Postmortem: RGB Ninja

Well, I managed to finish the game for the Windows Phone 7 (http://www.windowsphone.com/en-US/apps/5606a89c-eee7-4e69-a15b-d5c498d79c0c), and release it on the Marketplace. I’d like to talk a little bit about what went right and what went wrong (like Gamasutra’s postmortem :)). Before I begin, I’ll post two images, one with a graph showing the number of downloads over time, and the other showing the number of downloads per country.
 

What went right:

Actionscript 3 /Flashpunk

The game started as a small Flash prototype. The initial idea was to mix Canabalt with Portal, and from there, it changed and evolved to what it is today. Making those prototypes in AS3, and utilizing Flashpunk, allowed me to test ideas quickly, and have feedback from my friends easily, by sharing a website link so they could play it. Below are a list of the prototypes I created:

http://megaswf.com/serve/1113136/ (3 lanes, Shooting and Near Miss mechanics, from racing games like Need for Speed and Burnout)
http://megaswf.com/serve/1114413/ (Near Miss, mouse control)
http://megaswf.com/serve/1152954/ (Colors mechanics introduced)
http://megaswf.com/serve/1153836/ (Colors mechanics changed, similar to final version)
http://dl.dropbox.com/u/1053748/RGBNinja12.swf (Close to the final version of the game)

C# / XNA

Making the game released version using XNA, was the best choice for producing it in the shortest amount of time. Since I already had participated in Microsoft’s competition, Dream.Build.Play, I already had some familiarity with the framework, as with the C# language. Besides being a really good and simple framework, Microsoft has a bunch of free game templates that you can get for free at their website (http://create.msdn.com/en-US/education/catalog/). Basically any simple thing you may want to do, there’s an example there, well commented, that you can use as a starting point.

Small Team

Having been a part of a few projects, I noticed that the bigger the team, the harder it is to finish something, and make it cohesive. From what I can tell, to have a big team, people need experience and leadership. So, having only 3 people on the team (me as a programmer, Rafael Fernandes and João Vitor Munduruca as artists) made the tasks easy to organize, and quickly delivered. To help us organize, we used Trello (https://trello.com/), a great online, and free, tool. It is really simple to use, allows you to create boards and tasks, distribute it to members, set deadlines, checklists, comments, upload pictures, and many others. It’s really good to see how a project is coming along and what each member is doing.

Portability:

Fortunately, I found a tool called MonoTouch (http://xamarin.com/monotouch). It allows developers to use C# to develop applications for the iOS and Android. It’s not cheap (starting price at $400 USD), but they have a student discount, so I could get it for $80 USD. Using it still doesn’t allow you to use XNA. Fortunately again, there’s a open-source project called MonoGame (http://monogame.codeplex.com/), that can be utilized with MonoTouch, to use XNA. I have already made some tests on the iPhone 3GS, and iPad 1, and I could port a version of the game in 1 day, running smoothly on both devices.

What went wrong:

Windows Phone 7

Windows Phone 7 is a new platform, being released about one year ago, and only recently released it in Brazil. That means that the number of devices is much lower than the other platforms (iOS and Android), which means that less people will have access to the game. Besides that, we didn’t do any marketing for the game, besides posting it on our Twitter and Facebook accounts. However, as most of our friends are brazilian, and almost no one has an Windows Phone here... Well, not many people got to know about it, and those who did couldn’t play.

Launching / Exposure

RGB Ninja was the first game I ever published. I had heard that Microsoft submission process was really thorough, and I would probably have to do some programming to handle a few cases (for example, the user is listening to music on the device, launches the game that has a soundtrack. The game should ask the user if it wants to keep listening to its song, or switch to the game’s audio). That’s why I decided to upload the game, in a playable state, but with a couple of bugs. While it went through the submission process, I would work on it, and after it didn’t get approved, I would be able to see all the specific cases I had to deal with. Turned out that the game did get approved, and I had chose to automatically publish it after the submission process (my fault :D). Luckily, I found an option to hide your application from the marketplace. The problem is that hiding it doesn’t change the fact that your application already got published, and it should appear on top of the “New” list in the Marketplace. For a game with no marketing, that’s probably the period where it gets the highest exposure. The result was that, after two weeks I updated the game, and “unhid” it. By that time, the game was already really far down the “New” list, and it got no exposure at all.