GDC 2010

It’s been a year since I posted anything here, and I haven’t been updating gamesound.org with the vim and vigour I aspire to. Sorry folks. I rather naively thought that after LittleBigPlanet had shipped I’d have a nice quiet year and be able to do a bit more writing – that’s certainly what I demanded in my letter to Santa having just emerged from seven months of crunch. Turned out that supporting our community with lots of DLC made 2009 pretty hard going. That’s set to continue, so things are still going to be pretty quiet round these here parts.

But after another round of inspirational talks at GDC (which just wrapped up a couple of hours ago!) I’ve got a little bit of spare time in my hotel room and a lot of cool ideas buzzing around my head, some of which I’ll share here. You lucky ducks.

Clint Bajakian (THE CLINT BAJAKIAN THE) gave an excellent talk on adaptive music techniques this afternoon. Near the end he touched on where he’d like to see ‘horizontal’/‘vertical’ streaming systems go in the future, which was basically fading out/in a phat pile of stems at different times, demonstrating this with an example mocked up in Pro Tools. One of the big problems this solves is the “stuff getting cut off when transitioning” issue, because you get rid of those elements elegantly at an appropriate point before the transition. Awesome, and the examples he had worked beautifully, but it raised the question in my mind of how do you remove something before a transition when the thing which instigates that transition has not yet happened? Bumcakes…

Then I had the idea that in some circumstances you can quite easily predict a future game state (especially in rather linear, controlled gameplay experiences) and use that information to drive the music system. The example I gave in the Q&A;/discussion at the end was that in a God of War combat-esque situation you can track player health and enemy health – if the player is totally kicking ass, has good health and the baddie is nearly dead then you could quite confidently initiate the first stage of the transition process. The tricky part is when the probability is ambiguous, such as when the player is running low on health and the baddie is similarly just as likely to die – it could go either way, and this makes it impossible to know whether to transition to the EPIC FAIL music or back down to the low-intensity (or whatever) music. Yet more bumcakes served up right there…

I left my braindump hanging at that point because I hadn’t finished thinking about it, but there was some more discussion along those lines. Lennie Moore picked up on it and approached the problem from the angle of thinking about the player’s emotional state – this was an excellent point and it triggered more brain juice but we ran out of time so I didn’t get to share.

The day before, Sid Meier’s GDC keynote had promoted the idea of going out of your way to make the player feel good. The example he gave was the fudged probability model in Civilisation Revolution, which does funky things like reduce the chances of the player failing a battle twice in a row because this would upset them – when players feel like they’ve been cheated they are more likely to throw in the towel. This certainly explains why Civ is so bloody addictive! The key point is that it’s about what the player feels should be the outcome of a battle and has nothing to do with the actual probabilities of success versus failure.

Tying these two ideas together is the cool bit, and it’s not unreasonable to suggest it would work well because they both deal with issues of probability. So, when the player is nearly dead and they are close to defeating the baddie, but not guaranteed to do so, the game seamlessly (and totally under the radar of the player) cheats big style by making the player invincible so that the music can safely do an awesome transition. It’s win-win. The player is happy for being bad-ass and their gameplay experience has been scored to perfection.

For me, these idea jams are what GDC is all about – it’s also a really good example of why audio folks should try and go to some non-audio talks. Don’t be shy in sharing your thoughts, because this is what makes a conference successful in the first place. Be inspired and be inspiring. That feedback loop = Awesome to the power of Win.