Coin Flip Probability Calculator
Flip a coin 10 times and exactly 6 heads feels like it should be common, yet it happens only about one time in five. Intuition is a poor guide here because dozens of sequences compete for the same ten flips, and this calculator replaces the guesswork with the exact binomial probability for the count you actually care about.
Set the number of flips, pick exactly, at least, at most, more than, fewer than, or a between range, and the tool returns the probability as a decimal and a percentage, the odds, the expected number of heads, a distribution chart with your range highlighted, and a plain reading of how surprising the result really is. Fair and biased coins are both covered.
Strict modes exclude the boundary: more than 5 means 6 or more, fewer than 5 means 4 or fewer. Probabilities assume every flip is independent and the coin bias stays constant.
Quick Answer
Key takeaways
- One exact count is rarely likely on its own: the most probable single outcome in 10 flips, exactly 5 heads, is still only 24.6094%.
- Ranges behave very differently from single counts: 4 to 6 heads covers 65.625% of all outcomes, which is why at least, at most, and between questions usually matter more.
- At least and more than are not the same question. The boundary count k is included in one and excluded from the other, and in 10 flips that single count can hold a quarter of the probability.
- A previous run of heads changes nothing. Independence is built into the formula, and it is also how real coins behave.
- Quick Answer
- How to use this coin flip probability calculator
- The binomial formula behind every answer
- Worked example: exactly 6 heads in 10 flips
- Why streaks and lopsided counts feel rigged when they are not
- What competing calculators miss
- Frequently Asked Questions
- Add This Calculator to Your Website
- References
How to use this coin flip probability calculator
Start with the number of flips, n. Ten is a good size for building intuition because every outcome can be listed, but the tool accepts up to 1,000 flips for classroom simulations, quality sampling, and game balancing work. Then choose the comparison that matches your question. Exactly k is the textbook case. At least k and at most k are the cumulative questions people usually mean in real life, such as the chance of at least 7 heads when a best-of series is on the line. More than k and fewer than k are the strict versions that leave k out, and between k1 and k2 covers a band, which is the right shape for asking how often a result lands in the ordinary middle.
Leave the coin set to fair unless you have a reason to change it. The custom bias setting exists for weighted coins, for sensitivity checks (how much would the answer move if this coin were slightly unfair?), and for any repeated yes-or-no process with a known success rate, from free throws to defective items on a line. In that wider use, heads is just the label for success, p is the success rate, and the same binomial math applies. If you only need the counting piece on its own, the number of sequences behind these probabilities comes from the binomial coefficient calculator, which computes C(n, k) directly.
The binomial formula behind every answer
The formula counts sequences first and weighs them second. There are C(n, k) different orders in which k heads can appear in n flips, each individual order has probability p^k (1-p)^(n-k), and multiplying the count by the weight gives the probability of the count itself.
The treatment in OpenStax Introductory Statistics, Section 4.3, Binomial Distribution sets out the four conditions that make this the right model: a fixed number of trials, two outcomes per trial, independent trials, and the same success probability on every trial. Coin flips meet all four by construction. The cumulative modes are sums of the same building block: at least k adds the exact probabilities from k up to n, at most k adds from 0 up to k, and the strict modes simply move the boundary by one. The calculator performs those sums exactly rather than approximating them, which is why a result like 37.6953% for more than 5 heads can be checked by hand from the outcome table.
Worked example: exactly 6 heads in 10 flips
The calculator is pre-filled with this case. C(10, 6) = 210, because there are 210 orders in which six heads can land in ten flips. Each order has probability 0.5^6 x 0.5^4 = 0.000977, and 210 x 0.000977 = 0.205078. The table below shows the full spread for 10 flips, which is also the table the tool prints under its result.
| Heads (k) | P(X = k) | As % | Reading |
|---|---|---|---|
| 0 | 0.000977 | 0.0977% | All tails: a genuine rarity |
| 3 | 0.117188 | 11.7188% | Off-center but routine |
| 5 | 0.246094 | 24.6094% | Single most likely count |
| 6 | 0.205078 | 20.5078% | The worked example |
| 8 | 0.043945 | 4.3945% | Noticeably lopsided |
| 10 | 0.000977 | 0.0977% | All heads: same rarity as all tails |
Two comparisons from the same distribution are worth memorizing. Between 4 and 6 heads covers 0.205078 + 0.246094 + 0.205078 = 0.65625, so the middle band holds about two thirds of all outcomes. And at least 5 heads, at 62.3047%, is a completely different claim from more than 5 heads at 37.6953%, even though the two phrases sound nearly identical in conversation. Most arguments about coin flip odds are really arguments about which of these questions was asked. For questions that stack two separate events rather than repeat one trial, the compound probability calculator handles the and-or combinations instead.
Why streaks and lopsided counts feel rigged when they are not
A fair coin produces results that look patterned far more often than people expect. In 10 flips, the chance of at least one run of four or more identical results somewhere in the sequence is substantial, and a 7 to 3 split, which feels like evidence of a loaded coin, has a probability of about 11.7188% in one specified direction and roughly double that counting both directions. The distribution is simply wider than intuition suggests: the standard deviation at n = 10 is 1.5811 heads, so counts from about 3.4 to 6.6 cover only the middle 68% or so, and everything outside that band is collectively common.
As n grows, the distribution tightens in relative terms even as it widens in raw counts. At 100 flips the standard deviation is 5 heads around an expectation of 50, and at 1,000 flips it is about 15.8 around 500, which is why long-run frequencies settle so reliably toward 50%. That settling behavior is exactly what the central limit theorem calculator demonstrates from the sampling side. The practical lesson cuts both ways: short sequences prove almost nothing about a coin, and a genuinely biased coin, say p = 0.6, needs hundreds of flips before its tilt separates clearly from ordinary luck.
What competing calculators miss
The coin flip tools reviewed for this page, including Vedantu, CalcVenue, QuickAI, FlipA Coin Online, and OnlyCalculators, get the core binomial formula right. The differences are in how much of the question set they answer and how much of the working they show.
Miss 1: the strict modes are often absent. Several tools offer exactly, at least, and at most, then stop. Anyone asking about more than k or fewer than k has to redo the problem with a shifted k and hope they shifted the right way. This calculator includes all six modes, with the boundary behavior stated under the tool so there is no guessing.
Miss 2: a percentage with no shape. A lone percentage hides where your outcome sits in the distribution. The highlighted chart and, for n up to 30, the full outcome table show whether your count is central, off to one side, or out in a tail, which is usually the thing you actually wanted to know.
Miss 3: no reading of the result. A figure like 0.205078 does not tell a student whether to be surprised. Every result here carries a plain-language interpretation with odds and a rarity band, plus the step-by-step working, so the answer can be checked, quoted, and understood rather than just copied.
One honest limit: the model assumes a constant p and independent flips. A real coin flipped by a machine, spun on a table, or caught the same way every time can depart slightly from that ideal, and no calculator setting can capture technique.
Frequently Asked Questions
At least k includes k itself, so at least 5 heads in 10 flips means 5, 6, 7, 8, 9, or 10 heads, which comes to 62.3047%. More than k excludes k, so more than 5 heads means 6 through 10, which is 37.6953%. The same split applies at the low end: at most k includes k, fewer than k excludes it.
It is 24.6094%. There are C(10,5) = 252 sequences with exactly five heads out of 1,024 equally likely sequences, and 252 divided by 1,024 is 0.246094. Five heads is the single most likely count, yet it still happens less than one time in four.
No. Each flip is independent, so a fair coin stays at 50% heads on the next flip no matter what came before. Believing a tails is due after a run of heads is the gambler’s fallacy. Runs feel surprising, but in 10 flips some kind of streak is normal, not suspicious.
Yes. Switch the coin setting to Custom bias and enter p, the probability of heads on a single flip, such as 0.6. The probability, chart, full outcome table, expected value, and standard deviation all recompute for that bias. With p = 0.6, exactly 6 heads in 10 flips rises to 25.0823%.
The expected value is a long-run average, not a prediction for one set of flips. In 10 fair flips the expected count is exactly 5 and 5 is also the most likely single count, but with a biased coin they separate: at p = 0.6 the expectation is 6 heads, and the distribution leans toward the high side without guaranteeing any single outcome.
One bar for every possible head count from 0 to n, with the bar height equal to that count’s probability. Bars inside the range you asked about are highlighted in dark blue, so you can see at a glance how much of the total probability your question covers. For more than 120 flips the counts are grouped into 120 bins so the chart stays readable.
The calculator accepts up to 1,000 flips. The full outcome table, which lists P(X = k) for every k, is shown when n is 30 or fewer because longer tables stop being useful to read. The chart, expected value, and standard deviation are produced at every size.
For a fair coin it is (1/2)^n for each side. All heads in 5 flips is 1 in 32, or 3.125%. All heads in 10 flips is 1 in 1,024, or about 0.0977%. Use Exactly k with k set to n, or k set to 0, to get either figure for any number of flips.
Add This Calculator to Your Website
References
- OpenStax, Rice University, “Introductory Statistics, Section 4.3: Binomial Distribution”: the four binomial conditions, the probability formula, mean np, and standard deviation. OpenStax: Binomial Distribution. Accessed October 2026.
- OpenStax, Rice University, “Introductory Statistics, Chapter 4 Formula Review”: consolidated binomial notation and formulas used to check the expected value and standard deviation shown by this tool. OpenStax: Chapter 4 Formula Review. Accessed October 2026.
- National Institute of Standards and Technology, NIST/SEMATECH e-Handbook of Statistical Methods and Dataplot reference material on binomial proportions: reference definitions for binomial counts and proportions. NIST: Binomial Proportion reference. Accessed October 2026.
- Vedantu, “Coin Flip Probability Calculator”: competitor formula presentation and conversion-table format reviewed during the competitive analysis for this page. Vedantu: Statistics Coin Flipper. Accessed October 2026.
- OnlyCalculators, “Coin Flip Probability Calculator”: competitor step-by-step layout and rarity bands reviewed during the competitive analysis for this page. OnlyCalculators: Coin Flip Probability Calculator. Accessed October 2026.