It's been a busy January at Zipline Games. In addition to the Moai platform and community growing by leaps and bounds as we get close to our 1.0 launch, our in-house title Wolf Toss now has over 1 million players and a 4.5 star user review average across Android, iOS, and the Chrome Web Store. As part of the success of Wolf Toss, we want to share our experience from the launch to help other Moai developers. Mike has started a Gamasutra-featured blog series on the topic, and I've just shared some of our top takeaways on MobileDevHQ.  Here's an excerpt of the good stuff from that article:

Lesson #5: Device proliferation is real, and it will keep you up at night.

When you’re launching a game for iOS, you have to decide how many versions of the hardware and iOS system software you’re going to support. Performance differences can be substantial, some of the new iOS 5 APIs won’t be there for every player, and of course the resolutions are different as well. We had a crash on iOS 4.3 due to a dependency on the new Twitter API. We supported 3GS (which is still being sold) but dropped support for 3G. We bought several used phones just so we’d have the right devices in house for testing.

All in all, iOS is reasonably tractable. Android is an order of magnitude harder. Originally we launched Wolf Toss just for Android 2.3+, and Google reported that we were available for over 250 devices. After some further performance optimization, we expanded our Android compatibility to versions 2.2+ and Android Market now reports 565 compatible devices. We’ve wrestled with a few force close events and performance problems that reproduced only on specific devices and were found first by players.

Mitigate the situation as best you can. Get a Perfecto Mobile account and test your game on several popular models and Android system versions. Plan for a very active support period to work out device-specific kinks. And get system captures from people with issues (many Wolf Toss players have happily shared system dumps so we could fix their problems). Dheck our FAQ on the topic to see how we did that.

Lesson #4: Letterboxing may be standard for movies, but it’s not good for mobile games

Multi-platform games have to support a lot of resolutions and aspect ratios. Moai handles the resolution adjustments for Wolf Toss, and we thought we had done a good job of supporting common aspect ratios like 3:2 (iPhone and some Android phones), 4:3 (iPad and Android tablets), 16:9 (QHD Android phones) and 16:10 (other Android tablets) by optimizing for 3:2 and using the minimal amount of letterboxing.

But by optimizing for the 3:2 displays, QHD Android users got small black bars on either end of their screens and they were not happy! We were surprised with the amount of negative feedback on this topic and now have an update planned to make full use of 16:9 screens.

Lesson #3: App piracy - get used to it.

Within 24 hours of launch, the Wolf Toss packages for both iOS and Android were showing up on filehosts and rogue app stores all over the net. For Android, there seems to be an implicit assumption that if the app is free then the developer has given carte blanche to re-distribute it.

This is a problem for many reasons. The biggest one is the potential introduction of malware, especially on Android, causing potential damage to players, not to mention your brand. Secondly, if you’ve carefully tested your game on certain platforms, people will sideload their way around that restriction, and then complain the app doesn’t work. Lastly, if you rely solely on Google Checkout for payments, that will fail on any phones that don’t have Android Market installed, which includes the majority of Chinese phones.

What can you do other than fire off a blizzard of DMCA takedown notices to those hosts that acknowledge them? Not much, especially when many of the offending app stores are Chinese. Ironically, one Chinese Android store wrote to ask us if they could add Wolf Toss to their app store after they’d already posted a pirated copy!

Lesson #2: Android is huge. Ignore it at your peril.

Folk wisdom among many developers is that iOS users more often pay for content. That may be true, but Android is rapidly catching up. In WolfToss, iOS users are monetizing only slightly better (~20%) than Android users on a per-user basis. But as we posted in the post-mortem bloglast night, Wolf Toss has over 25 times the number of installs on Android as it does in either iOS or Chrome. On top of this we still have major future promotion possibilities open on Android that we don’t on iOS (feature positions in storefronts owned by the wireless operators, possible re-releases to the Amazon App Store or for specific devices like the Kindle Fire or the Nook). Overall, the much broader reach we’ve had with a three platform launch has opened us up to many more opportunities for success despite our small title promotion budget. We’re also learning a ton about how each platform operates that we’ll apply to our next game. An extra 20% per user pales in comparison to these benefits.

Lesson #1: A little extra time before launch is worth it.

We’re strong believers in iteration and learning from users, but the reality of the situation in a multiplatform mobile game launch is that you have to work through the platform providers and store operators and you need to have some wiggle room to handle issues that crop up. We quietly launched the iOS version of Wolf Toss in Canada a few weeks before the official launch, which gave us time to observe player behavior and find additional bugs. Sure enough, we found the original transition between tutorial levels was too confusing for users, and we were able to ship a corrected version before full launch.

Even with this preparation and a solid 12 hour window before a launch on December 8th, we barely made it to the finish line on time, and we had some serious hiccups. The propagation of the release to all geographies across iOS devices took about 19 hours, which meant our press release and first news stories went live 7 hours before some people could find the game on their phones. Certain screen shot updates appeared to have been updated correctly in one of the app stores but never got pushed to phones until we deleted and resubmitted the same screen shots. In future, we’ll be setting the final deadline for app store adjustments 24-36 hours before the actual launch event.

 - Todd