Good sports learning is not only about performing a skill. It is also about seeing the situation, choosing an action, explaining the choice, and learning from the result. This was the starting point for the Softball Hit Map interactive.
In an email discussion with the SLS team, Jason Lum from the Physical Education and Sports Teacher Academy (PESTA) described a practical Assessment for Learning challenge. During gameplay, teachers want to assess both batting skill and the strategic logic behind where a pupil chooses to hit. A paper hit map can record the decision, but it is onerous to manage and makes timely, individual feedback difficult when many pupils are active at the same time.
The question was therefore simple but ambitious: could a digital hit map make pupils' tactical thinking visible and provide immediate, contextual feedback?
Starting with Gemini in Google AI Studio
The first working version was developed as a Google AI Studio app with Gemini. The source code was then downloaded as a project ZIP so that it could be inspected, run locally, and improved beyond the limits of the original prototype.
This is an important part of the workflow. Generative AI can quickly turn a teacher's idea into something tangible, but a generated prototype is not automatically ready for classroom deployment. It still needs testing, pedagogical refinement, privacy decisions, platform integration, and careful packaging.
The Gemini prototype provided the foundation: a visual softball field where pupils could arrange players, identify a hit or throw target, and think through different game situations.
Improving the prototype with the Codex app
I used the Codex app to inspect the downloaded source code and develop it into a standalone interactive that could work inside SLS. The improvement process was not simply a visual redesign. It focused on making the activity dependable, assessable, and useful in an actual PE lesson.
The resulting interactive now allows pupils to:
- select game scenarios, the number of outs, and a grounder, line drive, or high fly;
- position fielders, teammates, and a hit or throw target on the field;
- drag markers, refine a tactical plan, and use undo or redo;
- explain why the chosen target is appropriate;
- receive immediate feedback about gaps, nearby fielders, runner movement, force-out risk, double-play risk, and possible collisions;
- revise an explanation after receiving coaching prompts;
- use a four-part reasoning rubric covering the open gap, runners, trade-offs, and risk;
- copy an evidence summary or export the completed work as a PDF.
For SLS use, Codex also helped add xAPI state capture. The package can record the scenario, field layout, target zone, pupil justification, revision, self-check results, score, and targeted coaching questions. This gives the teacher more than a final mark: it preserves a trace of how the pupil was thinking.
The interactive was also rebuilt as a self-contained SLS package. Its scripts, styles, and xAPI bridge are bundled so that the activity does not depend on a live Gemini call or a browser-side API key. The coaching response is generated locally using transparent softball rules and evidence checks. This makes the activity more reliable in a managed school environment and avoids exposing an API credential.
Why this is useful for sports learning
The strongest feature of the Softball Hit Map is that it separates physical execution from tactical understanding without treating them as unrelated.
On the field, a pupil may not always get enough successful contacts to demonstrate every tactical idea. In the interactive, the teacher can present many situations quickly: runners in different positions, different numbers of outs, and different defensive arrangements. Pupils can predict, justify, receive feedback, revise, and then test the idea in play.
- See: Read the game situation and defensive space.
- Decide: Place the intended hit or throw target.
- Explain: State what the runners should do and why the choice is suitable.
- Check: Consider gaps, alternatives, and risks.
- Revise: Improve the plan before returning to gameplay.
The interactive also supports differentiated teaching. A beginner can focus on identifying open space and avoiding a nearby fielder. A more advanced pupil can discuss hitting behind a runner, tag-up decisions, force plays, launch angle, defensive depth, or the risk of a double play.
A practical pitch for PE teachers
This is not intended to replace observation, gameplay, or teacher judgement. It is a formative tool that helps a teacher see what is normally invisible: the reasoning behind a pupil's action.
It may be used before a game as a tactical briefing, during a station rotation, after a play as a reflection task, or as an SLS follow-up. Because pupils receive an immediate prompt, the teacher can spend more time listening to higher-value explanations and supporting pupils who need direct coaching.
The larger lesson from this project is also worth sharing. Gemini made it possible to move rapidly from an educational idea to a working prototype. Codex then helped turn the downloaded code into a more robust, transparent, mobile-friendly, SLS-ready learning object with evidence capture and a deliberate feedback loop. The value came from combining teacher knowledge, generative prototyping, and careful engineering.
Project files
- Live interactive: Launch Softball Hit Map
- Shareable source package:
softball-hit-map-source.zip - SLS upload package:
softball-hit-map-sls.zip
The source package excludes node_modules, generated build folders, environment files, and other unnecessary bulk. Dependencies can be restored with npm install.
Acknowledgement: The educational problem and prototype context came from Jason Lum and the PESTA digital think tank team's exploration of scorable interactives, immediate feedback, and richer sports learning in SLS. The initial app was prototyped in Google AI Studio with Gemini, then further developed and packaged using the Codex app.