Public · Sign in
MT
← Back to topic
LESSON PLAN

Likelihood of Simulated Events

A
Apothem Team
Grade 3 · Data & Probability
LESSON AT A GLANCE
Warm-up
5 min
Explore
15 min
Formalize
10 min
Practice
12 min
Exit ticket
3 min

Warm-up

If I flip this coin 10 times, how many times do you predict heads? (5.) Let us test. Flip 10 times, record. Class results: did everyone get exactly 5? (Likely not.) Why not? (Randomness.) If we all combined our results (30 students x 10 flips = 300 flips), would we be closer to 50%? (Yes.) Why?

Explore

Probability experiments stations: (1) Coin: predict 10 flips, record actual, calculate total heads out of 30 flips across the group. (2) Die: roll 12 times, which face appeared most? (3) Spinner: spin 20 times, compare to predicted. Pool class results at each station: do larger samples give better predictions?

Formalize

Compare predicted vs. actual: display all class results for the coin flip. Create a bar graph of heads counts across all groups (how many groups got 4 heads? 5 heads? 6 heads?). What is the most common result? (5, likely.) What is the range? This is a distribution, and Grade 3 students can describe it qualitatively.

Likelihood of Simulated Events

Snowsnake game connection: in a skill-and-chance game, what determines the outcome? (Both player skill and random variation.) Can we predict who will win before the game? Not with certainty, but we can say which outcome is more likely based on skill levels. This is applied probabilistic reasoning.

Practice

Students conduct a 20-flip coin experiment, record results, compare to prediction, and pool results with the class. Write 2 sentences about the difference between theoretical and experimental probability. Exit ticket: in a bag with 2 red and 8 blue marbles, which colour are you more likely to draw?

Exit ticket

Students conduct a 20-flip coin experiment, record results, compare to prediction, and pool results with the class. Write 2 sentences about the difference between theoretical and experimental probability. Exit ticket: in a bag with 2 red and 8 blue marbles, which colour are you more likely to draw?

TIP  The discrepancy between predicted and experimental results is the lesson, not a failure of the experiment. Explicitly celebrate: the results are not all 5 out of 10. That is what chance means.
WORKED EXAMPLES
Example 1 — Predict, then simulate: the three-colour spinner

Step 1: Examine the spinner: half red, quarter blue, quarter yellow. Elicit ranked predictions with reasons: red should win most spins (owns half the space); blue and yellow should roughly tie for the rest.

Step 2: Decide sample size — and let students feel why more is better: "Would 4 spins convince you? 40?" Settle on 40 as a class.

Step 3: Simulate: spin 40 times, tally live. Sample outcome: red 22, blue 9, yellow 9.

Step 4: Compare results to predictions: red ≈ half of 40 ✓; blue ≈ yellow ✓. Not EXACTLY 20-10-10 — and that's the teachable wobble: chance results hover NEAR the space-based prediction without hitting it on the nose.

Step 5: Pool all groups' tallies (suddenly 200+ spins): watch the proportions tighten toward half-quarter-quarter. Bigger samples steady the picture — the deepest idea in the room, earned with a spinner.

Example 2 — Design the dice game, then discover it's rigged

The game: roll two dice, add. Player A scores on a sum of 6, 7, or 8; Player B scores on 2, 3, 11, or 12. B has MORE winning numbers — is B luckier?

Step 1: Vote first (most pick B — four numbers beats three).

Step 2: Play 30 rounds in pairs and tally. The table fills with A wins.

Step 3: Hunt the reason — list the ways to roll each sum: 7 can happen as 1+6, 2+5, 3+4, 4+3, 5+2, 6+1 (six ways!); 6 and 8 have five ways each. But 2 is only 1+1, and 12 only 6+6 — one way each. A's three numbers own 16 ways; B's four numbers own just 6.

Step 4: The moral, stated for the ages: what matters is not how many WINNING LABELS you hold but how many PATHS lead to them. Middle sums are highways; edge sums are goat trails.

Step 5: Redesign the game to be fair — students propose and defend new number splits. (Fairness ≈ equal path-counts, about 18 each.)

Example 3 — 100 flips came up 58 heads: is the coin broken?

The claim: a group flips a coin 100 times, gets 58 heads, and concludes their coin is "a heads coin."

Step 1: Take the claim seriously — what SHOULD 100 fair flips give? "About 50" — but press on the word ABOUT: exactly 50, or near 50?

Step 2: Gather evidence from the room: every group flips 100 times. The board fills: 47, 53, 58, 44, 51, 55… Nobody hit exactly 50; everyone landed in the neighbourhood.

Step 3: Place 58 in that landscape: it's on the high side of normal wobble, not outside it. A fair coin wanders this much routinely.

Step 4: Ask what evidence WOULD justify suspicion: 90 heads? 85? (Intuition: way outside everyone's wobble zone.) There's a real threshold idea here — Grade 3 meets, informally, the logic of statistical significance: don't call a coin crooked for doing what fair coins ordinarily do.

Exit sentence for journals: "Fair doesn't mean exactly even — it means close to even, most of the time."

MATERIALS
Fair coins
Standard dice
Spinners (equal and unequal sections)
Recording sheets for experiments
Snowsnake game or alternative traditional game (if available)
WATCH FOR
!Students may think that if a coin shows tails 3 times in a row, heads is due. Address the gambler's fallacy directly: each flip is independent.
!Students may think experimental results must exactly match theoretical predictions. They should be close (especially with many trials) but will always vary.