Postmortem: Pangalore's Knightly Adventure
June 21, 2013 Page 6 of 6
5. Skewed Focus on the High End
Our game design focused on high-end graphics, leading to performance issues at launch, especially on older mobile devices. The takeaway for us was to be brutally honest about which devices will work with your game and target your game experience toward those devices. As both the iOS and Android markets continue to be fragmented, this is a critical issue. If you have a 3D action-oriented game like ours you can run into memory and performance bottlenecks very easily.
We knew as we approached launch that we had problems with devices like iPod touch 4th generation and iPhone 3 -- they had worked under iOS 5 but with the iOS 6 upgrade and our own additions we were running into memory crashes. Unfortunately, in our case we could not change the supported iOS platforms text in our iTunes description. Devices such as iPod touch 4 that were now problematic were baked into our app information as being supported.
As a workaround, we updated our main iTunes description and added device checks to warn players that they would experience problems. If a player with an unsupported device tries to load the game they get a pop up warning advising them of this and the game exits so they can uninstall it. However, this was not ideal and some players were understandably upset when they game they had downloaded would not run!
With our most recent update, the issues were largely fixed by compressing our graphics much more heavily. Our environments are not as pretty but ratings are trending toward 4 stars on iTunes. Our next update will restore some graphical sheen with improved resolution and memory optimization so things will get even better (we are stealing some animation frames that are not necessary to the play experience, among other tweaks). However, would have been better to fix this BEFORE launch.
Note higher compression in the 'after' screenshot
With a multi-platform, cross-genre game developed in an international environment, we gave ourselves a high degree of difficulty for our first flagship game. Part of the fun of developing games is constantly challenging yourself, though, and we have no regrets for trying to do something different.
Even though our team was battle-tested in PC online development, the combination of developing in Unity for mobile platforms and trying to support a large range of devices brought fresh challenges. While veteran mobile developers may say, "I told you so!" when they see some of our missteps, there's no substitute for first-hand experience and our team got plenty that will be useful in the future. And hopefully some of our experiences will help other mobile teams do an even better job on their first titles!
Fortunately for us, we were able to achieve a good degree of success with our first title and we hope to achieve even better player experiences with our upcoming titles. Our next game, ZooVale, takes some of play mechanics from Knightly Adventure (although in a completely different world populated with cartoon-style talking animals) while executing them in HTML5 for a less graphically intense experience that will run in any browser, including mobile browsers, as well as on iOS as an app. We're excited to see how far we can push HTML5 gaming and thus far we happy with how much performance we can generate through our new HTML5-based engine. And we are still in the early phases of our Knightly Adventure -- we are going to keep updating it to make it better by applying everything we have learned so far.
Page 6 of 6