Skip to content
Hi, Bot
All Hi, Bot games

› Hi, Bot Games

Robot Route Program it for one maze. Teach it for every maze. Find where each one breaks.

Act 1: write a program that walks a paper robot through a maze. It works, and it crashes on the very next maze. Act 2: teach the robot instead, by showing it what to do and tallying every move. Then it runs three mazes it has never seen, alone. One it solves, one it freezes on, and one it circles forever with the goal two squares away. Programs vs learning, and where each one breaks.

1–4 players · ~9–14 + parents · ~30 min · screen-free

Print tonight

Plain paper, one side, Actual Size / 100%. Print one Habit Sheet per two teams. You also need a pencil and a robot token: cut one from the Maze Book, or use a LEGO figure or an eraser with a drawn arrow.

How to play

Act 1: write a program that walks the robot through a maze. Act 2: don't program it. Teach it, by showing it what to do in four training mazes. Then let it loose on three test mazes it has never seen. Score 2 for a program that works and 3 for every test maze your robot finishes by itself.

  1. Quick start — read this, start playing

    • Every move is: face that way, then step one square. A = ahead, L = left, R = right, B = back. Hit a wall or the edge and the robot crashes.
    • Act 1: in pencil, write a program for Training maze 1 on your Program Strip. No touching the robot. Then run it. Reached the goal? +2. Now run the same program on Test maze 1.
    • Act 2: walk the robot by hand through Training mazes 1-4. Before every step, look at what's open next to the robot, find that row on your Habit Sheet, and tally the move you make.
    • Then test: on Test mazes 1-3 the robot moves alone, doing the move with the most tallies in each situation. Reach the goal within 30 steps? +3.
    • That's the whole game. Everything below is for when a question comes up.
  2. Setup

    • Lay out the Maze Book. Put the robot token on the start square of Training maze 1, its arrow pointing the way the printed arrow points.
    • Play solo, or split into two teams. Each team takes one Habit Sheet table and one Program Strip.
    • The robot only ever knows one thing: which of the three squares next to it (left, ahead, right) are open. It can't see the goal.
  3. Act 1 — program it

    • Everyone writes a program for Training maze 1, one move per box, in pencil. Don't touch the robot while you write. The shortest one is 16 moves.
    • Run each program move by move. Hit a wall or the edge = crash, and the run is over.
    • Reached the goal? +2.
    • Now put the robot on Test maze 1 and run the same program. Watch what happens. A program only knows the maze it was written for.
  4. Act 2 — teach it

    • Walk the robot through Training mazes 1, 2, 3 and 4, one at a time. Take turns moving it. Choose any route you like.
    • Before every step: look left, ahead and right of the robot. Find the matching row on your Habit Sheet, then tally the move you're about to make.
    • A situation not on the sheet? Then the robot has nothing to go on there. Don't tally it.
    • Tip: pick one habit and show it every time. A robot shown three different things in the same situation learns nothing useful.
  5. Test it

    • Put the robot on Test maze 1. Nobody steers now. Each step, find the situation's row and make the move with the most tallies.
    • No row for it, no tallies in it, or a tie: the robot freezes. That run is over.
    • Goal reached within 30 steps? +3. Still wandering at step 30? It's going in circles. Stop.
    • Run Test mazes 2 and 3 the same way. Most points wins. Ties share the win.
  6. What the robot teaches

    • Program: exact steps for one problem. Perfect there, useless anywhere else.
    • Learning: a rule picked up from examples. It works on mazes it never saw. That's the whole reason to learn instead of program.
    • Only what it has seen: Test maze 2 has a crossroads. No training maze does, so there's no row for it, and a habit-taught robot freezes. Real AI gets stuck the same way on situations missing from its training data.
    • Edge cases: a habit that works almost everywhere can still fail badly. On Test maze 3 the keep-the-wall-on-your-right habit goes round in circles with the goal two squares away.
    • The robot can't see the goal. Neither can a learned rule. It only knows what its sensors tell it.
  7. Variants

    • Left hand: train the keep-the-wall-on-your-left habit instead. It beats Test maze 3, which the right-hand habit never can. Neither habit is smarter. They fail in different places.
    • More data: add a crossroads row to your sheet in pencil, teach it on Test maze 2, then try again. That's retraining.
    • Shortest program race: who can solve each training maze in the fewest moves?

Downloads

  • Instruction sheet

    Plain paper · 1 page · 8.5 × 11

    One 8.5 × 11 page for the table — quick start, setup, turns, and scoring.

  • Full rules

    Plain paper · duplex · 2 pages

    Same game, longer type — plus what the cards teach and the variants.

  • Maze Book

    Plain paper · 2 pages

    Four training mazes, three test mazes, and robot tokens to cut out.

  • Habit Sheet

    Plain paper · 1 page · pencil

    Two teams' tally tables (what the robot learns) and Program Strips (what you write).

Rules preview

Same rules as above, laid out for the table. Keep a printed copy face-up while you score.

On a phone, open the PDF in a new tab — in-page previews are unreliable in mobile Safari.

Played Robot Route?

Playtest notes — round length, rule confusion, print quality, teach-back moments — help us tune the game.

Send playtest feedback →

Want more than card games?

Want facilitator-ready printables the day they drop, plus the full Activities library for grades 6–12? Activities Pass — 14 days free, then $5/month or $39/year. Browser games like this one stay free either way.

Start 14-day free trial →

Homeschooling? Print tips & FAQ · Want the screen-free activity library? Free printables · Start 14-day free trial