Running Experiments
📋 What it is
Careful experiments with many trials give reliable experimental probabilities.
🗣️ Coach says
A good probability experiment is designed: decide the trial, repeat it MANY times, record every result, then compute the fraction. One or two trials prove nothing — the law of large numbers needs lots.
🧠 Memory hook
Few trials = unreliable. MANY trials → the experimental fraction settles near the true chance.
😂 Giggle
Why did the bar graph break up with the pie chart?
It said, "You're too busy going in circles!"
😲 Whoa!
To trust a new medicine, scientists run experiments on thousands of people — because small samples can mislead by pure chance.
✅ Quick check: Why is flipping a coin 100 times more trustworthy than flipping it twice?
Say your answer out loud first — then reveal.
More trials average out the randomness, so the result lands closer to the true 1/2.
The law of large numbers pulls experimental results toward theory as trials grow.
🧪 Try it! (2 minutes)
Roll a die 10 times, then 40 times. Does the fraction of 6s get closer to 1/6 with more rolls?