Mobile game development / DIJION
What should you test in a first mobile game prototype?
Explore a core loop with one clear question, neutral tasks, an observation sheet and a decision for the next iteration.
A first prototype does not need to represent the whole game. Start with one uncertainty: whether a control is understandable, a movement feels satisfying or a goal is visible. Asking whether it is fun rarely tells you which part to change. Break that question into observable behavior that can guide another experiment.
Describe the loop as action and response
Consider a fictional game: tapping makes an object jump, clearing an obstacle awards a point and another obstacle appears. This is not a description of a Dijion release. The action → response → next decision chain can be examined before adding a theme or progression system.
Locate the uncertain point. Does the player understand when to tap? Can they connect the tap to movement? Can they distinguish a successful attempt from a failed one? Each question needs a different prototype and observation.
Choose one main question for the session
For example: Can a player deliberately change their timing after the first obstacle? This can explain more than the number of obstacles cleared. Write the observation criterion before the session, while avoiding broad conclusions about the player population from a small exercise.
Scroll the table horizontally to see all columns.
| Decision | Example |
|---|---|
| Question | Is the relationship between tap timing and outcome understandable? |
| Required pieces | One tap control, one obstacle, a clear result and restart |
| Excluded pieces | Shop, mission list and extensive progression |
| Observation | What does the player change on the second attempt? |
Give the task without giving away the solution
“Press the button on the right to pass the obstacle” prevents you from learning whether the control can be discovered. Try a neutral introduction such as “Explore this section in your own way; you can describe what you think is happening.” Record when assistance is requested and what you explain. Assisted performance is different from independent discovery.
Explain the purpose, voluntary participation and note-taking method. A session code can support useful observations without collecting names or contact information. Obtain separate permission if recording is necessary. Clarify that the prototype is being examined rather than the player’s ability.
Record behavior before interpreting it
Scroll the table horizontally to see all columns.
| Moment | Behavior | Player explanation | Next question |
|---|---|---|---|
| Entry | Where do they look and tap? | Goal in their own words | Is the goal visible? |
| First result | Do they notice the response? | How they explain the outcome | Is feedback sufficient? |
| Retry | Do timing or actions change? | What they want to try | Is the control understood? |
| Exit | Where do they stop or ask for help? | Reason to continue or stop | What should change next? |
“The control is bad” is an interpretation. “They tapped the decorative area on the first three attempts” is an observation. Keep them separate. One person’s behavior can reveal a question worth investigating; it does not establish prevalence or a success percentage.
Change one meaningful variable next
If the goal is unclear, reconsider the explanation or hierarchy. If the action is understood but its result goes unnoticed, improve feedback. Once the relationship is understood, investigate difficulty or the wish to repeat the loop. Changing everything at once makes it harder to learn what helped.
- Keep: A relationship the player understood without help.
- Change: A specific part supported by an observation.
- Revisit: The same task in the revised version.
- Unknown: Long-term interest, revenue or broad audience success.
Keep the limits visible
A short session does not validate acquisition cost, retention or commercial success. Devices, grip, sound and interruptions need their own later checks. The value of an early prototype is a better-informed next decision, not an unsupported success claim.