Over the weekend, we tested some revisions to our tutorial. We did a quick prototype where we slanted some flat slope pieces into very gradual fake “hills” in order to quickly find out how much of a difference more gradual slope pieces would make. What we found was a little surprising: The more gradual slope pieces not only made a difference–in combination with our better balance controls from last week, they almost entirely solved the problems we were having.
Up until this point, we had been planning a very elaborate revision of our tutorial on the assumption that we needed to drastically change how the tutorial communicated how to balance Seven, but it turns out the problem was much simpler than that. We simply weren’t giving players the opportunity to practice balancing in a safer environment so that they had a chance to learn the feel of the controls.
Based on this, we created two new slope pieces to allow us to make more gradual curves, both in the tutorial and in early levels. What used to be level one in the game is now level three. We also spaced out mechanics in the tutorial and in early levels so that, for example, players can now practice jumping by collecting bones before they attempt jumping over obstacles.
We conducted one last formal playtest to see how effective the changes we began last week had been. In addition to our usual playtest survey questions, we also did some comparisons of the success rate of playtesters at our game versus two other highly successful endless run games, Doodle Jump and Jetpack Joyride. We chose these because Doodle Jump has accelerometer controls, albeit very different ones from our game, and Jetpack Joyride has a lot going on gameplay-wise, which is also similar to our game.
While we were not looking to get scientific results, we were able to use the other games as a baseline to determine if our game was giving people enough success on the first three playthroughs. We found that, on average across all playtesters and all three playthroughs, our playtesters died within the first 10 seconds of Doodle Jump, within 25 seconds on our game, and within 45 seconds on Jetpack Joyride. That may sound extreme in all cases, but it’s important to keep in mind that this is a genre meant to be played as a quick diversion for only 30 seconds to maybe 2 minutes.
Considering that in earlier playtests, we used to have a few people dying repeatedly on the first hill in the tutorial, this was a huge improvement. Our survey questions also showed far more “challenging but in a good way” responses rather than “frustrating but fun” responses, and people almost universally answered that they could feel themselves improving upon each attempt and that they would keep playing if given the chance. After a long semester of iteration in order to help our game be accessible to casual gamers, we feel like we ultimately landed in a very good spot.