The Rotex Recap
Rotex, the abstract strategy board game I released in March, was a bit of a flop, unfortunately. I wasn’t expecting wild success (it doesn’t quite have that addictive quality Five-O has), but it definitely performed below my expectations, so I wanted to reflect on that and tease out some lessons from the failure.
Background
Six months ago, I left my day job as a programmer/designer in the mobile games industry and embarked on a new adventure to develop my own games full time. I was motivated by the fact that Five-O, even after several years and very little personal attention, was continuing to gather positive reviews and earning a stable, albeit small, income. If one app could earn $10-15 per day seemingly indefinitely, and if I could quickly develop several more similar apps, maybe I could grow this into a business!
Thus, with wide-eyed enthusiasm, I began prototyping new game ideas. My goal was to design a game that could sit on top of the Five-O “engine.” Basically, that meant designing a two-player game that would be fun to play online, asynchronously (via turn-based multiplayer). Since hours might pass between your turns in an asynchronous game, every turn should be interesting and require careful thought. A lot of card games and other traditional games don’t fit this model because they have very quick turns. Also, players shouldn’t be expected to retain a lot of information in their heads about each individual match. After all, they could be playing in lots of matches simultaneously! Games with hidden information, or with sequences of actions that span multiple turns, can be very difficult to play asynchronously. So with all this in mind, I came up with an idea for a hexagonal variation of Pentago, which I initially called Hexi, and which later became Rotex.
One of the strategies I employ when working on a game design involves building an AI opponent. It seems crazy to write an AI at such an early stage, but I find that I save time in the long run because it helps me discover balance issues quickly without requiring a lot of user testing. In my opinion, a proper AI opponent is also an essential feature for a digital board game, so I also need to make sure that it’s possible to create one. Plus, I just like coding board game AIs.
What I discovered was that the game was more frustrating than fun. My AI opponent would always beat me with a move I didn’t anticipate, and if I let the AI play against itself (another useful testing strategy), games would almost always end in a draw, as if it were just a complicated version of Tic-Tac-Toe. My attempts to fix these issues through new rules and changes only made the game more complicated and less elegant, doing more harm than good. In the meantime, I had also begun working on another, completely different game prototype which seemed to have a ton of potential – a 2D kart-style game with real-time multiplayer (more on this in the next blog post). So, my focus shifted away from Rotex for a while.
Fast forward to mid-January. An issue had cropped up in Five-O where the app would sometimes crash when loading certain Game Center matches. I was able to address the issue, after which I made a few other much needed improvements and released an update. As a side effect of working on the update, I now had the Five-O codebase fresh in my mind again, so I contemplated taking a stab at finishing Rotex. I estimated it would require about a month to take Rotex from prototype to finished product, using the Five-O codebase as a template. So, even if the game wasn’t perfect, at the end of the month I would have another product in the App Store which would energize me and help me learn even more about the market.
So, I spent the month of February finishing Rotex. During that time, I developed an art style that I liked, and discovered some small tweaks to the rules which greatly improved gameplay. I even found time to implement an interactive tutorial, which I felt was absolutely necessary for an abstract game like Rotex. I struggled to develop an AI opponent that wasn’t either God-like or hilariously bad, and I’m not sure I succeeded in my efforts. But, overall, I was quite pleased with how the game turned out.
The Numbers
Rotex was released on March 16th, 2014, and got about 1000 downloads in the first week. That would have been great, except that I had decided to make the game free, using only advertisements to generate revenue, and the number of daily active users (DAU) was low, averaging around 150 per day. My marketing and cross-promotion efforts were succeeding in generating a nice trickle of new downloads, but players weren’t sticking around to play very often.
After a month and a half in the App Store, Rotex has been downloaded 2,189 times. During that time, only 328 online multiplayer matches of Rotex have been completed. Single player mode is much more popular, with almost 10,000 matches completed. But, for comparison, Five-O players, during the same time period, completed about 211,000 single player matches!
In terms of revenue, the advertisements in Rotex have only generated about $22 so far. On March 28th, I released an update with an in-app purchase option to remove ads for a dollar, and that has generated $5. Even though the game has been a commercial failure, I’ve learned some valuable lessons, a few of which are summarized below.
Lessons
1. Ad supported business models don’t work for niche games. Ads work best when a game is addictive and appealing to a broad audience. This might explain why there seems to be a higher ratio of paid apps to free apps in the more niche App Store categories, e.g. board games. If I could start over, I would drop the ads and sell Rotex for one or two dollars instead. Also, having the option to pay to “remove ads” is not a good alternative. It’s fine in principle, but users would be much happier to hand over their money if they got something else in addition to having the ads removed.
2. User testing should never be skipped. If I had spent even a small amount of time just watching other people try to play the final build of Rotex, I would have noticed an obvious user interface issue. In Rotex, after placing a tile and rotating the board, users must “submit” their move by tapping the play button, exactly like in Five-O. I discovered, too late, that many people expected moves to submit automatically after they made them. I suspect this is why a larger percentage of online matches go unfinished in Rotex; players may be making their moves but never submitting them. I recently released another update with a modified tutorial and some animation to draw the eye to the play button after a valid move is made. If that doesn’t work, I may need to rethink the user interface design.
3. Cross-promotion is a cheap and viable source of downloads. Everyone knows the App Store has discoverability issues, especially for indies. If you’re not lucky enough to get featured, you have to fight for every download. By building your own network of high-quality apps, and cross-promoting between them, you can make all of your apps a lot more visible without an expensive investment (besides building the apps in the first place). Currently, my network is only two apps, but I plan to add several more in the coming months.
4. Leveraging existing codebases leads to much shorter development times and fewer bugs. This is pretty obvious, but I was surprised at how much of a difference it made. I’m currently developing another game from scratch, and it is taking months of work to get to a minimum viable product. In contrast, I developed Rotex from prototype to finished product, complete with a battle-tested online multiplayer system, in about one month by using Five-O as my “template.” Not only that, I also fixed a few bugs and made improvements that I was able to funnel back into Five-O. From now on, when considering game projects, I will definitely prioritize the ones which leverage my existing work.
What’s Next?
I have some ideas to turn Rotex around in the future, perhaps by re-releasing it as a paid app. Currently, however, I’m working hard on my next project, which I hope will be a successful entry into a new and exciting genre. If all goes well, it should be ready to release within the next couple of months. Watch for a more detailed announcement soon!