Linked up a couple of my old talks on the blog. I personally don't really like watching talks on the computer - too slow! - I'd almost rather read a transcript. But if you're not like me, there they are.
But skimming over my talk on Prototyping and New Product Science - I'm honestly not sure how much of that I still believe.
There's a couple of strategies you can use when developing new games:
"Throw stuff at the wall and see what sticks" - the idea here is that you put a lot of minimum viable products out there, looking for something that'll catch fire. We could call this the Rovvio strategy - the story is, after twenty-something tries they made Angry Birds.
"Spend a crapload of resources" - the idea here is that nobody's going to care about your crappy little minimum viable products. You need to build a great game, spending years on it, and build an audience at the same time. (So in some ways it's the same, because maybe you put out a crappy minimally viable alpha version to start building that audience.) The Minecraft strategy.
So my talk more-or-less says to combine the two. Make a bunch of prototypes to see what's promising, and then spend a crapload of resources to turn that into your shipping game. Make some experiments, find your Tower of Goo, and turn it into World of Goo. One of the places where my talk falls down is in its methods: determining 'what's promising' is basically a gut-check by senior management. Would they recognize a Minecraft or Sims if the minimally viable prototype was in front of them?
Putting a free version out on the internet is one way to see where the traction is, but what if none of your products have much traction? How do you know how much traction is enough? Most of us aren't going to have a Tower of Goo that goes viral - we're going to have a handful of stuff and some of it will be more popular than others but none of it will set the internet on fire.
While doing that talk we were working on a prototype with Microsoft. We actually did a couple different prototypes, and my favorite got cancelled after weak testing.
Now, less work went into my favorite than the crowd favorite. It wasn't ready to be shown. But we had limited time and opportunity to get it in front of people. So we ignored my gut feeling that my favorite was the good one and went ahead with the other.
(And what happened to that project? It got greenlit - but then the project was handed to a different studio. Nice. And the other studio messed it up and it got cancelled, deeper into the funnel.)
The problem, I now see, is that in trying to reduce game development to rules-of-thumb and New Product Science was I was trying to bypass my brain. I've been making games for a long time and I should just trust myself to make the right ones.
I'm not getting rid of the idea of the funnel - I've got all these prototypes (and one minimally viable product) in the funnel: We're No Angels and Numbergun! and sixty second shooter and Energy Hook and Total Party Kill and a couple more that I haven't put Out There.
Where I'm making sure to use my brain is this: We're No Angels and Numbergun! are the ones that have gotten the most press. The old me would say those are the ones that are closest to catching fire. And yet I'm continuing to work on and promote sixty second shooter and Energy Hook. Because those are the strongest games. There's clever design in them. The others just have catchier visuals.
Is there a take home lesson here? Maybe it's, "Don't Let Your Strategy Override Your Common Sense." Hmm, sounds like common sense, doesn't it? But I bet people let their strategy override their sense all the time. I suspect that these funnels and greenlight systems and new product science that big game companies indulge in nerf the minds of their best people.