Case Study: Touch Control in Squarefall

Project Overview

Duration: 10 months (2021)
Role: Full Stack Developer
Category: Case Study, Usability Testing, Programming
Software: Unity, C#, Blender, Photoshop, Aseprite

Squarefall is a solo-developed game using Unity coded in C#. I created all the scripts and assets except the soundtracks.

The development process included user research, rapid prototyping, usability testing, and user flow analysis. I worked closely with players to validate design decisions and iterations weekly. The project was launched on Google Play in August 2021.

case study

During Squarefall's development, usability tests revealed that half of the players had difficulty understanding the game controls. While initial adjustments helped these players, the other half however, was then confused. The final solution was an adaptive tutorial that automatically detects users' preferences.

The problem

Players were moving in the opposite direction from what they desired.

Squarefall's control is primarily using drag and release for the character's movement. However, half of the players had problems learning the controls.

Usability test - a

Touch control dilemma

Tests revealed that there were two types of players: "slingshotting" versus "flicking" and they each consisted about half of my testers. Whichever control I chose appeared to confuse the other half of the testers even when informed of the intended controls.

I later realized that the placeholder of the direction indicator shared resemblance with drag indicators, which confused players. The direction indicator was later reworked.

Usability tests were conducted both in person and remotely. Before each test, new subjects were recruited to ensure there were subjects that had first time experiences. Research data were collected from interviews and scripted events in the app.

Usability test - B

providing options

To accommodate players that prefer flicking and slingshotting, I added a toggle for inverting the controls. A tutorial with video demonstrations and text descriptions was also added. The findings were as the following:

Players didn't read nor watch the video. They just tapped through everything to jump right into the game.
Players still missed the option to toggle control.
Some players quit the game during the tutorial due to the unfavorable controls.

problem reframed

How might I make the onboarding process the shortest and smoothest possible?

The default control settings could only satisfy half of the players. The game had limited time to guide players to the settings best fit them.

A script was added to detect if players initiated control with slingshotting or flicking. Settings were applied automatically without the need for players to figure out if they had the ideal settings.

To ensure players initiated the control that was the most natural to them, they were presented with multiple nudges: objective text, target enemy, animated trajectory and direction indicator.

Results

I used Unity's metrics to track the conversion rate of the tutorial:
‍
Before the implementation of the adaptive settings, less than 50% of testers were able to complete the tutorial and 30% testers were stuck in the first task.

After the implementation, almost 85% of the players were able to finish the tutorial during the open beta. Of 40 individual feedback received out of over 400 downloads, only one incident of control problem was reported.