UDK level done with only 10 brushes, 2 wepaons and 1 powerup. This seems like a great exercise for students to learn how to do the most with little detail. All you can do using these tools I whitebox your level.

The slides for this part of the presentation are available online in ZIP format.

Level Design in a Day: Rapid Prototyping using BSP
This section was from Jim Brown of Epic Games. In it he talked a lot about the Unreal Editor as a tool to quickly create levels. Most of it felt like an ad for the tool but there were some nice points, like the fact that the Gears of War team used Kismet (UED scripting language) to even prototype some enemies or items before they were then created in full.

The Anatomy of a Social Gamer: Why Do They Come, Play and Pay?
So after a while I decided to head back to the Social Games Summit to check out this lecture. It sounded interesting on paper but unfortunately wasn’t. The setup was that Marianne Borenstein from playdom moderated the panel made up of average social-gaming joes and janes.

The session then consisted of Marianne asking questions to her guests about their personal history and involvement with games of all kind. The idea was to use this to represent the Social Gaming audience and shed some light on their motives and expectations.

Unfortunately the panel was quite boring and I couldn’t really agree with the basic idea. 4 people are not able to accurately represent the breath of the social gaming audience, regardless of the method used to pick them.

Level Design in a Day: Narrative Support throgh Level Design
This was the last part of the tutorial that I have notes for. Held by Joel Burgess of Bethesda Softworks, it dealt with the way a level designer can actually tell a story. This was one of the high points of the tutorial with a lot of good thoughts and info.

He started off by talking about the tool of a storyteller, which he separated into two categories: Language and Visual information. The former is direct and unambiguous but brings with it the fact that it’s often tiresome and can be a lot of work to localize. Visual storytelling suffers less from these drawbacks but at a cost.

The idea is to use visual means to create patterns that the player can use to draw conclusions. The idea is that the player looks at the world and makes up his own stories based on what he sees. The stories created this way are very powerful, because the player has created them, he’s become a designer himself.

Joel had a few examples taken from Fallout 3 that emphasized his points. As said, I did enjoy the presentation and you can take a look at Slides in PPTX format yourself.

I’ve been meaning to do this for a long time but kept putting it off. Now it’s finally here: the No More Wrong Turns article in handy PDF format for you to download and print:

• Download No More Wrong Turns PDF

The article is released for personal use only. If you want to do something else with it, please get in touch with me. I’ll propably say yes but it’d be nice to know about it.

Another excellent article I just want to quickly point out to you guys: It’s a short piece by the MTV Multiplayer blog about the design process behind the Shadow Complex world.

It basically boils down to the Chair Entertainment guys prototyping the entire game world on paper and playing through it. Accoding to the article this method helped identify and prevent at least a few issues that might otherwise have ended up in the final game. That just goes to show that there’s a lot of trouble you can avoid if you building a good prototype.

There’s a new video (one of hundreds) for Dragon Age: Origins out. Compared to the Marilyn Manson gore & tits fests of the other trailers this one is actually worth a look. Granted it’s still primarily a piece of promotional video but they touch on a few interesting points. A little bit about the visual design and the workflow.

They even bring up a few ideas I have lined up for yet another article about game spaces. And no, it’s not the one about boss fights…

Anyway, enough bla bla, here’s the video:

Alright my dear readers, the Gamsutra embargo for No More Wrong Turns is up so it’s time to post the article on this blog. In here I’m going to talk about navigation through the spaces of video games and the different tools that help us with that. This is mostly going to be relevant for 3d environments but much can be applied to 2d graphics as well. Also please bear with me, this is going to get quite long.

Read more…

In this post I’ll talk a bit about the goals and requirements of good level design. This thing has grown out of my architecture degree and the numerous lectures I held on the subject of level design – especially those with architects-in-training as their audience. So if you’ve sat in on one of them, this will propably be familiar.

So before I get into level design, I’ll first have to talk a bit about architecture. Thousands of years ago, the roman architect Vitruvius identified what he called the three “qualities of architecture”. They are as follows:

• Firmitas, stability. The building stands stable on it’s own
• Utilitas, usability. The spaces created by the building are suited for their intended use
• Venustas, beauty. This building has a beautiful aesthetic

They do make sense, right? Admittedly categorizing things is often arbitrary and can be argued for or against but I kinda liked that setup, where each layer is building on one another. If your building isn’t standing safely, it doesn’t matter much if your kitches is perfectly laid out. And if your doors are too small for people to get in properly, then it doesn’t help that the aesthetics are wonderful – the building itself is still a failure.

Now I tried transporting this divide over to level design, and this is what I got:

• Firmitas, stability. The level runs well without any technical or performance issues
• Utilitas, usability. The space does a good job at leveraging the game mechanics
• Venustas, beauty. The environment creates an atmosphere and provides affordances.

I’ll elaborate a bit on these three qualities and how I think they relate to level design as a discipline:

Firmitas
As mentioned before, this is a pretty basic category. A sort of minimum requirement if you will. If a level fails to fulfill it, it ceases to be properly playable due to technical issues. The two main areas here are stability and performance.

Stability as I see it refers to the level not crashing and breaking the game. Nowadays it’s usually quite hard to do that with a level itself since most engines are quite robust. It could be doable though with the level referencing a file that doesn’t exist or something similar. Also if there’s some sort of level script then that might be another source of errors. Of course if such a crash is very very rare it might not affect playability too much because it’s really hard to cause it. Still it’s a flaw in the “foundation” of the level. This can be avoided by working cleanly and paying a lot of attention. It might still creep in but that’s what the QA is for…

Performance is the other side and means that the level is running properly. Again this has become a bit less of an issue with higher end systems but it basically just means that your level doesn’t suffer from stuttering and frame-rate drops. In small amounts this can still be bearable but if your multiplayer level slows down to a crawl because there’s 6 players and 2 explosions, then you’ve done something wrong. This is often at odds with the graphics of a environment – the more detail you have the more your performance will drop. In the end it comes down to a balance between the two. And of course there’s a lot of clever tricks and thught out optimization involved. Everything from detail brushes to visportals and distance fog/culling. Whatever will do the trick.

Utilitas
This aspect of level design deals with the design of the actual play space. That is to say it defines the space within the player moves. I’ve talked (or at least tried to) about my definition of the digital play space some in my last (somewhat controversial) rant about how level design is game design. I’m afraid I failed somewhat since there were a lot of misunderstandings there – but I’ll try again sometime, maybe visually. Alright, slightly off course here, Let’s get back on track.

Utilitas defines the boundaries within which the player(s) move(s). It directs the flow of play and it’s job is to iteract closely with the game design to provide the intended experience (which doesn’t neccessarily have to be “fun”). It’s worth to note that these spatial constraints are what make the game possible and if they are changed, the possible experience is too.

The details of what Utilitas actually entails are too many to mention. This surely would be a great project quite similar to the 400 Project from Noah Falstein and co. but goes beyond the scope of this article, especially as these are highly dependent on the genre and game design involved.

These goals can range from anywhere between controlling the flow of players within the space to providing obstacles and challenges for the player to overcome.

Venustas
The words original meaning and the quality as it relates to architecture refers to beauty. That is to say that buildings are supposed to be pleasant to look at. This makes a lot of sense if you consider the fact that buildings are quite permanent construct and having a disgusting building in the middle of town could be considered visual terrorism for some sensible eyes. Of course back then Vitruv didn’t know or write about theme parks and other types of experiential architecture. There being “ugly” can sometimes be goal.

And the same’s true for levels. They don’t need to be beautiful, instead their goal is to evoke a mood, an atmosphere. So following that tenet, everything that relates to the creation of mood and visual references is part of this category.

Admittedly there’s some overlap with Utilitas since the visuals affect usuability. For example when putting moody lighting in a scene the designer needs to take care that the important areas are still visible. Or when trying to use visuals to steer the player. That said, for the sake of easier categorization it’s safe to put everything that doesn’t affect the space (textures, light etc.) directly into this category.

Besides mood and atmosphere, there’s a few other jobs that the audiovisuals of a level can do. A quite important (and easily overlooked) one is that they can create associations. Familar visuals can cause certaun expectations in the player and they establish a frame of reference. If there’s a castle on a hill, odds are that there are going to be guards and that there’s something worth guarding there. Expecting some sort of ruling body there wouldn’t be too far fetched either. Of if there’s a door it might suggest that it can be opened. All of this are helpful cues to help the player figure out what he can do.

Additionally the audiovisual elements can help tell a story. Both in the larger sense of selling the setting (castles for traditional fantasy, spaceports for science fiction) and in smaller instances of telling small stories within the world. The entire topic of spatial storytelling is actually a fascinating one that I plan to write another article on in the near future.

So that’s it. That’s my rundown of the three Qualities. Now what do you guys think? Is that a sensible system to categorize and analyze by?

An interesting article over at the BLDGBLOG about procedural generation of cities. That’s a topic I’m really interested in and that article is a good quick read with some interesting links to further stories.

Especially noteworthy are the game Subversion from Introversion and the procedural city project from Shamus Young of DM of the Rings webcomic fame. Also Viktor Antonov, designer of City 17 is mentioned in there. If you remember, I’ve already stumbled over him in the EDGE podcast I linked to a few days back.

Check it out.

Also if anyone has some additional info on Marco Corbettas Structure, I’d love if you could share…

EDIT: Here’s a YouTube video on Structure.

Level Design is Game Design!
Have I got your attention? Alright, so I admit the two are not perfectly the same but I am convinced that Level Design is nothing but a specialized application of Game Design. And I can prove it! First let’s start by looking at a pretty okay definition of Level Design:

Level Design is the process of designing and implementing the digital spaces of a video game

Sounds simple enough, right? Of course if we were scientific about this we’d have to take a closer look at what digital spaces are. However that’s beyond the scope of this article and will be something I’ll talk about another time. For now let’s look at board games. Why? Because they’re quite similar to video games and because it makes it a lot easier to understand the topic.

Now with board games, you have the rules of the game, which encompass how pieces can move, how a turn plays out and what the victory conditions are. The other half of that puzzle is the design of the actual board, which could be seen as the level design of the game. It’s evident that when the board is changed, the entire game changes with it. Imagine a game of chess, where the board is not 8×8 fields but instead 8×12 or maybe something more outlandish like an L shape. This will drastically change how the game is played. The same is true when a level is changed.

So if you think about it, the game space is nothing but another set of rules. They’re just visualized as a space to make it easier for us. Instead of rolling a die and tracking the movement pieces on a board, we could simply use numbers to denote a token’s position. Say if you’re on field “3″ and roll a 5, you wouldn’t move your piece 5 spaces, you’d write down “8″ instead. And then if there’s “special fields” you’d have a table where you could look up the number and see if there are any special rules for it. It’s obvious that this is a lot more complicated than moving a piece on the board, which is why we use spaces for a lot of our board games.

So if game spaces are just rules why is there a split between game and level design? There is no such thing in board game design, right? Well the answer’s simple: In the early days of game development there was no seperation: there often was just one person making the entire game, designing the rules, the spaces, the graphics and the sounds. But Because the increasing amount of content and complexity required specialization we now have different people doing different parts of the same job.

So with all that said, let’s look at that definition again:

Level Design is the process of designing and implementing the (spatial) rules of a video game

And let’s look at Wikipedia’s definition of Game Design

Game Design is the process of designing the content and rules of a game

Sounds awfully similar, doesn’t it? There’s just a few differences:

• Level Design only deals with those game rules which are manifested as spaces
• Level Design is also responsible for the implementation of these rules in the game
• Level Design only really exists as a discipline in the realm of video games

I rest my case.

The article I’ve been mentioning and working on in the past couple of weeks is finally online. You can read No More Wrong Turns over at Gamasutra. It deals with ways that Game and Level Designers can steer the movement of players.

Give it a read and please let me know what you think. I’m eager to get some feedback. Feel free to post a comment over at Gamasutra or here on my blog.