June 4th

2:00 PM

I made good progress on the convergent algorithm this morning, but today’s report concerns something else entirely. I’ve been having an email discussion with Alex Vostrov about this project. He pointed me towards Fate of the Planet, a recent game that addresses environmental issues directly. I took a look at their website (I can’t play the game because they don’t have a Mac version — yet) and tried to read between the lines with my designer’s eyes to figure out the critical factors at work in their game.

It’s definitely a good game; I was really surprised at the lush cosmetics combined with the low price, so I took some time to estimate their costs. Their team includes five directors, seven development people, three promotional people, and eight consultants. That’s a full-time staff of 15 people plus some contractors. This thing was not slapped together overnight; let’s assume that it took them a year to build. At $50K/head/year (that’s starvation wages), they probably spent nearly a million bucks on this baby. I am astounded that they were able to find that much capital. Even more striking is the fact that they’re charging just $10 for it. They have to sell 100,000 copies just to break even! Maybe I’m out of touch with the games market, but that seems a steep break-even point. Of course, it could have taken them less than a year, and some of their people could have been working on other projects as well, but even if we halve the cost estimate to $500K, that’s still 50K copies they need to sell. I don’t understand the economics of this project.

Anyway, the game has lots of nice features and is certainly lovely to behold. The guts of the game are difficult to guess from their website, but their simulation looks fairly good. Not as thorough as Balance of the Planet 1990, but still pretty good. It looks to me that they made a big effort to make this more like a game and less like a simulation — certainly an imperative if they want to reach break-even. I definitely want to get this game when the Mac version comes out.

My own strategy with BotP 2011 is to keep the overhead way low so that I can make break-even with lower sales. In order to compete with FotP, I’ll have to come in at a significantly lower price.

Getting back to Alex Vostrov, he pointed out that the original BotP was rather user-hostile. It is a complicated and difficult game, I agree. I think he’s put his finger on a good point. After mulling it over during lunch, I have come up with two possible ways to make this monster a bit friendlier. The first is to have an intelligent agent popping in to give advice as needed. I know, everybody remembers Mr. Paperclip and thinks that the idea is a loser, but I think that, done intelligently, such an adviser can greatly help the player. I think he should use audio rather than video, as that puts a bit more character into him, and audio costs only about 3K per second, meaning that I can have a thousand seconds of audio for just 3 MB — no sweat these days. I can do the audio myself. As for the character himself, I’m torn. My preference would be to use Bullwinkle, but I can’t afford the licensing costs. Perhaps I could use a custom character, but that would entail artwork costs. I could simply use myself, but I’m ugly. There’s some benefit to using my own visage, as it would make the connection between designer and user closer. Even though the advisor window will be small, I don’t want to include animation; that’s probably too expensive. I definitely have to have a “Shut up!” button to dismiss the agent; but I also want the agent to re-appear if the evaluation algorithms conclude that the player’s need for advice exceeds the degree of determination with which he’s dismissed the agent.

The other factor I need is humor, and lots of it. I think I’m up to the challenge of loading this thing with humorous elements, even if those elements aren’t knee-slappingly funny. I won’t reveal my ideas at this time, but I’m already laughing about all the amusing things I can put into it.

Enough fun. Back to the grind of getting those convergent algorithms working