10:00am–11:00am: City of Wonder: Postmortem”, by Troy Whitlock and Scott Jon Siegel

Actually, only Scott spoke, presumably because Troy is no longer at Playdom. (He left a week or two before I did.) At any rate, I was glad to see a postmortem of this game, it was one of my favorite social games from last year.

The game‘s original pitch: expand your city state, discover technologies, battle with armies, from the dawn of history to the present. Take the Social City engine, add in more choice and player-versus-player combat.

Scott thought the original pitch was a crazy idea, had little interest in working with it. Social City was genre-defining, which meant that it had spawned a lot of competition. And Studio 24 (the Playdom studio that he was part of) had previous experience with market saturation: that’s a big reason why Fish Friends failed.

Also, the pitch looked like a niche game; it wasn’t clear it would get the same nice demographic that Social City got. So why do it?

Reason 1: PvP + isometric city builders. Both of those factors do a good job of leading to item purchases. The former affects game mechanics (combat), but doesn’t affect customization. The latter has opposite problem: only aesthetic. Maybe you can merge the virtues: give a stat boost plus expression?

Reason 2: The challenge of making “hardcore” features appeal to casual players, let people do what interests them.

Social City had a good core loop: money, population, happiness: need to work on all three, leading to virtuous cycle. City of Wonder kept the same base mechanics, except it gated new buildings by progress along the technology tree instead of strictly by experience.

Also, the research / technology tree led to long arcs, longer-term goals. Leads to diversity of city appearance. But traditional tech trees look like spaghetti maze: definitely didn’t want that. So they made it much less intertwined: even that, though was potentially too overwhelming for casual players. So they left the full tech tree as a depth on demand feature, but presented a few choices through the main view, with advisors giving specific recommendations. (And, as a bonus, the advisors give the game more personality.)

Also wanted to bring across point of progressing through time, e.g. by giving a parade when you reach the Bronze Age. But in early playtests, players thought it was a stone age game, which not all players were as interested in. So they addressed the sense of progression in the loading screen, in splash text, and in having each age’s buildings having a distinctive look.

Title: Social Civ was its working title. They tested various names; City of Wonder consistently won. Scott didn’t like it – e.g. no use of “wonder” within the game. But they went with it; his current theory is that people like it because of a song lyric.

Embassies: they wanted a neighbor mechanic, and wanted to do that in a way that’s not just a one-time thing, but instead has a concrete ongoing effect on your own city.

PvP: provides a second trio of attributes for buildings, cultural / economic / military. It’s hard to do well on all three, so people could specialize in one or two. The combat outcome (and what played into it) is hard to explain; they iterated a lot on that screen. Emphasized differences between wins and losses, and reinforced distinction between different types of expeditions.

Legends were most fun part of development: enjoyed coming up with quotes (hat tip to Steve Meretzky).

Game as a whole was his favorite social game development experience. Havin said that:

What went right:

  • Casual focus, or democratization is a better term – the iPod is an example of what they’re going for, letting user find any song in at most three button presses. So they made main game features accessible with minimal clicks; other features are optional, scalable.
  • Iteration. Everywhere: e.g. the placement of buildings/roads in starting layout turned out to make a big difference. Supported via continuous integration, and always having a stable build available. Scrubbed the game every day: errors in text, glitches, feel of an aspect of gameplay, …  Started as 3-hour meeting, got shorter as game progressed.
  • Amazing team, split between Mountain View and India.

What went wrong:

  • Legends. So appealing that management encouraged them to double down with legends as a trading card concept. Sounded good, but trading card aspects were added in last minute, so not enough iteration: badly balanced, badly explained. (Most successful ones were promoted during pinch points.) Ultimately not sure if feature did more harm or good.
  • Embassies. Added somewhat late; main issue they didn’t deal with at launch was “Embassytown”, where your city is cluttered with embassies. Thought it would only affect a few users, but affected many, especially your most engaged users. Eventually solved via multi-embassy buildings.
  • PvP. Unexpected side effects of choices players could make, sometimes made players unhappy. No reward in attacking higher-level players: you know you’ll lose, or even if you won’t, there’s not a big reward. So don’t get the excitement of working towards big wins.

The game is still a live service. They recently released the first colony, and there’s room for more. Colonies let them grow the game, even adding new features (e.g. naval expeditions in the water colony) without hurting the core gameplay in the main colony.

11:15am–11:45am: “Humanities Unlocked: The Value of Liberal Arts for Your Game Design Program”, by Mia Consalvo

First, her slides:

She started with a word on advising: “advising itself can be problematic”. At her first job, she received signals that it wasn’t important. At her second job, it was easy to fall into rhetoric like “you only need to take four of these five areas”. Now at MIT: students like math / technology, hard to get them to take humanities.

What is the worth of the humanities? College costs a lot; what will you do with an English major? Humanities programs are getting cut. She’ll present four case studies showing how humanities courses are valuable for game design.

1) Philosophy

Mentioned Miguel Sicart, and a Project Horseshoe white paper on ethical issues.

Ethical theories: virtue ethics (good habits of character, but with balance); utilitarianism (look at consequences of action: are they more favorable or unfavorable (to everyone); duty theory (have specific foundational principles of obligation: e.g. stealing is wrong).

When playing games with good/evil split, players generally take he good side in the first playthrough: it’s easier, more natural, how they were trained. (Even when trying to be bad, players accidentally did good when playing on autopilot.) Why, exactly? Is the underlying theory duty-based? Is it the rules: the good guy always wins? More content: evil in games often means selfish, so you don’t get to do side quests?

Discussion of Zevran in Dragon Age: Origins: she thought she’d been being a good manager, but Zevran turned, she killed him. Then in Awakenings, she had a similar issue; this time, though, it was a character that she’d gone out of her way to cultivate. Ultimately, though, they disagreed about something important, no amount of convincing worked. Which annoyed her at first, but she rather respected the designer’s choice once she’d thought about it more.

2) Foreign languages

Western otaku: lead to interest in culture, languages. Many people took organized courses; some learned on their own, though. And then learned about Japanese people, culture, not just about their favorite anime/games. Almost everybody either had one to Japan or had plans to, almost always in study abroad context, not standard tourist stuff. So: “the cosmopolitan western otaku”.

Can even lead to business: e.g. Carpe Fulgur, who brought Recettear to the U.S., and were quite successful in doing so. Fan interest can open up new, viable markets.

3) Classics

Talked about Sophocles project from the GAMBIT lab at MIT. They had to figure out how to allow player agency while being true to the tragic ending. 

In general: there are a lot of stories out there, but we retell the same ones over and over again in games. How can we tell other sorts of stories? Classics have great source material.

4) Literature

Talks about Edith Wharton: read great example of characterization. Wouldn’t want that much text in games, but very much want characters who are such individuals. Also talked about roles in society: so much games can learn from that, too. (Not least “social games”.) How do we function as groups, as people?

1:45pm–2:45pm: “We Don’t Need No Stinkin’ Badges: How to Re-invent Reality Without Gamification”, by Jane McGonigal

I rather liked this talk. Or at least with a caveat: I didn’t take too many notes because it was going by so fast and because her slides are on Slideshare, so part of the talk turned into my nap. (I’m essentially incapable of going to a conference without napping through parts of one talk a day.) But I really like her emphasis on the positive aspects of what it could mean to bring games to a broader context, and her using the term “gameful” to refer to those positive aspects (emphasizing the intrinsic benefits of games, as opposed to “gamification”, which is generally used more for the extrinsic reward aspects of games). “You don’t need a badge when you have real power.”

3:00pm–4:00pm: “Intuition vs Metrics: How Social Game Design Has Evolved”, by Laralyn McWilliams and Brenda Brathwaite

This was my favorite talk of GDC so far. I don’t think I’ll be able to give a good feel for that here, however; go to Brenda’s blog for the slides, but imagine that they’re delivered with even more character than in the slides.

But beyond that: it was just a super sensible talk: of course you want data when developing games, of course you need to develop and apply judgment to make best use of that data, of course there are minefields that you’ll encounter when navigating this. (Especially because of the culture clash of traditional game design with traditional web app design.) This is a message that still bears repeating over and over again, and we still need a great deal of compassion while figuring out how to move forward.

My notes are below, but I’m not sure there’s anything in them that you won’t get from the slides.

Collision of worlds: traditional game industry versus web apps. Could be great, or one side could consume the other, with bad results. Both worlds have a way of doing things, both think they are right.

Traditional game design: level 30. Social game: level 3. But social game success is undeniable. “Yay! Cake. Ooh! Money!” (Is the “real cake” that traditional game developers make even something that Facebook gamers even want to eat?) Brenda: to make a social game to me is harder than making a Wizardry game.

Both sides nervously hire people from the other side.

Game designers: “I just know it’s going to be fun.” Worked for Wizardry, but to people from the web space, it’s crazy not to ask users and try things out.

Traditional game design: jam shit in, rip shit out, until you find the fun. Clash with web space PMs: your idea of fun is nice, but I think X will be a better user experience, how do we really know? Any data?

MMO experience: sure, jam shit in, but start with shit that’s similar to what you’ve seen work in previous games. And shipping the game is just the start.

To traditional game designers, shipping an unfinished game is terrifying. It might be horrible; what’s worse, people might see a horrible it and love it! Intellectually, alpha testing makes sense; and maybe a couple of years ago you could have released quietly; these days, everybody will notice as soon as you launch.

“The business and Silicon Valley culture is the game.”

Numbers sometimes tell you what you already know. Even if not, though, they’re just symptoms. So: don’t leap immediately from numbers to solutions.

Creativity flourishes within constraints: metrics are one way of providing those constraints.

Don’t blindly follow metrics. Game design is the map and the plan; metrics is the weather report which may cause us to re-evaluate.

Success story one: accidentally released Ravenwood Fair with super easy numbers on various design parameters. To make an appropriate challenge, set numbers back; monetization plummeted. Turns out that players wanted a lot less challenge than Brenda expected; good thing they made that mistake and had metrics to help.

Success story two: Free Realms launched; play time was great, but players didn’t progress beyond level 1. Tried simple fixes, didn’t work. Eventually, observing players, found: camera was a problem, but more important, they were partying! Having a great time, they didn’t care about quests. Would have been hard to find root cause with metrics (according to Laralyn, at least), but very useful info going forward.

4:15pm–5:15pm: “Rapid-Fire Indies”, by ten speakers

Not much to say here: I’ve pretty much given up on trying to convey the experience of microtalks. A pretty good series of microtalks, though.

Chris Hecker: AAA Indie games: polished to perfection, clearly contains lots of love, highly anticipated before launch. First two obvious, talk about the latter.

Petri Purho: Didn’t say much, then played a music video.

Eddy Boxerman: Old devs: he’s turning 40. (Yay!)

David Hellman: Just pictures, but good ones.

Kyle Pulver: Game jams are awesome.

Chris DeLeon: Lots of 20-second bits, followed by Thoreau quotes

Andre Clark: How to be an Indie Punk: The Story of pOnd.

Markus Persson (Notch): Piracy. Favorite bit: the idea of a “lost sale” is the stupidest thing ever. Also: game development as a service. Maybe people who pirate this week will buy next week.

Scott Anderson: Using Technology for Gameplay Innovation. E.g. using shadow physics for gameplay. Inspirations: demo scene; creative coding; molecular gastronomy.

Anna Anthropy: The words we choose as labels define the limits of our discussion. With new tools, distribution methods, game development is available to anyone: “indie” needlessly carves out a group.

Post Revisions: