Run one focused exercise
Ask each student to start with a familiar action such as sorting, timing, or route choice. Change exactly one rule, predict its effect, then play a short version and record what actually changed. For instance, compare a sorting puzzle with unlimited bins against one with a single limited-capacity bin; the design question is whether the new constraint creates a choice or only extra taps.
Use a repeatable prompt and note sheet
Example prompt: “Create a one-screen sorting game with four pieces and two matching bins. First make each bin unlimited. Show the goal and restart clearly. Keep the piece order fixed.” After playing it, change only one instruction: “Give each bin room for two pieces and keep the same piece order.” Students can note their prediction, the action they tried, and whether the observed result matched. This is a prototype exercise; do not treat one attempt as evidence of an educational effect.
Compare observations, not scores
Have students report a concrete moment: which piece they chose first, where they hesitated, or what feedback explained the rule. Compare predictions with observed play, then let each student revise one sentence of the prompt. A clear report separates the designer's expectation from what the tester actually understood.
Use fictional content and follow the school’s own account and device policies. The platform is for people 13 and older, and game versions publish publicly; avoid identifiable student work unless the institution has approved that use. If private class-only projects are required, choose a tool that provides that scope.
Questions about game design exercises for students
What age group is this exercise intended for?
This page is for teens aged 13 and older. It is not a primary-school activity, and any use in a school setting should follow the school's own policies for accounts and public work.
How can a student tell whether a rule change helped?
Compare a specific prediction with an observed action or point of confusion. Avoid treating a higher score or a single session as proof of a learning benefit.