🪙 Coin Flip: Flip a Coin Online for Heads or Tails
By Shihab Mia · Updated 2026-07-01
Press Flip to toss the coin.
Total flips this session: 0
A coin flip gives each of two outcomes, heads or tails, an equal 50 percent chance, so it is the fairest way to settle a two-way choice. Press Flip below for an instant random result, or set how many coins to toss at once to see the split. A running tally tracks heads and tails for your whole session, with live percentages, so you can watch the numbers even out the more you flip.
What is the Coin Flip?
A coin flip is the classic method for making a fair, unbiased, two-way decision. A coin has two faces, heads and tails, and a fair toss gives each an equal 50 percent (1 in 2) chance. People reach for it whenever they want an impartial choice with no favouritism: who goes first, which of two plans to pick, who pays the bill, or simply yes or no when they cannot decide. Because neither party controls the result, both sides accept it as fair, which is exactly why sports referees and courts have used coin tosses for over a century.
This simulator generates each result with the browser's cryptographic random number generator (crypto.getRandomValues), not the weaker Math.random function. That source is engineered to be unpredictable and evenly distributed, so neither heads nor tails is favoured. To keep the odds exactly even it reads a random byte and discards the one value that would otherwise nudge one outcome slightly ahead, a standard technique called rejection sampling. The result is a toss that is as fair as mathematics allows, and in one important sense fairer than a real coin.
That is not a figure of speech. In 2007 Stanford statistician Persi Diaconis and colleagues showed with a physical model that a real tossed coin is slightly biased toward landing the same way up as it started, and predicted about a 51 percent same-side chance. A 2023 study collected 350,757 real human flips and confirmed the effect, measuring a same-side rate close to 50.8 percent. The bias comes from a tiny wobble in the coin's spin, and it vanishes only if the caller cannot see which side started up. A software simulator has no starting face and no wobble, so it sidesteps this bias entirely.
The running tally illustrates a core idea in probability: the law of large numbers. In a short run you might see five heads in a row and the percentage can sit far from 50 percent. That is normal, not a fault. As the number of flips grows, the proportion of heads drifts steadily toward 50 percent, even though the raw gap between the counts can keep widening. A single flip is unpredictable, but the long-run average is not, and that is precisely the behaviour the tally makes visible.
A related trap is the gambler's fallacy: the belief that a run of heads makes tails somehow due. It does not. Every flip is independent, so the odds reset to 50 percent on each toss no matter what came before. The coin has no memory. Understanding this is the difference between reading randomness correctly and being fooled by it, which is why classrooms use coin flips to teach independence and probability in the first place.
For decisions with more than two options, a single coin is not enough, but coins still work. To pick fairly among three people, run three simultaneous flips and let the odd one out win, or assign the three outcomes of two flips (heads-heads, heads-tails, tails-heads) and re-flip on tails-tails. Some people prefer best of three flips to feel more decisive, though statistically a single flip is already perfectly fair and best of three changes nothing about the underlying 50/50 odds.
When to use it
- Making a quick, fair decision between two options when you genuinely cannot choose.
- Deciding who goes first in a game, who pays, or who picks the restaurant or movie.
- Settling a dispute impartially so both parties accept the outcome without argument.
- Running a classroom or probability demonstration by tossing many coins and watching the percentages settle near 50 percent.
- Teaching the law of large numbers and the gambler's fallacy with a live, honest random source.
- Replacing a physical coin when you have none to hand, or want to avoid the slight same-side bias of a real toss.
How to use the Coin Flip
- Leave the count at 1 for a single coin, or enter a larger number to toss many coins at once.
- Press Flip to generate a random heads or tails result.
- Read the result and the updated running tally of heads and tails counts and percentages.
- Keep flipping to build up the session totals and watch the split approach 50/50.
- Press Reset tally to clear the counts and start a fresh session whenever you like.
Formula & method
Worked examples
You flip a single coin 4 times in a session and get heads, tails, heads, heads.
- Heads count = 3, tails count = 1
- Total flips n = 3 + 1 = 4
- Heads percent = 3 / 4 x 100 = 75.0%
- Tails percent = 1 / 4 x 100 = 25.0%
Result: Tally shows 3 heads (75.0%) and 1 tail (25.0%) after 4 flips, a wide split that is normal in a short run.
You set the count to 100 and press Flip once, landing 53 heads.
- One press tosses 100 coins, so total flips n = 100
- Heads = 53, tails = 100 - 53 = 47
- Heads percent = 53 / 100 x 100 = 53.0%
- Tails percent = 47 / 100 x 100 = 47.0%
Result: A 53 to 47 split, close to the expected 50 to 50 over 100 tosses.
You want the probability of getting three heads in a row.
- Each flip is independent with probability 0.5 for heads
- Multiply the odds: 0.5 x 0.5 x 0.5 = 0.125
- Convert to a percentage: 0.125 x 100 = 12.5%
Result: Three heads in a row happens about 1 time in 8, or 12.5 percent of the time.
How the heads percentage settles as the number of flips grows (typical fair-coin behaviour)
| Number of flips | Expected heads | Typical range of heads % |
|---|---|---|
| 10 | 5 | roughly 20% to 80% |
| 100 | 50 | roughly 40% to 60% |
| 1,000 | 500 | roughly 47% to 53% |
| 10,000 | 5,000 | roughly 49% to 51% |
Probability of common short outcomes with a fair coin
| Outcome | Probability | As a percentage |
|---|---|---|
| One specific result (heads) | 1 in 2 | 50% |
| Two heads in a row | 1 in 4 | 25% |
| Three heads in a row | 1 in 8 | 12.5% |
| Five heads in a row | 1 in 32 | about 3.13% |
| Ten heads in a row | 1 in 1,024 | about 0.098% |
Fair ways to decide among more than two people with coins
| Number of people | Method | How it works |
|---|---|---|
| 2 | One flip | Assign heads to one person, tails to the other. |
| 3 | Three simultaneous flips | The person whose coin lands opposite to the other two wins; re-flip if all three match. |
| 4 | Two flips (2 rounds) | Flip to split into pairs, then flip within the winning pair. |
Common mistakes to avoid
- Thinking a streak makes the next flip "due" to change. After five heads in a row, tails is not more likely on the next flip. Each toss is independent and stays at 50 percent. Expecting the coin to balance itself out flip by flip is the gambler's fallacy.
- Expecting an exact 50/50 split in a short run. Over a handful of flips the percentages will often sit far from 50 percent. That is ordinary randomness, not bias. The proportion only closes in on 50 percent over many flips.
- Assuming a real coin toss is perfectly fair. Research shows a physical toss is slightly biased toward landing the same side it started, about 51 percent. It only becomes fair when the caller cannot see the starting face. A simulator has no starting side, so it avoids this.
- Reading the count box as total flips. The number you enter is how many coins one press tosses, not your session total. Pressing Flip several times keeps adding to the running tally below.
- Believing best of three is fairer than one flip. A single fair flip is already perfectly 50/50. Best of three feels more decisive but does not change the odds for either side; it just adds flips.
Glossary
- Heads
- One of the two faces of a coin, traditionally the side showing a portrait or main design.
- Tails
- The other face of a coin, the side opposite heads.
- Fair coin
- A coin or simulation where heads and tails each have an exactly equal 50 percent chance.
- Probability
- A number from 0 to 1 (or 0 to 100 percent) describing how likely an outcome is. A fair coin flip has a probability of 0.5 for each side.
- Law of large numbers
- The principle that the average of many independent trials moves closer to the expected value as the number of trials grows.
- Independent events
- Outcomes that do not affect one another, so a previous flip never changes the odds of the next one.
- Gambler's fallacy
- The mistaken belief that past results make a particular future outcome more or less likely in independent trials.
- Same-side bias
- The slight tendency of a real tossed coin to land the same way up as it started, measured at about 51 percent in physical experiments.
Frequently asked questions
Is this coin flip truly random and fair?
Yes. Each flip uses the browser's cryptographic random number generator (crypto.getRandomValues), which is designed to be unpredictable and evenly distributed. The tool also discards a single byte value through rejection sampling so heads and tails stay at exactly 50 percent each.
What are the odds of heads or tails?
On a fair coin each side has a 50 percent (1 in 2) chance on every flip. The two sides are equally likely, and that does not change no matter what came before.
Is a real coin flip actually 50/50?
Not quite. Physics research by Persi Diaconis predicted, and a 2023 study of 350,757 flips confirmed, that a real toss lands the same side it started about 51 percent of the time. The bias disappears only if the caller cannot see the starting face. This simulator has no starting side, so it is a true 50/50.
If I get five heads in a row, is tails more likely next?
No. Each flip is independent, so the chance is still 50 percent for each side regardless of the streak before it. Believing past results change the next flip is the gambler's fallacy.
Why is my tally not exactly 50/50?
Short runs are naturally uneven, so the percentages can sit well away from 50 percent. As you flip more times the proportion of heads drifts toward 50 percent, even though the raw count gap can still grow. This is the law of large numbers.
How do I use a coin to decide between three people?
Have all three flip at once and the person whose coin lands opposite to the other two wins; if all three match, flip again. Alternatively, use two flips and assign heads-heads, heads-tails, and tails-heads to the three people, re-flipping only on tails-tails.
What is the probability of getting three heads in a row?
Because each flip is independent, you multiply the odds: 0.5 x 0.5 x 0.5 = 0.125, or 12.5 percent. That is about 1 chance in 8. The same logic gives 1 in 1,024 for ten heads in a row.
Can I flip more than one coin at a time?
Yes. Set the count box to the number of coins you want one press to toss (up to 1,000). The result shows how many landed heads and tails, and every coin is added to the running session tally.
Is best of three flips fairer than one flip?
No. A single fair flip is already exactly 50/50 for both sides. Best of three can feel more decisive, but it does not give either side an advantage; it only adds flips.
Does the running tally save after I close the page?
No. The tally is kept only for the current session and lives in your browser memory. Reloading or closing the page clears it, and you can also reset it any time with the Reset tally button.