Our first own IP project BrainJewel has been published on Facebook in August 2011. Ever since we have been trying to improve the game and drive up the retention and virality to a higher level. In total more than 30,000 players installed BrainJewel on Facebook in the period of six month without any paid advertisement or other marketing measures. BrainJewel is now running on Facebook with an average of 10,000 Monthly Active Users and 1,000 Daily Active User, with most traffic coming from Applifier. This is not such a big success for a Facebook game, but we are still happy to get continuous traffic and great feedback from our players.
We know that we made a terrific game with wonderful artwork and smooth game play. But we also know that it was created for a niche-market (brain games) and did not have the most innovative concept out there. Besides that the individual game session was much lower compared to classic social games, which made it harder to monetize. Given these factors combined with the tough competition on Facebook, the traffic and retention is somewhat understandable but still disappointing.
MAU/DAU Facebook BrainJewel
When Adobe released AIR last year, we saw a new opportunity for BrainJewel. The game felt more like a mobile/touchpad game all the time and TribePlay is always eager to explore new possibilities. The whole project started pretty much as an experiment with the hope of new opportunities.
Why iPad?
So we decided to go through with the transfer. There was not a lot of information about other companies out there that had already done it successfully. A couple of developers tried it, but mostly for smaller Flash games and BrainJewel currently includes twelve individual mini games, which meant much optimization was needed. Anyway we decided to try it and TribePlay chose iPad first for three main reasons:
1. Touchpad’s give the user in our opinion the optimal gaming experience
2. The iPad is more powerful than the iPhone, so a conversion should be easier
3. For the iPad conversion we only needed little UI changes
Issues during the conversion
With BrainJewel the biggest challenge was to get the performance to an acceptable level on the iPad. We took the web browser version of the game as a start – porting it to iPad was fairly easy. In about no time we could use BrainJewel on the iPad, Android Pad/Phone and iPhone. BUT it ran very slowlyas it was still full of vector images and the programming was not fine-tuned. PCs are very powerful compared to the iPad. We got the performance up eventually by combining a lot of graphics into bitmaps and finding some performance pitfalls. Combining graphics into bitmaps was also necessary to improve the looks of the game, as all vector graphics are rendered on lowest quality by AIR on iPad.
BrainJewel on iPad Promotional Material
At the time BrainJewel for iPad was developed, Flash did not support GPU acceleration meaning the CPU is responsible for both rendering and game logic, whereas GPU acceleration accelerates rendering by using the GPU. Halfway through our project Adobe AIR started to support GPU rendering. Unfortunately GPU rendering puts a lot of requirements on graphics in the game and there was no way for us to reasonably adjust the graphics to take advantage of it.
Social Integration, In App Payments, Language
Originally we planned to also offer In-App Purchases similar to the Facebook version, so that the user could potentially buy the soft currency Jewels (as in our tagline “Use our BRAIN to find the JEWELS” 
via the App Store. This turned out to be really difficult and we could only find one example, where it had been done before. Eventually we ended up dropping this plan, as the effort and risk seemed not to be in relation to the advantages and also TribePlay wanted to get BrainJewel out there to test grounds on iOs.
Update 2012/02/17 I found an interesting looking tool on the Web: This company is promising a solution for US$49.99. We have not test it yet, but it might be worth evaluating. Anyone tried this?
Another point was that BrainJewel was meant to connect with Facebook, so players could use their old Facebook score and progress. This also turned out to be quite an effort, so after all we did not made use of the social integration. Generally Adobe AIR does not seem to work really smoothly with Third Party APIs. This might change in future updates of AIR.
Another big issue we only discovered AFTER the release of BrainJewel to the App Store. Adobe AIR is not setting the localizations correctly, so after the release BrainJewel was suddenly marked as available in “English, Chinese, Czech, Dutch, French, German, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Swedish, Turkish”. Apple will do this automatically according to the specifications in the ipa file but the specifications cannot be changed in Adobe AIR. For our customers and us this is of course bad, because it seems we promise things that we cannot keep up with. So this had to be fixed as soon as possible and is currently awaiting approval by Apple. Anyway until now there is no direct fix provided but in the AIR forum you can find a way to work around it. Please check this before you upload.
Last time up in the AIR?!
In the end we managed to master all the obstacles Adobe had put in our way and BrainJewel is out there in the App Store. You can see the result and judge for yourself. You can also play the web version on Facebook. But overall the optimization and tweaking took almost the same time, then to write the app directly into iOS native. Using Adobe AIR might make more sense when you are building a new application from scratch and have already considered that the Flash application might be converted later. In theory it is actually great to build one application, convert it and explore four different markets at the same time. In practice there are still some flaws.
When you build an application from scratch Adobe AIR is a powerful and great tool and for that we would consider using it again. For the conversion of an “old” application it is a different thing. But the situation might change with the release of iPad 3, which will be most likely the time when most people stop using iPad 1. To make the game compatible for iPad 1 took most time to optimize due to the lower hardware and in the end we have discovered that it is not really possible due to problems in Adobe AIR with the memory of the iPad. Until that is solved by AIR converting a game for iPad 1 will always be a problem. Anyway the game is now out there and we LOVE it. Try it yourself in your local App Store 
Enjoy BrainJewel. And let us know about your experience with Adobe AIR.
