Leaving the Game

Leaving the Game was a project we did last year with the American Film Institute Digital Content Lab and a whole heap of different people from around the game, film, TV & IT industries, including Kuma Reality Games, Cartoon Network, Disney, IBM, Microsoft, Method, Georgia Tech, Furnace Media, and a whole bunch of freelancers. It was a pilot for a machinima series – one which we knew would never get made, but we wanted to prove that technically, it could be done, and artistically, that machinima has the ability to create TV quality entertainment. (You may argue that “of course machinima is good enough”, but the professionals still need to be convinced about what would happen if they started to use it.)

Looking back on it, it’s one of the things I’m rather proud to have worked on, and having it nominated for a technical achievement award at this year’s Machinima Festival is very satisfying.

So what’s so damn clever about it?

First of all, you have to understand that LtG uses old-style machinima technology. Those who’ve been around the machinima scene for a while will remember that machinima wasn’t originally a way to create video files. You didn’t get avi or wmv files you could upload to YouTube or stick on an FTP site. You had to watch machinima by running the game it was created in, and then loading a demo file which played everything back in the game engine in real time. So, no post-production, no clever edits, no mixing different engines. LtG went back to these roots, and everything you see is played in the Half-Life 2 engine, on a X-Box 360, in real time.

Early concept art for Amber's assassin costume

This, of course, gave us a whole heap of problems. Switching between locations – of which there are several in the film – would have meant huge loading times if we did each one as a separate level. Instead, we had to build all our sets on a single level and just teleport to them. We couldn’t cut away when we wanted to compress time, so we had to build identical sets running in different time frames and cut between them.

On the other hand, it enabled us to introduce some other features that would not have been possible if we had just made video files. What you don’t realize from looking at the video on the Web site is that Leaving the Game isn’t just a straightforward video. What you see depends on who you are and where you are.

Leaving the Game is delivered to the X-Box via X-Box Live as a pile of assets (all the custom content, skins, models, levels, sounds, music, voices etc) and the script. That, in itself, makes for a completely new content distribution model. If we’d done episode 2, we’d only need to send assets we hadn’t already used earlier in the series, so we’d already have the lead characters, main locations, title sequence, theme music, etc.

But – and here’s the bit that’s unique – the assets you get aren’t the same for everyone. If you’re an adult, you get the unrestricted version with full-on gore, swearing (and, I think, sticky-out nipples on Amber’s catsuit – not sure if those stayed in the final cut, though). If you’re a child, you get the PG-rated version without the blood and moderated language. If you’re Spanish, you get the Spanish-language version, and because it’s real-time, the characters lip-synch correctly, and the timings of the scene adjust to fit the duration of the speech. (To be honest, although we did the tech demo of this feature, it didn’t make it to the final cut, simply because we ran out of time to do the second set of voice recordings.) And finally, the product placement is adjusted to fit you as well. In one scene, we had bottles on a shelf. Kids saw a soft drink, adults in some states saw one brand of beer, adults in another state saw a different brand of beer. We experimented briefly with making the lead character either black or white, male or female, at the viewer’s choice, but decided that although it was technically possible, we didn’t want to make the story or the character too radically different.

So what?

This is a completely radical approach to the way we produce, deliver and view content.

By using a connected games console and game engine as the core of the viewing medium, we change the way you get content. Instead of delivering movies, you deliver everything you need to have that movie reconstructed locally. This in turn means that you can produce content for a wider audience: the same content can appeal to both kids and teens, and you can create foreign language versions faster and cheaper.

And lastly, and possibly most importantly for the professionals, advertisers can see the potential in being able to target audiences more precisely. They can place a product into a piece of content, and know that it will only be seen by relevant people. In these days of dwindling ad revenue, and as content creators find it harder to get funding, this is highly attractive.

I’ve believed for a long time that machinima has the power to shake up the media world, and Leaving the Game proved it. It was also a hell of a lot of fun to make, and a great opportunity to work with all sorts of people I wouldn’t have met otherwise. Hats off in particular to Keith & Dante at Kuma, whose team did most of the hard work.

Amber as she appears in the film