Claw Game Maker teaching notes
For middle school science and STEM teachers, co-op leaders and parents. One period on why a prize slips, with the Grip Bench and the replay; a second on building a machine that passes the honest check; and the wrong ideas students bring, with where the app answers each one.
Using this with a class? Make a free class and add your students yourself: each gets a username and password, with no email or Google account needed. Or give them one join code. Using Google Classroom? Post the link there; with a free class, the link also puts each student in your class, so you see their work. mytekDev is also available from the Google Workspace Marketplace, so your school's Google admin can install it for everyone.A first lesson: One period: play the starter three times, read why each prize dropped, find a bear's holding grip on the Grip Bench and compare it with its weight and friction, then send twenty robot players at the machine and compare their win rate with a prediction. A second period to build and share a fair machine of their own.
Set up a free class Share to Google Classroom- About these notes
- Forces and friction in one period
- A fair machine, a second period
- What students get wrong, and where the app answers it
- What the app simplifies, on purpose
- Kid, middle school and high school
- The sheet, the puzzles and the setup
- Standards
About these notes
Claw Game Maker is a claw machine the student builds and then plays with real physics. The prongs are hinged bodies closed by motors whose force limit is the grip setting, in newtons; friction at the tips is what holds a prize up; nothing is attached to anything and no play is decided in advance. When a prize drops, a replay draws the forces and says why.
It is written for MS-PS2-2 (the motion of an object depends on the forces on it and its mass), MS-ETS1-1 to 4 (define, evaluate, analyze and iterate on a design) and 7.SP.C.6 and 7 (probability from repeated trials). The claw machine is the hook: most students have lost a dollar to one and called it rigged, and the app lets them find out what is true in that.
Three levels. Kid speaks in weak, medium and strong grips and plain sentences. Middle school shows the newtons and the grams. High school adds the friction line, the torque and the formulas. Setup can lock the level.
Forces and friction in one period
- Play three times (8 minutes) Everyone plays the starter three times. The starter is a 70 cm box with three rubber-padded prongs and a mild grip; careful players win about two in five. After each loss, Why did it drop? plays the grab again at quarter speed with arrows for the prize's weight, the push from each prong and the friction that push could make, and one sentence: never lifted, slipped off one side, tipped out, let go when the grip eased off at the top, or stuck across the chute.
- Read the replay together (5 minutes) Pick one lost play on the projector. The weight arrow points down from the middle of the prize; the push arrows point in from the tips; the friction arrows point up along the tips. Ask what holds the prize up. It is friction, and it has a limit set by the push and how grippy the two surfaces are.
- The Grip Bench (12 minutes) Test, Grip Bench: one bear alone under the claw. Guess whether it holds at the grip shown, run it, then turn the grip down until it slides and up until it holds. A 180 g bear weighs 1.8 N; three rubber tips on plush hold it by squeezing alone above about 0.8 N each, and on the bench it needs a little more. Then try the capsule ball: it holds at a far lower grip, because the curled tips close under it and carry it on their shape. That is cradling, not squeezing.
- One variable at a time (8 minutes) Keep the grip and swap the tips for bare metal: the same bear slides. Keep the tips and double the mass in the prize's card: it slides. MS-PS2-2 asks for exactly this, one variable changed and the motion observed.
- Robot players (8 minutes) Test, Robot players: predict how many of twenty plays a careful robot will win, then run them. The robots aim with a set wobble, so the rate belongs to the machine and the pile. Run twenty more and the number moves; that is the spread of a small sample, and 7.SP.C.6 in one screen.
- Talk it through (4 minutes) Ask the questions under How to tell whether it landed on the tools page. A student who can say what holds the prize up and why the ball is different has it.
A fair machine, a second period
The challenge: build a machine that is fun to play and fair, and share it. Build has three sub-tabs. Cabinet sets the box, the chute and its lip, the colors, the sign and the lamp and strips inside. Claw sets the prongs (how many, their length, the tip curl, the pad angle, how wide they open), the tips, the three grips for the grab, the lift and the carry, the speeds and the timer, and the look. Prizes fills the box at real masses, or with a 3D model of the student's own from the library.
Fair means the honest machine check: the maker has won a prize from it, and every prize kind in it can be lifted at the machine's own grips on the Grip Bench. It is modeled on the state rules that every prize on show must be winnable with the claw as set. A machine that fails names the kind that cannot be lifted, and the student changes the grips, the tips, the prongs or the pile until it passes. Fun is the robot win rate: a machine the robots win 90 percent of the time is a vending machine, and one they never win is a thief. Somewhere between 20 and 60 percent is a game.
Sharing puts the machine on a share page where classmates play it and leave comments. Ask students to play two machines that are not their own and say what the maker did to make it hard.
What students get wrong, and where the app answers it
| What students say | Where the app answers it |
|---|---|
| The claw holds the prize by squeezing it | The replay's arrows: the push is sideways, the friction is what points up. Turn the grip down on the bench and the push is still there when the friction runs out. |
| A stronger grip always holds better | The capsule ball on the bench: a hard squeeze pops a round prize out, and curled tips cradle it at almost no grip. Pad angle and tip curl in Build. |
| Claw machines are rigged, so skill does not matter | The carry grip in Build is the operator's setting, and the replay shows a prize held off its middle turning out of the claw. Both are true: the setting and the aim. |
| Heavier is harder, full stop | On the bench a heavy boxed toy slides where a lighter bear holds, but a heavy bear with curled tips under its arms holds where a light smooth ball slips. Mass, friction and shape all count. |
| Twenty plays tell you the odds | Run the robots twice. The rate moves by several plays out of twenty; the sheet asks why. |
| The claw dropped it because the machine decided to | The sentence under the replay names a physical reason every time, written from what the engine recorded, and the arrows show it. |
What the app simplifies, on purpose
- The prizes are stiff. Real plush squashes, the prongs sink in, and that helps a grab. How it works in the app says so.
- Friction is one typical number for each surface, and a pair of surfaces averages the two. Real friction belongs to a pair and is measured; the app's numbers sit in textbook ranges (rubber on cardboard about 0.65, bare metal on hard plastic about 0.3).
- On soft materials friction is not one fixed number the way it is here; it changes with how hard the surfaces press.
- The friction the replay draws is what the contact could have made at that push. Whether the prize held is what the physics decided, and the two agree in the plays we checked.
- The lift's acceleration adds a little to the load, about five percent of the prize's weight; the bench keys say so.
- A pile is luck. The same machine filled from a different seed wins anywhere from one play in ten to four in five for the robots. The starter's pile is chosen to sit near two in five.
Kid, middle school and high school
| Kid | Middle school | High school | |
|---|---|---|---|
| The grips | Weak, medium, strong, very strong | Newtons on every slider | Newtons, and the friction line from weight and the coefficient |
| The replay | One plain sentence | The reason and what to do about it | The same, with the torque clause: the farther from the middle, the bigger the turning push |
| The bench | Held or slipped | The push and the friction the tips made | The same, with the line a straight squeeze would need |
| Puzzles | Five, in plain words | The same with the numbers | The same, with the formulas in the keys |
The sheet, the puzzles and the setup
The Sheet in the app's header is the student's write-up: a picture of their machine, its claw and grips, their bench rows with the guess beside the result, their robot runs with the prediction, which puzzles they did and their written answers, and whether the machine passed the honest check. It prints, and with a free account it comes back to you as a saved project.
Setup is for you: the level and whether it is locked, which tabs students can open (Play is always on), the sheet's questions and up to ten of your own, a target win rate for the robots, and a printed answer key. Save the project and hand it out as the starter; the machine and the settings travel with it.
Code, beside the Sheet, is the programming rung for middle and high school. Students write chooseDrop, a short JavaScript function that reads the pile (each prize's name, mass, position, how high its top is, how many prizes sit on it) and returns where to drop, and the robot players play it on their machine beside a careful robot aiming the usual way. The second function, gripPlan, sets the squeeze over time for the grab, the lift and the carry; the honest check still judges the weakest carry. Both run in a sandbox that cannot reach the page or the network, and the sheet gains one question about the rule they wrote.
Standards
- MS-PS2-2: plan an investigation to provide evidence that the change in an object's motion depends on the sum of the forces on the object and the mass of the object. The Grip Bench, one variable at a time.
- MS-ETS1-1 to 4: define a design problem with constraints, evaluate competing solutions, analyze data from tests to improve a design, and develop a model to generate data for iterative testing. The fair machine challenge, with the robot players as the model.
- 7.SP.C.6 and 7.SP.C.7: approximate the probability of an event by collecting data on repeated trials, and compare a model's probability with observed frequencies. The robot win rate against a prediction.
- HS-PS2-1, as an extension: the friction line, the torque from a grab off the center of mass, and the formulas in the puzzle keys.
More help
Is something on this page wrong or not clear? Write to us.
Page last reviewed November 2026.
Google, Google Classroom and Chromebook are trademarks of Google LLC.