Rock-Paper-Scissors Predictor
Play rock-paper-scissors against a human 'AI' that keeps a tally of your habits, then predicts your next move and plays the counter.
Materials You'll Need
- This printable sheet
- Pencil
- 1 friend or family member to play the 'AI'
Hi, Bot Track Projects · rock-paper-scissors-predictor · Print-friendly · hibot.space/track-projects
How It Works
- 1
The adult plays the 'AI.' Its only tool is the pattern tally on this sheet. It never guesses from your face.
- 2
Rounds 1–10: play normally. After each round, the AI adds a mark to the tally: what did you play right after your last move?
- 3
Rounds 11–20: before each throw, the AI looks at your last move, finds the most-marked box in that row, and plays whatever beats it. Keep adding marks.
- 4
Track every round on the scoresheet, including what the AI guessed you'd play.
- 5
Reflect: Did the AI win more in rounds 11–20 than in 1–10? What did it learn about you? Try one more round of 10 where you try to fool it.
Your Work
Fill in the pattern tally and scoresheet. Count the AI's wins in each half. Write how you'd beat it.
Pattern tally
| After you play… | Then Rock | Then Paper | Then Scissors |
|---|---|---|---|
Scoresheet
| Round | You | AI guessed you'd play | AI played | Winner |
|---|---|---|---|---|
AI wins: rounds 1–10 and rounds 11–20
How I'd fool the predictor
Parent Notes
You are the model. The tally is its training data, and 'most marks wins' is its whole algorithm. Play it straight: no reading faces. Most people repeat habits (like throwing rock after a loss), so the predictor usually wins more in the second half. Ask: 'What happens if you play completely at random? Can it still learn anything?' (No. Random has no pattern to learn.)
Reflect
What did you learn? What was hard? What would you do differently next time?
Made by Hi, Bot. Learn more at hibot.space/track-projects