I haven’t talked much about SteamBirds on my blog. Now that things are post-release and over-the-hump, I think it’s about time I discuss the history behind SteamBirds, and some other random thoughts. Numbers post to follow.
First up, about the Secrecy; I had often knocked Closed Development practices but hadn’t actually tried them myself. I thought it would only be fair if I gave it a go. Everything I’ve done before SteamBirds has been very open (just see all these posts on Protonaut and Space Squid), and in many cases my productivity apps have had their source code released to someone.
I often thought closed development was an entirely useless proposition, especially in the Indie Space. In the end though, I made an analytical post on the subject and have now declared that Closed Development was a success for me; it helped me focus and gave me more time to put into the project. I’m not concerned about exposure and building a fan base, because this version of the game is designed to build exposure and a fan base for a later iteration of the game. I can’t say I’d make the same choice if this was my final release!
Anyway, let’s recount the tale of SteamBirds, how it came to be, and how I did it (in a very general sense).
September 4th, 2009: It all started with the Penny Arcade Expo in Seattle, Washington. I was busy working on Protonaut at the time. Greg Wohlwend, partner in the project and graphical genius, needed to take a short break from the game and it was a good chance at a mini-vacation.
I’ve been a long-time fan of Penny Arcade. I won’t say I had the pre-cognition to be a day-one fan; I think I first graced their pages in around 2000. Even when the comic fails to entertain, I’ve often found – at the very least – inspiration in Tycho’s textual truths.
I had never been to a PAX event before, despite being very close to the West Coast Venue. I was very excited to go to the September 2009 show, but was not prepared for the level of disappointment I felt. I’m fairly abreast of all industry knowledge/announcements/gameplay videos, and seeing games “first hand” a few days before launch held no special magic for me. I don’t put a premium on games that can only tease me; the same way I’d rather have private sexy-times than go to the strip club.
The table-top floor, however, saved the experience. I had a blast. I soaked up new games, unpublished games, games in testing, and old games. I reveled in it. It was like bathing in a big pool of gaming sexy-times. Board games, of course, have a turn-based monopoly… And I quickly found myself whisked away to thoughts of yester-year, where games like X-Com occupied my every waking moment. Where did all the turn-based games go? Why are there almost no turn-based flash games? Have we been swept up in a digital era of quick-action and forgotten our roots?
September 13th, 2009: Still high on the gaming experience a week before, I tried my hand at creating a turn-based table-top game. Using a set of movement mechanics and pieces of paper as placeholders, Aubrey (my girl) and I raced dirigibles across our coffee table. BlimpRace was born. The blimps could bounce off of each other and we often ended up in comical – if frustrating – situations where we just kept bumping around trying to squeeze through narrow canyon-like checkpoints. I still think the game has promise, but it had a lot of issues in both gameplay balance and pacing.
I’m a licensed pilot myself, and Aubrey has an unhealthy enthusiasm towards all things airship-related. It was just fun “playing in the sky” as it were; it’s really enjoyable engaging in something you find fascinating.
To any prospective developers out there: Make sure you love your subject matter! I started making Protonaut when I knew in my heart that I dislike platformers. Bad idea!
October 24th, 2009: Nearly two months from my initial inspiration, the first line of code is written for what would eventually become SteamBirds. It was obviously called “SexyPlane” at the time (obviously).
October 28th, 2009: After 4 days of very lazy, part-time effort (I figure 2 days full-time), the first Prototype is complete. SteamBirds never deviated far from this initial prototype. This version of the game is full of programmer-art and had only the most basic human-generated sound effects (bang! pow! boom!). Daniel Cook is excited, and in the coming weeks we hammer out a collaboration deal.
Daniel Cook made some amazing contributions to the game. His experience and insight into the industry – and more importantly – the minds of the players, was invaluable. I’m absolutely certain that without his design and guidance the game would have ended up being a very different product.
October 29th, 2009: Daniel Cook hands me the first pass of art assets and I get to make the game look really beautiful. After a few days I had a revamped playable version. We ended up going in a completely different direction in the end, but this “cut-out” art style got a lot of praise from the people I showed it to:
This is right about when I started seriously working on Steambirds. Since the core game mechanic was pretty much final, the work Dan and I were doing was dressing it up to the nines. Inventing new compatible mechanics, implementing UI, balancing gameplay, and polishing the game. I never put in a full effort; my time-tracking software says I put in 3 hours per day, 4 days per week. Very part-time. I even did a lot of code from the pub on my laptop.
Some of my downtime, of course, was spent in the shower thinking about the game – sketching ideas and writing notes – or just plain thinking about problems. It really wasn’t a lot, though. I was busy watching the entirety of Deep Space Nine for the first time.
Goddamn I hate Sisko so hard. I don’t think he can act (for TV; he’s apparantly an accomplished theatre actor/instructor) worth a damn and the only reason the show was watchable for me was the supporting cast. Christ.
December 12th, 2009: I went for lunch with Colin Northway. I expressed my desire to have a top-notch musician on this project; someone who can crank out some epic tunes.
Colin: Why not DannyB?
Me: Pff, I can’t get DannyB. He’s all famous and shit.
Colin: Nah dude, give him a shout! He’s a great guy!
Sure enough, DannyB (Danny Baranowski) jumped on the project right away, and gave it that distinctive sound. I love how the game’s opening screen grabs you by the shirt collar and says “YOU WILL PAY ATTENTION TO ME NOW.” The music is even popular enough to have generated some sales on DannyB’s bandcamp page.
DannyB was also instrumental in securing Jordan Fehr to do the excellent sound effects for the game.
I attribute a lot of the success of the game to the audio design. It really makes the experience a complete package. Just check out the trailer I made and tell me it isn’t made a thousandfold more awesome by the sound:
January 19th, 2010: The gameplay and UI has entered a beta stage. Aside from a few bugs, I would almost be happy releasing the game at this point.
Again, Daniel Cook’s contribution was immeasurably valuable. I think collaborating with him on this project not only made the game better, it made it viable. He was also great at keeping me grounded; keeping me sticking to the basics and not wandering off into “things I want to code” land.
I wanted to make this launch count, though; I began large-scale user testing.
I uploaded a copy of my game to FlashGameLicense and used their FirstImpressions service to get some initial big-issue feedback. This service selects random people from around the world (any age group, country, preferences, etc.; only requisite is they know how to use a mouse and keyboard) and forces them to play your game for 5 minutes and write up some feedback. I occasionally re-used this service throughout the coming month, and eventually got the “Average Fun Rating” from 5 points up to 7.5.
I was also user-testing in-person. As valuable as FirstImpressions are, they aren’t very detailed and you can’t watch users get frustrated with things or quit at strange times. I cannot stress enough how important real-user testing is.
I even had Daniel Cook and Colin Northway do some user tests on my behalf. Do not skirt user-testing! I spent nearly an entire month on this stage!
January 30th, 2010: I posted the game up for Auction at this point. I’ll go into a lot more detail in the next blog post about the numbers and figures on this end of things. Development essentially stopped here, and the game is largely feature complete. Minor bug fixes and statistics-reporting APIs developed.
February 22nd, 2010: Armor Games wins the FGL bidding war and even commissions me to make a few bonus levels in the game for his site. I spend half a week of full-time dev working on the requested changes and preparing the game for final release to the wild.
March 3rd, 2010: SteamBirds goes live on Armor Games. Exactly* ten years previous, Penny Arcade posted this comic:
Shit. I was really hoping the comic would be a telling metaphor for my situation! Well.. damnit. If I fudge the date a few days earlier at least I can get a topical comic about videogames as art.
All in all, the game’s development timeline was very short. If you assume full-time days, it took me 2 days to code and around a month of polish to get things to the sellable point. Maybe a further week or two of full time coding setting up things for sequels, ads, sponsors, viral versions, and various things like highscore integrations and badges for sites that support them.
In real-world-time, I spent around 3 months on the game and a further month on post-sale issues and updates. The reason for part-time work? I’m lazy. I had some other work going on on the side, but I spent way too much time catching up on old unwatched StarTrek episodes. :)
I’ll post some details on the numbers and deals next! Until then, any questions you guys have on the process? Timing? Tools I used? Feel free to ask in the comments.