Coin flip probability, explained

Few ideas in probability are as approachable as the humble coin flip. You already own the equipment, the rules take one sentence to explain, and yet a single coin can teach you almost everything about how chance really behaves. This guide walks through it step by step, with worked examples you can check for yourself.

A single fair flip is the whole foundation

A fair coin has two sides, heads and tails, and nothing about its shape favours one over the other. That gives each side an equal chance: heads is 1/2 and tails is 1/2, which we usually write as 50%. Add the two together and you get 100%, because on any given toss one of the two outcomes has to happen. If you want to watch this settle in real time, our coin flip tool keeps a running tally so you can see the split drift toward half and half over many tries.

A coin has no memory

Here is the single most important rule, and the one people forget most often: each flip is independent. The coin has no memory. It does not know what it did a moment ago, it cannot store a result, and it feels no pressure to balance things out. Whether the last toss was heads or tails, the next toss is a fresh 50/50. This idea of independence is the key that unlocks everything else on this page.

The odds of a streak

Independence lets us calculate the odds of a whole sequence. To find the probability of one specific run of results, you multiply the probability of each flip together. Because every flip is 1/2, a run of n heads in a row has probability (1/2)ⁿ. Each extra flip you demand cuts the chance in half.

Notice the pattern: two heads is 1/4, and every step down doubles the denominator. Five heads in a row happens only about 3% of the time, so it is uncommon but far from impossible. Ten in a row is genuinely rare, roughly one chance in a thousand. One important detail: these are the odds of getting that streak on a specific set of flips you name in advance. If you keep flipping all afternoon, short streaks will still show up regularly, because you are giving them many chances to appear.

The gambler's fallacy

Suppose you have just flipped five heads in a row. What are the odds the sixth flip is tails? Many people feel that tails is now "due", as if the coin owes them a correction. It does not. The sixth flip is still exactly 50/50. The (1/2)ⁿ formula only looked steep before you started; once five heads have already landed, they are history, and the next flip cannot see them. Believing that a streak makes the opposite result more likely is called the gambler's fallacy, and it has emptied a lot of wallets. Results even out in the long run not because the coin corrects itself, but because a few early flips get swamped by thousands of later ones.

Flipping several coins at once

So far we have flipped one coin at a time. A different and equally useful question is this: if I flip several coins at once, how many come up heads? Counting heads across a batch of coins gives you the binomial distribution, and it behaves in a satisfying, predictable way.

Two coins

Flip two coins and list every equally likely outcome: HH, HT, TH, and TT. There are four, each with a 1/4 chance. Now count the heads. Only HH gives two heads, so two heads is 25%. Only TT gives zero heads, another 25%. But one-of-each happens two ways, HT and TH, so exactly one head is the most common result at 50%. See it build for yourself with the flip 2 coins tool.

Three coins

Add a third coin and the number of outcomes doubles to eight. Sorted by how many heads appear, they fall into the pattern 1, 3, 3, 1. That means three heads happens 1 way out of 8 (12.5%), and so does zero heads. Two heads happens 3 ways out of 8, which is 37.5%, and exactly one head is the same 3/8. The middle results are always the fat part of the distribution, and the all-heads or all-tails extremes are always the thin ends. The flip 3 coins tool lets you stack up trials and watch that 1-3-3-1 shape emerge.

One hundred coins and the law of large numbers

Scale this up to one hundred coins and the pattern becomes beautiful. On average you expect 50 heads, and the typical spread around that average, the standard deviation, works out to exactly 5. As a rule of thumb, results land within one standard deviation of the mean about two-thirds of the time, so roughly 73% of the time you will see between 45 and 55 heads.

Now here is the twist that surprises almost everyone. Fifty heads is the single most likely outcome, yet you will actually hit exactly 50 only about 8% of the time. "Most likely" is not the same as "likely". No single count is common, because the probability is shared out across dozens of possibilities, from roughly 35 up to 65 heads. Fifty simply gets the biggest single slice of a thinly divided pie. Run a batch on the flip 100 coins tool and you will rarely land on 50 on the nose, but you will almost always land nearby.

This is the law of large numbers, and it says something more subtle than "everything evens out". As you flip more and more coins, the proportion of heads closes in tightly on 50%. The raw count, though, does the opposite: the more coins you flip, the further the exact number of heads can drift from the halfway mark in absolute terms. With 100 coins, being off by 5 or 10 heads is ordinary; with a million coins you might be off by hundreds, yet still sit at 50-point-something percent. The percentage tightens while the headcount spreads.

The five things to remember

• One fair flip is 50/50, every single time.
• Flips are independent, so a specific run of n heads has probability (1/2)ⁿ.
• A streak never makes the other side "due": that is the gambler's fallacy.
• Counting heads across many coins follows the binomial distribution, bunched in the middle.
• Over many flips the proportion converges on 50% even as the exact count keeps wandering.

Understand those five points and you understand more about probability than most people ever will, all from a coin you probably have in your pocket.

More guides

• Is a Coin Flip Really 50/50? The Surprising Science of the Toss
• How to Make Random Decisions Without Overthinking
• How to call on students randomly without singling anyone out
• How to split a class into fair groups

Or jump straight to the coin flip tool and spinner wheel, or browse all SpinPicker tools.