So Jamie and Jason are talking about game production issues, specifically Jason’s recent Escapist article on the matter.

I think they’re both right, although I think Jamie’s heart is closer to mine when he says that he’s tired of hearing how we’re all fuckups. Jamie’s money quote:

Game development is one of the most difficult endeavors known to man – it combines the unpredictability of software engineering with the unpredictability of film-making and the unpredictability of creating an interactive experience that’s different for every user…unpredictability cubed.

‘Fun’ isn’t easily scheduled. It’s a squishy concept, that takes time and iteration to get to. But those who worry about things like marketing are insistent that the game has to be sold in a certain timetable (preferably 2-4 months before Christmas) to be successful. Unschedulable task + hard deadline = almost inevitable crunching.

One of the disconnects between Origin and EA back in the day was that EA wanted Origin to come up with an iteration of Ultima and Wing Commander every year. The logic was that Madden and NHL Hockey could do it. But those products had very limited innovation from year to year – Madden historically adds only a couple of features to the design of the game on top of its annual roster update, and that works for them. EA has since seen the folly of their original logic, and are trying to find ways to add innovation to their games, while still keeping things vaguely schedulable.

Jason asks what a perfectly realized development pipeline would be. Personally, I’d like to see one focused on an ‘incubator’ philosophy, where studios ask very small groups to make something fun. Have five of six of these groups going at once, but keep the burn on them very low. Keep the expectation that there’s a very darwinian reality going on (only the best of the incubated projects gets staffing, most get killed) and people won’t feel heartbroken and betrayed when they do. More than half of all game projects end up getting killed – so find a way to make the system acknowledge that reality.

Anyway, keep funding and staffing as light as possible, until you have come up with something fun and have eliminated most of the design (and if possible, technical) risk – THEN add people. At that point, at least in theory, what you’re building is at least clear enough that the project is schedulable.