A Gambler, a Plague of Coincidences, and Four Probabilities Your Gut Gets Wrong

In the 1650s a French gambler known as the Chevalier de Méré kept a pair of bets he was sure about. Bet one: roll a single die four times and at least one six appears. He won this more often than he lost. Bet two, which he believed was equivalent: roll two dice 24 times and at least one double six appears. He was losing money on it. The puzzle, sent to Blaise Pascal, led to a famous exchange of letters between Pascal and Pierre de Fermat in 1654, and historians often mark it as an early step toward modern probability theory.

We can see the answer in two lines. For bet one, the chance of no six in four rolls is (5/6)^4, so the chance of at least one is 1 - 0.482 = 51.8%. For bet two, the chance of no double six in 24 rolls is (35/36)^24, so the chance of at least one is 49.1%. A bet that looks like a rescaled version of the first one sits just below even odds. The Chevalier's intuition said "one in six, times four" and "one in thirty-six, times twenty-four" were equivalent, and they are not, because the right way to combine repeated chances is through the complement, not by adding.

That one habit, work out the chance of nothing happening and subtract it from one, solves a large share of everyday probability questions. Four more cases show where the gut goes wrong.

Coincidence 1: the birthday match

How many people must be in a room before there is a better than even chance two of them share a birthday? Most guesses are above 100. The answer is 23. With 23 people there are 253 pairs, and each pair has a small chance of matching. Multiplying the no-match chances for everyone in turn gives about 49.3%, so the chance of at least one match is 50.7%.

• 10 people: 11.7%
• 20 people: 41.1%
• 30 people: 70.6%
• 40 people: 89.1%
• 57 people: 99.0%

The mistake is picturing a match with you. The chance that someone else shares your particular birthday needs about 253 people to reach 50%. A match between any two people among 23 is a different, much easier event. This gap explains why coincidences feel eerie: of all the possible pairings in your life, some will match, and you notice the ones that do.

Coincidence 2: the test that is 90% accurate

A screening test catches 90% of people who have a condition. It also gives a false alarm for 5% of people who do not. The condition affects 1% of the population. You test positive. What is the chance you have it? Most people say around 90%. Here is the count for 10,000 people:

• 100 have the condition, and 90 of them test positive
• 9,900 do not have it, and 5% of them, 495 people, also test positive
• Total positives: 585. True positives: 90.

Your chance is 90 / 585, or 15.4%. The test is good and the positive is still probably a false alarm, because the condition is rare. Change the prevalence and the answer shifts sharply: at 0.1% prevalence a positive means 1.8%, at 10% prevalence it is 66.7%, at 30% it is 88.5%. The same test means very different things in a high-risk group and a general population, which is why doctors ask about symptoms and history before ordering one. This is a pure illustration with made-up rates. Real tests have their own published accuracy figures, and your clinician is the right person to interpret them.

Coincidence 3: streaks look suspicious, but they are normal

Flip a fair coin 100 times. How likely is a run of at least five heads somewhere in the sequence? About 81%. At least six? About 55%. At least seven? About 32%. These come from a short recurrence calculation, and they surprise people because a human asked to write a "random" sequence by hand tends to avoid long runs. Real randomness clumps.

The opposite error is the gambler's fallacy: after five heads in a row, tails is "due". It is not. The coin has no memory, and the next flip is 50/50 as it always was. The chance of five heads in a row from a standing start is 1 in 32, but once four heads have already landed, the chance the next one matches is simply one half.

Coincidence 4: the lottery ticket that feels almost winnable

In a lottery where you choose 6 numbers from 49, the number of possible tickets is 49 x 48 x 47 x 46 x 45 x 44 divided by 720, which is 13,983,816. One ticket has a 1 in 14 million chance. Buying 100 tickets raises the chance to about 7 in a million, which is still not something to plan around. The permutation and combination calculator does this count in one step, and it helps you see why combinations (order does not matter) give a much smaller number than permutations (order matters) for the same selection.

A short toolkit

• For 'at least once' questions, compute the chance of never and subtract from 1.
• For 'and' questions with independent events, multiply. For 'or' questions with mutually exclusive events, add.
• For test results, imagine 10,000 people and count them. It beats any formula for building intuition.
• For 'someone, somewhere' versus 'me specifically', count pairs, not people.
• Treat independent events as having no memory, whatever the last result was.

If you want to check a result of your own, the probability calculator handles single events, repeated trials and combined events. Work a few of the examples above through it, change one number at a time, and watch how quickly intuition and arithmetic part ways.