Chebyshev's Theorem Calculator
Chebyshev's theorem (also written Chebyshev's inequality) gives a guaranteed lower bound on the proportion of data that falls within k standard deviations of the mean - for ANY distribution, regardless of shape. Enter k to get the minimum percentage, or enter a desired probability to find the k you need. The result updates instantly and the steps panel shows every line of working.
Formula
Worked example
For k = 2: minimum proportion = 1 - 1/4 = 0.75, so at least 75% of any distribution lies within 2 standard deviations of the mean. Reverse: to guarantee 90%, k = 1/sqrt(1-0.90) = 1/sqrt(0.10) = 3.162.
What is Chebyshev's theorem?
Chebyshev's theorem (also called Chebyshev's inequality after the Russian mathematician Pafnuty Chebyshev) is one of the most general results in probability theory. It states that for any distribution with a finite mean and a finite variance, the proportion of observations that fall within k standard deviations of the mean is at least 1 - 1/k^2. The theorem requires only two conditions: k must be greater than 1, and the distribution must have a defined mean and variance. No assumption is made about the shape - the data can be normal, heavily skewed, multimodal, or drawn from a completely unknown distribution. This universality is the theorem's main strength and the reason it appears in quality control, finance, data science, and actuarial science.
How to use this calculator
In forward mode, enter a value for k (the number of standard deviations from the mean). The calculator applies the Chebyshev formula 1 - 1/k^2 and returns the minimum proportion of data guaranteed to fall within that symmetric interval. Optionally supply the mean and standard deviation to display the actual numeric interval in your data's units. In reverse mode, enter a target percentage (for example, 90 to guarantee that at least 90% of data lies within the interval) and the calculator solves k = 1 / sqrt(1 - p) to find the minimum k you need. The steps panel shows every line of arithmetic. The reference table lists pre-computed values for common k choices.
Chebyshev's theorem vs. the Empirical Rule
The Empirical Rule (68-95-99.7 rule) states that for a normal distribution, approximately 68% of data falls within 1 standard deviation, 95% within 2, and 99.7% within 3. These are tighter guarantees than Chebyshev's bounds because they exploit the specific bell-curve shape. Chebyshev's theorem guarantees only 75% within 2 standard deviations and 88.9% within 3, but it works for every distribution. When you know your data is approximately normal, use the Empirical Rule for precision. When the distribution is unknown or clearly non-normal (income, asset returns, waiting times), Chebyshev's inequality is the right tool because it cannot be invalidated by skewness or heavy tails.
Practical applications
Quality control engineers use Chebyshev's theorem to set control limits that are valid regardless of whether the process output is normally distributed. Financial analysts apply it to place bounds on portfolio returns without assuming a specific return distribution. Data scientists use it as a sanity check: if more than 1/k^2 of observations fall outside k standard deviations, that is statistical evidence of an outlier-heavy or asymmetric distribution. In medicine and public health, the theorem provides conservative confidence intervals when the underlying distribution of a biomarker or outcome is unknown. The reverse calculation is especially practical: you decide what proportion of cases you need to capture, and the calculator tells you how wide your interval must be in standard deviation units.
Chebyshev's theorem: common k values
| k (std devs) | Chebyshev min % | Empirical Rule %* | Max outside % |
|---|---|---|---|
| 1.5 | 55.56% | 86.64% | 44.44% |
| 2 | 75.00% | 95.45% | 25.00% |
| 2.5 | 84.00% | 98.76% | 16.00% |
| 3 | 88.89% | 99.73% | 11.11% |
| 4 | 93.75% | 99.994% | 6.25% |
| 5 | 96.00% | >99.999% | 4.00% |
| 6 | 97.22% | >99.999% | 2.78% |
| 10 | 99.00% | >99.999% | 1.00% |
Minimum percentage of observations guaranteed within k standard deviations of the mean, for any distribution. Empirical Rule values (in parentheses) apply to normal distributions only.
Frequently asked questions
What does Chebyshev's theorem tell you?
Chebyshev's theorem gives a guaranteed lower bound on the proportion of data within k standard deviations of the mean, valid for any probability distribution with a finite mean and variance. For k = 2, at least 75% of the data must lie within 2 standard deviations; for k = 3, at least 88.9% must. The bound is conservative: the actual proportion is often higher, especially for bell-shaped distributions.
Why must k be greater than 1?
For k = 1, the formula 1 - 1/k^2 gives zero, which is trivially true but useless. For k less than 1, the formula gives a negative number, which is meaningless as a probability. Chebyshev's inequality only produces a useful (positive) lower bound when k > 1. The theorem is therefore most commonly applied at k = 2, 3, 4, or 5.
How is Chebyshev's theorem different from the Empirical Rule?
The Empirical Rule (68-95-99.7) applies only to normally distributed data and gives precise percentages for k = 1, 2, and 3. Chebyshev's theorem applies to any distribution but gives weaker (more conservative) bounds: at least 75% within 2 standard deviations (vs. 95.45% for normal) and at least 88.9% within 3 (vs. 99.73% for normal). Use the Empirical Rule when you know the data is normal; use Chebyshev when the distribution is unknown.
How do I find k given a target probability?
Rearrange the Chebyshev inequality: if the desired minimum proportion is p, then k = 1 / sqrt(1 - p). For example, to guarantee at least 90% (p = 0.9), k = 1 / sqrt(0.10) = 3.162. The reverse-solve mode in this calculator does this automatically.
Can Chebyshev's theorem be used with sample data?
Yes. When working with a sample, substitute the sample mean (x-bar) for the population mean and the sample standard deviation (s) for the population standard deviation. The theorem still guarantees that at least 1 - 1/k^2 of the sample values lie within k sample standard deviations of the sample mean. This is sometimes called the sample Chebyshev inequality.
What is the weakness of Chebyshev's bound?
The bound is often very conservative. For a normal distribution, Chebyshev says at least 75% within 2 standard deviations, but the actual figure is 95.45%. The gap exists because Chebyshev must cover the worst-case distribution. In practice, if you know more about the shape of your data (even just that it is unimodal), tighter inequalities such as the Vysochanskii-Petunin inequality can be applied.