Flip a Coin — Virtual Coin Toss
Quick Answer
Flipping a fair coin gives heads or tails with equal 50 percent probability; this virtual coin uses crypto.getRandomValues for true randomness, matching the statistical fairness of a physical coin toss described by Persi Diaconis's work on coin dynamics.
Also searched as: heads or tails, virtual coin toss, coin flip online, 50/50 decision maker
Tap Flip
Heads
0
Tails
0
Total
0
How a Virtual Coin Flip Works
A virtual coin flip is a computer simulation of tossing a fair coin, designed to produce heads or tails with equal 50 percent probability on every press. This page generates each outcome with crypto.getRandomValues, the browser's built-in cryptographically secure random number generator defined by the W3C Web Crypto API specification. It reads entropy from the operating system's random pool (for example /dev/urandom on Linux or BCryptGenRandom on Windows), which is continually seeded from hardware noise such as disk timing, network jitter, and dedicated RNG chips on modern CPUs. Because of this, the generator passes the NIST SP 800-22 statistical randomness tests and the sequence of flips is indistinguishable from an ideal coin. For related random tools see our RPG dice roller, random number generator, and probability calculator.
The Math of Coin Flips
A single fair coin flip is a Bernoulli trial with probability p = 0.5 of heads. For n independent flips, the number of heads follows a binomial distribution B(n, 0.5) with mean n/2 and standard deviation sqrt(n)/2. For 100 flips, you expect 50 heads plus or minus 5, meaning results between 45 and 55 are inside one standard deviation and results outside 40 to 60 are rare (about 5 percent probability). The probability of exactly k heads in n flips is C(n, k) times 0.5 to the n. A worked example: flipping 10 coins has 2^10 = 1024 equally likely outcomes; the probability of exactly 7 heads is C(10, 7) / 1024 = 120/1024 = 11.7 percent, and the probability of at least 7 heads is (120+45+10+1)/1024 = 17.2 percent. These formulas are the foundation of hypothesis testing and A/B testing in modern statistics.
Key Terms You Should Know
Fair coin: one where each side has an equal 50 percent chance of landing face up. Bernoulli trial: a random experiment with exactly two outcomes (success or failure) at fixed probability. Independent events: flips where the result of one does not affect any other; all coin flips are independent. Expected value: the average long-run outcome; for n flips it is n/2 heads. Law of large numbers: the principle that the observed ratio of heads approaches 50 percent as the number of flips grows. Gambler's fallacy: the mistaken belief that a long run of one outcome makes the other outcome more likely on the next flip; it does not. Streak: a consecutive run of the same outcome; in 100 flips, the longest streak averages about 7. PRNG vs. CSPRNG: a pseudo-random number generator is fast but predictable from its seed; a cryptographically secure PRNG (like crypto.getRandomValues) is unpredictable even if you observe prior outputs.
Coin Flip Statistics — Reference Data
The table below shows binomial probabilities for common coin-flip questions that appear in statistics textbooks, probability courses, and gambling strategy guides. The values come from the binomial distribution formula applied to a fair coin with p = 0.5. Notice how the probability of deviating from the expected 50/50 split shrinks rapidly as the number of flips increases, a direct illustration of the central limit theorem. These numbers are used routinely in A/B testing to determine whether an observed difference between two variants is statistically significant or just random variation.
| Scenario | Probability | Odds |
|---|---|---|
| Single flip heads | 50.0% | 1 in 2 |
| 2 heads in a row | 25.0% | 1 in 4 |
| 5 heads in a row | 3.125% | 1 in 32 |
| 10 heads in a row | 0.0977% | 1 in 1,024 |
| 20 heads in a row | 9.54e-5 % | 1 in 1,048,576 |
| Exactly 50 of 100 heads | 7.96% | 1 in 12.6 |
| 45-55 heads of 100 | 72.9% | 1 in 1.37 |
Practical Examples
Example 1 — Deciding between two restaurants: You and a friend cannot agree whether to eat pizza or sushi. Assign heads = pizza, tails = sushi, commit to the result, and tap Flip. This is the classic use case for a coin flip, called a tie-breaker in decision theory, and it is guaranteed to be fair for both parties. Example 2 — Testing randomness: Use multi-flip mode to run 1,000 flips. You should see heads between roughly 470 and 530 times (95 percent confidence interval), confirming that the generator is not biased. If you consistently see results outside that range across many batches, either the generator is broken or you happened to observe a rare event. Example 3 — Starting a game: Football, cricket, and tennis all use coin flips to decide which side serves or kicks off. The NFL uses a real silver commemorative coin, while virtual coin flips like this one are increasingly used for online tournaments and remote matches where no physical coin is available.
Tips and Best Practices
Decide the mapping first: before flipping, commit which side represents each option; deciding after the flip defeats the purpose. Accept the result: coin flips only work as decision tools if you follow the outcome; second-guessing turns them into a coin-based rationalization. Notice your gut reaction: if the result disappoints you, you have just learned which option you actually wanted; this meta-use of coin flips is a well-known trick for breaking indecision. Avoid the gambler's fallacy: after five heads in a row, the next flip is still 50/50; streaks are not predictive. Use multi-flip mode for real experiments: single flips give noisy data; 100 or 1,000 flip batches show the law of large numbers clearly. Do not use a coin flip for weighted choices: if the options are not equally attractive, use a weighted randomizer instead. Prefer the virtual tool for remote decisions: when people cannot be in the same room, a virtual coin flip removes any dispute about whether the coin was caught fairly or the call was made in time.
Frequently Asked Questions
Is a virtual coin flip truly random?
Yes. This flip-a-coin tool uses crypto.getRandomValues, the browser's built-in cryptographically secure random number generator defined by the W3C Web Crypto API. It pulls entropy from operating-system sources such as /dev/urandom on Linux and CryptGenRandom on Windows, which pass NIST SP 800-22 statistical randomness tests. Over millions of flips, the ratio of heads to tails converges on the theoretical 50/50 split, exactly matching a fair physical coin. Standard Math.random is also roughly fair but is not cryptographically secure and should be avoided for any sensitive decision.
What is the probability of flipping heads 10 times in a row?
The probability of flipping heads 10 times in a row on a fair coin is (1/2) to the 10th power, which equals 1/1024 or about 0.0977 percent. This also means on average you need 1,024 attempts before the streak appears, though any single trial can produce the streak or never produce it. The same math applies to any fixed outcome: 5 heads has a 1 in 32 chance, 20 heads has a 1 in roughly 1 million chance, and 30 heads is about 1 in 1.07 billion. For statistical tests of coin fairness, these probabilities are the baseline you compare against.
Are real coin flips 50/50?
Not exactly. A famous 2007 paper by Persi Diaconis, Susan Holmes, and Richard Montgomery at Stanford showed that a physical coin, caught in the hand, has a slight bias of about 50.8 percent toward the side that was facing up when the flip started. The bias comes from the precession of the flipping axis, which means the coin spends slightly more time on its starting side. A 2024 replication study with 350,757 recorded flips confirmed a 50.8 percent same-side bias. Virtual coin flips do not have this physical bias because they are not constrained by real-world dynamics.
How do I use a coin flip to decide something?
Pick which side represents each option before you flip, commit to following the result no matter what, and then press the Flip button once. If you feel disappointment when you see the result, that is useful information about which option you actually preferred and can override the coin flip with that preference. Coin flips are most useful for tie-breaking between roughly equivalent choices where any decision is better than no decision. Sports referees use them at the start of games for the same reason, as codified in Rule 2.3 of the NFL Rule Book and similar rules in soccer and cricket.
Can I flip more than one coin at a time?
Yes. Set the multi-flip count to any number up to 1,000 and click Flip to run that many independent tosses at once. The tool reports the counts of heads and tails, the percentage, and the longest streak observed. Running large batches is a good way to see the law of large numbers in action: with 10 flips the ratio will often deviate wildly from 50 percent, but with 1,000 flips the ratio typically sits within 1 or 2 percentage points of 50 percent, demonstrating how randomness converges over long runs.
Why does the coin sometimes show long streaks of the same side?
Streaks are normal in random processes because the coin has no memory. The probability that 100 flips contain a streak of at least 6 heads or tails is roughly 80 percent, and a streak of 8 in the same 100 flips is about 32 percent, according to the formula for longest runs in binary sequences. People often mistake streaks for bias because human intuition expects randomness to alternate more than it actually does. If you see a run of 5 heads in a row, the next flip is still exactly 50/50, an effect known as the gambler's fallacy.