In my last post, I mentioned that my last game, Amnesia Island, had done reasonably well on the Nook. Over a 3 month period, the game had been downloaded over 50k times, which is good performance (in my opinion) for a game which is based on a set of memory tests for rats.
Currently, I’m working on a new game (titled Taxi Dash), which I’m hoping to release on the Nook in the next couple of weeks. As I get closer to submitting the game to Barnes and Noble, I have become a bit concerned about how the game performs on my Android test devices. According to my in-game counter, Taxi Dash is running at about 50 frames per second on my Samsung Player 4.0, and my first generation Kindle Fire. But, that same version only runs at about 20 fps on my Nook Color (the original tablet released by Barnes and Noble). And, at that frame rate, the game feels a bit laggy to me.
So, since the Nook Color is no longer sold by Barnes and Noble (though you can still get on on Amazon, ironically), I’m curious how many Nook Colors are still in use (because I don’t want to miss out on a large portion of the Nook market, and I don’t want users of the game to have a bad experience). I haven’t found much information online on Nook device stats, so I turned to my own experience with Amnesia Island. Of the 50k installs that Barnes and Noble reported, about 46,980 were recorded on my servers, so I would estimate that I have analytic data from most of the installations of the game. And for each game session, I collect the device model, so I can get a sense of what types of devices/computers my players are using.
Of the sessions that were reported to my server, about 30% were submitted by Nook Colors, and 50% by the Nook Tablet (which is the device that followed the Nook Color). Only 20% of the installs I observed came from the most recent devices (the Nook HD and HD+).
So, if the installs that I see for Amnesia Island accurately reflect the group of users who might try out Taxi Dash, I’ll expect to lose out on about 30% of my possible installs if I don’t support the Nook Color. But, if the Nook Color users are more likely to have a sub-optimal game experience, do I really want 30% of my potential game ratings to come from the Nook Color? Well, Amnesia Island had a similar frame rate on the Nook Color (though not as low as I’m seeing in Taxi Dash). To get a sense of how the user experience for that game was different for the different Nook models, I looked the Amnesia Island installations, and tried to see how much of the game each users completed. My assumption here was that if the users of the slower devices (Nook Color) enjoyed the game less (or were more likely frustrated by the game), they would be more likely to end the game early and less likely to complete each game session.
Of the roughly 47k installs of Amnesia Island that I looked at, I found that overall the users of Nook Colors were indeed more likely to end their first game session early (before finishing the memory test), and were less likely to ever complete at least one game session. Shown in the figure to the right, you can see that the percentage of installs where I never saw the user complete the first trial of their first game session. This percentage was highest for the Nook Color (the most dropouts), and lowest for the HD/HD+. The difference is not enormous (about 55% of Nook Color users dropped out on the first trial, compared to 43% for the HD/HD+), but a difference of greater than 10% suggests to me that the Nook Color users may be having a less enjoyable experience than the users with the later versions of the Nook.
Similarly, if I look at each install to see if a user ever completed at least one session of the game (so, they may have dropped out early in their first session, but did they ever come back and try it again?), I see a similar pattern. Only 10% of Nook Color users have completed at least one game session, compared to 14% for the Nook HD and 16% for the Nook HD+.
Overall, the ratings for Amnesia Island have been poor (average of 2 stars over 100+ ratings). I expected this, as Amnesia Island is less a game than a demonstration of some classic rat memory tasks. But, I would guess that those users with a Nook Color, who likely experienced a game that was loaded slower and ran slower than the same game on a higher end Nook, probably gave the game worse reviews. (Note to Barnes and Noble: it would be useful to me to see the device that was used to submit a rating!). I expect that if I had restricted Amnesia Island to Nook Tablets or higher, some of those negative ratings and reviews would not have been submitted (but, that is just a guess).
So, what do I take away from my own install data? Well, if Taxi Dash does not run well on a Nook Color, then I won’t target that device when the game is released. But, seeing the number of Nook Colors that are still in service, it is worth my time (and other Nook developers) to continue to support the device.