the onda

I've blogged about how much fun World of Goo has been (to be surpassed only by the recently discovered iPhone game Rolando). Gamasutra has a great post on the value of rapid prototyping in the creation of WoG's predecessor, "Tower of Goo," which was borne out of a Carnegie Mellon project on experimental gameplay.

Though the piece is nominally about techniques for designing video games, anyone involved with building any kind of consumer software should be very interested in this piece, especially those who work in the very malleable and fast moving environment of web services. I've often invoked the principle of "tracer bullets" on teams that I've worked on to try to get folks thinking about the minimal amount of work possible to get something that might demonstrate value in front of real users. From now on, I'll also be including references to this piece.

Some of my favorite pieces of advice in the post:
* give yourself less time than whatever you think the minimum time you need to get something out (< 7 days seems like a good rule of thumb).
* formal brainstorming sessions suck and aren't worth the time. Intersperse the creativity while you work.
* fake as much of the magic as possible; users care about the experience, not your great engineering.

And some of the things I learned from the piece:
* get a team of generalists who have the right attitude; they will be much better than skilled specialists with the wrong attitude.
* prototype in parallel to make sure people aren't blocked while prototyping.

I swear that if all of the VC-backed consumer Internet startups took this advice to heart, as a sector we would waste 50% fewer investor dollars than we currently do. Which might imply 2x the number of successes.

A great read.