When Intuition Fails
Your gut is wrong. These 20 interactive simulations prove it. From the Monty Hall problem—where switching doors doubles your odds—to Simpson’s paradox, where a treatment can win in every subgroup yet lose overall. Watch a thousand needles estimate π, see why your friends really are more popular than you, and discover that “random” isn’t even well-defined. Every paradox here is a place where mathematics and human intuition violently disagree. Run the simulations. Trust the numbers. Question everything else.
The greatest hits of probability—problems that have fooled mathematicians, philosophers, and game show contestants for decades.
Three doors, one car, two goats. The host reveals a goat. Should you switch? Run 10,000 trials and watch switching win 2/3 of the time.
Only 23 people needed for a 50% chance of shared birthdays. Add people to a room and watch the probability soar past your intuition.
Treatment A beats B in every subgroup, yet B beats A overall. Explore how aggregating data can completely reverse a trend.
One envelope has twice the other’s money. The math “proves” you should always switch—creating an infinite loop of regret.
“I have two children, at least one is a boy.” P(both boys) = 1/3, not 1/2. See how phrasing reshapes the probability space.
Probability hides in geometry and social networks—and what it reveals is deeply counterintuitive.
Drop needles on parallel lines. Count crossings. Divide. You just calculated π by throwing sticks on the floor.
Three ways to pick a “random” chord in a circle give three different answers: 1/3, 1/2, 1/4. “Random” is not well-defined.
Your friends have more friends than you do—on average, always, mathematically guaranteed. Explore why in an interactive network.
Average class has 35 students, but the average student is in a class of 90. Same data, wildly different answers depending on who you ask.
The next elevator always goes the wrong way. It’s not Murphy’s Law—it’s probability. Watch elevators prove your frustration is mathematical.
Time warps probability. Buses bunch, coins don’t remember, and the optimal strategy for life’s biggest decisions involves the number e.
Buses come every 10 minutes on average. Your average wait? Also 10 minutes, not 5. Longer gaps catch more random arrivals.
After 10 heads in a row, tails is NOT more likely. Flip thousands of coins and watch streaks emerge from pure randomness.
Random events cluster in time. Buses bunch, disasters come in waves, and “bad luck comes in threes” is pure probability.
Interview N candidates, can’t go back. Reject the first 37%, then pick the next one that’s best so far. The optimal stopping rule.
A game with infinite expected value that no rational person would pay $20 to play. Where expected value breaks down.
Numbers lie—or rather, our interpretation of them does. These paradoxes reveal how data can mislead even experts.
30% of real-world numbers start with 1, only 4.6% start with 9. Explore datasets and see this eerie pattern emerge everywhere.
Capture 5 tanks with serial numbers. Estimate total production. The Allies did this in WWII—and beat the intelligence agencies’ estimates.
Move the weakest from Group A to Group B. Both averages go UP. A statistical magic trick used in cancer staging.
In hospitals, two independent diseases appear negatively correlated. Sampling bias creates phantom relationships from nothing.
Flip a coin. Heads: wake once. Tails: wake twice with memory erased. P(heads) when awakened? Philosophers still disagree: 1/2 or 1/3.