If most buses run half-empty, what kind of bus does a random rider land on?
Did this make sense?
Your thumbs help us make it clearer.
Thanks! That helps us make it clearer.
After you watchIf most buses run half-empty, what kind of bus does a random rider land on?
The short answer
Picking a random rider weights buses by how many people they hold. In the tested fleet, eight small buses hold 24 riders and two busy buses hold 36, so a rider is more likely to be on one of the two busy buses even though most buses are small.
Try this next
- What if every bus carried the same number? Set the practice slider to even. Count the people per bus and explain why giving each rider a vote would make no difference.
- Will 25 random picks always match 36 chances in 60? Reset the fixed-fleet experiment and repeat several samples. Compare the counts with the exact census.
- Could most buses be small while most riders are on busy buses? Draw eight buses with 3 dots each and two buses with 18 dots each. Count each group before comparing it with the simulation.
Now you — bend it
- What if Push the unevenness slider higher and predict what happens to the gap between the rider-average and the bus-average.The gap equals the spread in how full buses are divided by the average fullness, so more lumpiness means a wider gap.
- What if Set every bus to carry the same number, then predict where the rider-average lands.With no spread, the two averages match exactly. The gap only opens when buses differ.
- What if Add one giant bus that holds most of the riders. Predict whether a random rider almost always lands on it.Picking by rider weights each bus by its own crowd, so the bus holding the most people gets picked the most.
Can you prove it?The average a rider feels is always at least the plain bus-average, and they only match when every bus is equally full. — Use the fresh test fleet: its plain mean is 60/10=6. Its rider-weighted mean is (8×3×3+2×18×18)/60=12. Random samples fluctuate around the latter; one sample is not a proof.
Design your own test:Before counting every rider, predict which group contains more people.
Explain it to a 6-year-old: Many people can fit on just a few busy buses, so counting people can tell a different story from counting buses.
The whole story
How it works
The practice slider redistributes 60 riders over 10 buses. The hidden test then uses a separate fixed fleet: eight buses with 3 riders each and two with 18. A random rider has 36 chances in 60 to be on a busier bus. The mean occupancy per bus is 6; weighting each bus by its riders gives an expected rider-experienced occupancy of 12. Small samples may differ from these expectations.
What people get wrong
People assume the crowding they personally feel must equal the plain average, so a packed bus means buses really are crowded. In fact a random rider over-samples the full buses, because that is where most people are, so their experience runs above the fleet average whenever buses vary in how full they are.
The catch
Counting the buses is honest: it correctly answers 'how is the whole fleet doing?' and the city is not lying when it says many buses are roomy. Counting the people is honest too: it correctly answers 'what does a typical rider feel?' and that answer really is 'crowded.' Neither number is wrong; trouble only comes from mixing up which question you are asking.
Questions kids ask
Is the city lying when it says the average bus is half-empty?
No. The fleet really does average out to half-empty when you give every bus one vote. But you ride as a person, not as a bus, and most people are crammed onto the few full buses, so the average a rider feels is much higher. Both numbers are true; they answer different questions.
Would this still happen if every bus carried the same number?
No. If every bus is equally full, the bus-average and the rider-average match exactly. The gap only opens up when buses vary, and it grows the more uneven they are.
Where else does this trick show up?
Everywhere people or things bunch together. Your class feels bigger than the school's average class size, popular restaurants always seem packed, and most people have fewer friends than their friends do. In each case you are sampling by the crowded thing, so your experience runs above the plain average.
Why does my bus also seem to come right after a long wait?
A related reason: buses tend to bunch up. A slightly late bus collects extra riders, which slows it down further, so gaps grow. You are more likely to arrive during a long gap than a short one, so your wait feels longer than the schedule promises.
Talk about it
- Guess first: if the city says most buses are half-empty but yours feels packed, who do you think is right?
- Why might counting buses and counting people give two different answers that are both true?
- Where else in your day do you think you only notice the crowded version of something?
For grown-ups
This is the inspection paradox, or size-biased sampling. Sampling by rider weights each bus by its own occupancy, so the rider-experienced mean equals the plain mean plus variance divided by mean. It is always at least the plain mean, with equality only when every bus carries the same number (zero variance). The same maths explains why a class feels bigger than the school's average class size, why most people have fewer friends than their friends do (the friendship paradox), and why waits feel long when buses bunch.
Keep going
What else makes you wonder?
- If the few packed buses carry most of the riders, what would it take to spread people out evenly?
- Does this same trick make your lunch line, your favorite playground, or a popular game always feel busier than it really is?
- If most people have fewer friends than their friends do, are you the crowded bus in someone else's count?
