When showing up at GDC, I leafed through the booklet to see if there was anything I’d missed. And, indeed, there was: it seems that the online schedule builder had left off most or all of the keynotes for the summits! (Or at least did when I looked at it last week, I guess that’s fixed now?) Oops. So I decided to go to two of those today instead of what I’d previously planned, for the Social and Online Games summit and for the Independent Games summit. (At 10am and 4pm respectively; I guess they don’t trust indie game makers to be awake at 10am?)

The other thing I discovered today: if you want to zoom when taking pictures on your iPhone, you have to buy a third-party application. It’s only digital zoom, of course, but it’s still convenient to be able to do that right on the phone so you can read pictures of the slides easily right there.

Today’s notes:

10:00 am: How Friends Change Everything, by Gareth Davis.

He’s Facebook’s platform manager; mostly Facebook boosterism and stuff that’s not too surprising to those in the social games business, but pleasant enough. As he notes, Facebook is interesting because it puts front and center the possibility of game interaction with people who aren’t essentially anonymous / pseudonymous (as is the place when playing strangers on XBLA); the idea of games being social in that sense goes back thousands of years, so in that light maybe Facebook games should be considered normal, while video games’ relatively brief history is the exception?

To that end, he thinks that the iconic Facebook game is still ahead of us, and that we should expect to have multiple games with over 100 million players; both of which I quite agree with (there’s certainly no reason to think that FarmVille isn’t going to be surpassed multiple times over. And, of course, he thinks that Facebook will be at the center of social gaming; hard to argue with him, for the time being.

11:15 am: The State of Social Gaming: Industry Overview and Update, by Justin Smith.

He’s from Inside Network, which publishes web sites like Inside Social Games and Inside Facebook. The talk was what the title suggests; so, for better or for worse, nothing here that I noticed that I wasn’t already basically familiar with.

11:45 am: Open Source Secrets: The Software Architecture Behind a Successful Virtual Goods Business, by Timothy Fitz.

He’s the technical lead at IMVU, and I found this talk hugely refreshing: an ode to continuous deployment and the testing and refactoring that makes this possible. They have a thick client that they push new versions of once or twice a day, and they update their website 35-50 times a day (though only pushing changes out to ten percent of users); so their cycle is “commit, go green, push, repeat”.

They learned this lesson the hard way; they once worked on a release for two and a half months. Fortunately, they did A/B testing when they did finally release it, which revealed a serious problem in chat functionality; it took them an additional 6 weeks to figure out what went wrong, because two or three minor changes interacted in unexpected ways to produce a bad effect. Oops. So they don’t want to have to go through that again.

Their heavy emphasis on automated testing doesn’t mean that they don’t like manual QA: instead, they want the QA people to use their brains, instead of acting as robots. So QA people think about how to test features (which engineers then automate), and they work to understand the impact of planned changes in interfaces; automated tests work to catch unplanned interface changes.

The title comes from the way they build their client: they realized that they have a lot of good UI experience on the web, so they build on that by embedding Firefox in their client, with custom C++ code only in one narrowly defined area. They have significant amounts of code in each of C++, Python, and JavaScript, and they find this works better for them than a single-language strategy.

1:45 pm: Why Are Gaming Veterans Flocking to Social Gaming?

This was a panel discussion between Steve Meretzky, Brenda Brathwaite, Brian Reynolds, and Noah Falstein. Quite pleasant to listen to; two themes that repeated themselves were: 1) they all missed times when games could be developed quickly by small teams, and see social games as recapturing some of that, and 2) if you’ve been in the industry for more than two decades, you’ve gone through a lot of changes, and this is another one of those; so having veterans around is valuable for that reason alone.

They didn’t have any crystal balls or anything, just a pleasant chat. One of the more amusing lines: Brenda Brathwaite saying that, now, with the emergence of metrics, game design has finally become a game for game designers!

3:00 pm: What Social Games Can Learn From Virtual Worlds, by Mike Goslin.

He’s the VP of product development at Hangout, and also has Disney theme park experience. His first claim is that retention is key: it increases lifetime value in obvious ways (lifetime increases) and slightly less obvious ways (people get drawn to status items and subscriptions), and decreases cost of acquisition (virality works better over time, better to invite fewer people over a longer time than to spam everybody you know at once). He thinks that concurrency leading to a sense of community is important; it certainly seems to be something that Hangout is betting on, and something that Facebook games so far have generally stayed away of. (Wild Ones being an exception here.) They actually use Unity instead of Flash, another exception, and they have games that are gateways to a chat / community area.

As to the title: social games have the advantage that player registration is very easy (because Facebook handles that), but getting players to monetize is harder than in virtual worlds. To that end, he made a push for subscriptions: not as requirements for players (the way World of Warcraft does it), but as an additional layer on top of virtual goods to increase core players’ commitment and to give them a way to get VIP accessories. (Hangout has a fashion show theme, so the latter probably works well there.)

4:15 pm: Increasing Our Reach: Designing To Grab and Retain Players, by Randy Smith.

I really enjoyed the talk he gave last year on “helping your players feel smart”, so I figured I’d see what he had to say this year. And it was my favorite talk of the day, though I certainly won’t be able to do it justice here: basically, making the point that games have to work at immediacy (make the first 10 minutes awesome to draw players in) as well as depth (to keep players satisfied over time). And he went through a bunch of games (which I want to play now) showing how they did well or badly on one or both of these.

For immediacy: provide a “game-toy” as a player’s initial experience. This should package up the game’s core ideas in a way that’s as fun to play as possible: have simple, intuitive controls; have strong, juicy affordances; and make it hard to fail, lowering the pressure. For depth: provide it on demand, so players can go into games as deeply as they wish but don’t have it forced on them; provide hints to players for depth that they might not otherwise notice (e.g. completion stats on a level, achievements suggesting things to try out); and have both low-level tactical gameplay loops and mid-level strategic loops.

He talks fast. I hope he’ll put his slides up somewhere?

5:15 pm: AI Developers Rant!

Insufficiently ranty; a bit of completely gratuitous guy culture, and a Bourdieu reference that might lead to something interesting. I napped through parts of it.

Post Revisions: