During the day I’m developing enterprise software, so you’d think I’d know how much work testing can be. Still, I was caught off guard by the amount of work it takes to test games.
Currently my typical unit of work is:
1) start game.
2) play for 30 minutes.
3) gleam in terror on newly created list of 15 tasks and 5 critical bugs.
4) spend next X hours working through list.
5) goto 1.
When I was mostly playing games as opposed to making them, I always found it bewildering when I ran into a bug: why did the developers leave that in? It’s not hard to find. How did they miss that?
Now I know it probably was very low on a list of hundreds of bugs that still had to be resolved.
There is something in the emergent behaviour of games that makes testing painful. When testing your typical enterprise CRUD application, you at least know all the usecases, and you can test them one by one.
In a game, especially in an open 3D world, there is an infinite number of usecases. The only approach to testing seems to be to brute force it – play and test as many hours as possible, in the hopes to catch all the show stopping bugs and problems before the game goes live.
At least I’m looking forward to a good rush of adrenaline, when I press the ‘release game’ button 😉