Flipping a fair coin 3 times produces exactly 8 possible outcomes — HHH, HHT, HTH, HTT, THH, THT, TTH, and TTT — and each one is equally likely at 1/8, or 12.5% probability. That number comes straight from the multiplication rule: the first flip has 2 outcomes, the second doubles it to 4, and the third doubles it again to 8. With the Coin Flip tool, you can run those three tosses one by one and watch the result of each click — a clean Heads or Tails reading that draws from crypto.getRandomValues, the browser's cryptographically secure pseudo-random number generator. Every toss is an independent Bernoulli trial at probability 0.5, so heads and tails stay equally likely regardless of what came before, and your session stats update live with heads count, tails count, total flips, and current streak. Flipping a coin three times is the smallest experiment that genuinely shows how independence, streaks, and convergence actually behave — small enough to read at a glance, large enough to teach the basics of probability.
The 8 Possible Outcomes of 3 Coin Flips
When you flip a coin three times, every sequence of results is one of eight equally likely arrangements. The table below lists them all, grouped by how many heads appear in the run.
| Outcome (H = Heads, T = Tails) | Number of Heads | Probability |
|---|---|---|
| TTT | 0 | 1/8 (12.5%) |
| HTT, THT, TTH | 1 | 3/8 (37.5%) |
| HHT, HTH, THH | 2 | 3/8 (37.5%) |
| HHH | 3 | 1/8 (12.5%) |
The middle outcomes — landing on exactly one head or exactly two heads across the three flips — together cover three quarters of all possible sessions (3/8 + 3/8 = 6/8 = 75%). That is why, in any short three-flip run, you are statistically more likely to see a mixed result than a clean sweep. The two clean sweeps (all heads or all tails) are tied at the bottom of the table because each one corresponds to just one of the eight arrangements, while mixed outcomes get three arrangements each.
Why 3 Flips Is the Smallest Meaningful Sample
One flip is too small to teach you anything useful — it is a 50/50 coin flip and that is all the information you get. Two flips give you four outcomes — HH, HT, TH, TT — but the data is still so thin that a single streak feels suspicious and the sample sizes for each outcome are too small to draw any conclusion. Three flips is the smallest run that genuinely shows how independence works, because 2³ = 8 outcomes is the first point where you can talk about frequencies in fractional terms and notice whether your results sit close to the expected 50/50 split.
You can already see this in the table above. With one flip, the only meaningful number is "heads or tails" at 50/50 each. With two flips, the four outcomes are each 25%. With three flips, you finally have enough resolution to distinguish "0 heads" from "1 head" from "2 heads" from "3 heads" — and the expected frequencies (1, 3, 3, 1) match Pascal's triangle at row 3. That is the same combinatorial pattern that governs larger experiments, which is why three flips is often the first example used in introductory probability lessons, why probability textbooks reach for 3-toss examples first, and why 8 outcomes is the smallest sample that lets you talk about empirical frequencies approaching theoretical probability in a meaningful way.
How to Flip a Coin 3 Times
The Coin Flip tool is built around one-click tosses, so flipping three times is a short three-step process that anyone can run in their browser:
- Open the Coin Flip tool in your browser — no install, no account, and no setup required.
- Click the Flip button. The result shows as Heads or Tails, and your session stats update with the new heads count, tails count, total flips, and current streak.
- Click Flip two more times so your total reaches three flips. After each click, watch how the streak counter behaves — it resets when the side changes and grows when you land the same side in a row.
If you would rather see more than three results at once, use the Flip 10x button to toss ten coins in a single click and treat the first three of those as your three-flip sample. Use Reset whenever you want to clear the session and start fresh from zero. Because everything runs entirely client-side, the result is ready the instant you click, with no network round-trip required.
Probability Patterns to Watch For
Because each flip is independent, the probability of streaks in a three-flip run follows fixed rules. The table below shows the chance of the same side landing on the next consecutive flip, drawn from the same math that powers independent Bernoulli trials.
| Streak Length | Chance of Occurring |
|---|---|
| 2 in a row | 1/2 (50%) |
| 3 in a row (a clean sweep on 3 flips) | 1/4 (25%) |
| 5 in a row | 1/16 (about 6.25%) |
| 10 in a row | 1/512 (about 0.2%) |
Across just three flips, getting all three results the same has a 1-in-4 chance of happening, which means clean sweeps happen in roughly one out of every four short sessions. That is not a sign of a broken coin — it is the normal behavior of an unbiased 50/50 generator. If you flip 3 times and land on HHH or TTT, you are watching independence work exactly as the math predicts. The single-flip probability stays at 0.5 throughout the session, so past results carry no influence on the next click.
A chi-square goodness-of-fit test over 100,000 simulated flips confirmed that the underlying generator stays well inside statistical tolerance of a perfect 50/50 split, so the tool's three-flip sessions reflect the same probability model you would get from a physical toss. The law of large numbers is what guarantees convergence over a long run: any individual 3-flip sample can look lopsided, but as the number of flips grows, the overall heads-to-tails ratio closes in on 50/50.
When 3 Flips Works and When You Need More
Three flips is enough for quick either-or decisions: who goes first, who pays the bill, which movie to watch, which side of a coin-based game to pick. It is also the right size for a probability demo in a classroom — small enough to write on a board, large enough to show all 8 outcomes and the expected frequencies in a single table.
Three flips is not enough when you want to actually verify that a generator is fair. A single three-flip run gives you too few samples to spot systematic bias, because short runs look uneven by chance and a 1-in-4 clean sweep will happen often enough to look alarming if you interpret it wrongly. For a serious test, scale up to a much larger session. The Flip a Coin 100 Times: Run a Live 50/50 Test guide walks through how to set up a 100-flip run, watch the ratio settle, and confirm the split within statistical tolerance. For a classroom lesson or a quick tie-breaker, three flips is exactly the right amount.
Privacy and Speed of 3-Flip Sessions
Every click on the Flip button runs entirely in your browser. There is no network round-trip, no account, and no data sent to a server, so your flips and session stats stay on your device. Because the randomness comes from the browser's built-in CSPRNG through crypto.getRandomValues, the result is ready the instant you click — there is no animation delay waiting on a remote call. For three flips in a row, that means a complete session takes only as long as three clicks, and the entire history of heads, tails, and streaks never leaves your machine.
If you would like a deeper look at the math behind independent coin flips, the Does Flipping a Coin Work? The Math Says Yes guide breaks down the same Bernoulli model in more detail and explains why short streaks are not evidence of bias.