Calculating The Convolution Of Two Random Variables

Interactive Probability Tool

Compute the distribution of the sum Z = X + Y for two independent discrete random variables. Enter support values and their probabilities, choose a preset if you want a fast example, and instantly view the convolved distribution, expected value, variance, and chart.

What this calculator does

Discrete convolution of PMFs

Best use case

Sums like dice, claims, counts, scores

Output included

Distribution table, mean, variance, chart

Expert Guide to Calculating the Convolution of Two Random Variables

Convolution is one of the central operations in probability, statistics, signal processing, queueing theory, reliability analysis, actuarial science, and modern data science. In probability terms, convolution gives you the distribution of a sum of random variables. If X and Y are independent random variables and you define Z = X + Y, the distribution of Z is obtained by convolving the distribution of X with the distribution of Y. This calculator focuses on the discrete case, where each variable has a probability mass function, often abbreviated as a PMF.

Why does this matter? Because many real-world outcomes are sums. A customer order total is the sum of item quantities. Daily arrivals can be the sum of arrivals from two independent channels. Insurance losses over a period can be viewed as sums of claim amounts. Exam scores can be modeled as sums of points from independent sections. Once you understand convolution, you can move from knowing the behavior of parts to understanding the behavior of the whole system.

What convolution means in plain language

Suppose random variable X can take values from one list and Y can take values from another list. To find the probability that their sum equals a specific number z, you look at every pair of values that adds up to z. For each valid pair, you multiply the probability of the X value by the probability of the Y value, assuming independence, and then add all those products together.

For discrete variables, the formula is:

P(Z = z) = Σ P(X = x)P(Y = z – x)

That summation is the convolution. It systematically captures every route that leads to the same total.

Independence is the key assumption for the multiplication step above. If the variables are not independent, convolution in this simple form does not apply, and you need the joint distribution instead.

Step by step method for discrete convolution

1. List the support values of X and their probabilities.
2. List the support values of Y and their probabilities.
3. Form every possible pair (x, y).
4. Compute each sum z = x + y.
5. Multiply probabilities P(X = x)P(Y = y) for each pair.
6. Add together products that produce the same z.
7. Verify the final probabilities add to 1.

This is exactly what the calculator above does. It parses the values, checks whether the probabilities are valid, multiplies all pairwise combinations, groups identical totals, and displays the resulting PMF.

Classic example: sum of two fair dice

The most familiar convolution example is rolling two fair six-sided dice. Each die has outcomes from 1 through 6, each with probability 1/6. The sum ranges from 2 to 12. However, these sums are not equally likely. There is only one way to make 2, but there are six ways to make 7. This is why the PMF of the sum forms a triangular shape.

Sum z	Number of combinations	Probability P(Z = z)	Percentage
2	1	1/36	2.78%
3	2	2/36	5.56%
4	3	3/36	8.33%
5	4	4/36	11.11%
6	5	5/36	13.89%
7	6	6/36	16.67%
8	5	5/36	13.89%
9	4	4/36	11.11%
10	3	3/36	8.33%
11	2	2/36	5.56%
12	1	1/36	2.78%

This table is a textbook illustration of discrete convolution. The support widens after summation, and the central outcomes become more likely because more combinations produce them. If you load the fair dice preset in the calculator, you will see this shape immediately in the chart.

Expected value and variance after convolution

One major reason convolution is powerful is that the moments of sums are often simple to compute. If X and Y are independent, then:

• E[X + Y] = E[X] + E[Y]
• Var(X + Y) = Var(X) + Var(Y)

These identities let you cross-check your convolution table. If your convolved PMF says the mean of Z is not equal to the sum of the component means, something is wrong with the setup or calculations. For example, a fair die has mean 3.5 and variance 35/12, so the sum of two fair dice has mean 7 and variance 35/6, approximately 5.8333.

Discrete versus continuous convolution

The calculator on this page handles discrete random variables because that is the most user-friendly way to interactively enter a distribution. In the continuous case, the same idea applies, but you replace the sum with an integral. If X and Y are independent continuous random variables with densities, then the density of Z = X + Y is:

f_Z(z) = ∫ f_X(x)f_Y(z – x) dx

Conceptually, the operation is the same: multiply matched pieces that combine to the same total, then aggregate. Practically, the continuous case usually requires calculus or numerical approximation.

Common applications of convolution

• Risk analysis: total loss from multiple independent sources.
• Operations research: total demand or total arrivals across channels.
• Quality engineering: combined measurement error from independent components.
• Finance: aggregation of count-based risks or scenario outcomes.
• Machine learning and data science: understanding sums of stochastic variables in simulation pipelines.
• Communications and signal processing: while the formal object may differ, convolution remains foundational.

Comparison table: common sum results for independent random variables

Input distributions	Parameter values	Mean of sum	Variance of sum	Resulting family or shape
Bernoulli + Bernoulli	p = 0.40 and p = 0.70	1.10	0.45	Three-point discrete distribution on 0, 1, 2
Poisson + Poisson	λ = 2.0 and λ = 3.5	5.5	5.5	Poisson with λ = 5.5
Fair die + fair die	Uniform on 1 to 6	7.0	5.8333	Triangular discrete PMF on 2 to 12
Binomial + Binomial	n = 10, p = 0.3 and n = 15, p = 0.3	7.5	5.25	Binomial with n = 25, p = 0.3

The table highlights a useful idea: convolution sometimes preserves a family of distributions. For instance, the sum of independent Poisson variables is again Poisson, and the sum of independent binomials with the same success probability is again binomial. In other situations, such as the sum of two dice, the result is still discrete but not in a named one-parameter family.

How to use this calculator correctly

1. Enter the support for X in one field and its probabilities in the matching field.
2. Enter the support for Y and its probabilities.
3. Make sure each list of probabilities sums to 1, allowing for tiny rounding differences.
4. Click Calculate Convolution.
5. Read the generated PMF table for Z = X + Y.
6. Use the optional highlighted value input if you want the calculator to call out P(Z = z) for a specific total.

Interpretation tips

Do not assume the sum distribution is uniform just because the original variables are simple. Convolution often creates concentration around middle values. Also note that the support of the sum is formed by adding the smallest values together and the largest values together. If X runs from 1 to 6 and Y runs from 1 to 6, then Z runs from 2 to 12. But not every total is equally likely.

Another important point is granularity. If one variable takes only even numbers and another takes only odd numbers, the sum can only take odd numbers. The support pattern of the result depends heavily on the support pattern of the inputs.

Frequent mistakes when calculating convolution

• Using probabilities that do not sum to 1.
• Entering support values and probabilities in mismatched order.
• Forgetting the independence assumption.
• Failing to combine equal sums after multiplying pairwise probabilities.
• Confusing the PMF of the sum with the PMF of one original variable.
• Rounding too early and creating a total probability slightly different from 1.

Why charts help

A chart makes the structure of convolution visible. You can quickly see whether the resulting PMF is symmetric, skewed, peaked, or spread out. For educational use, the chart is often the fastest way to understand why some totals are more likely than others. In the fair dice example, the bar chart climbs steadily to 7 and then falls symmetrically, which reflects the number of combinations producing each sum.

Authoritative learning resources

If you want to study this topic at a deeper mathematical level, these references are strong starting points:

• Penn State STAT 414 Probability Theory
• MIT OpenCourseWare probability materials
• NIST Engineering Statistics Handbook

Final takeaway

Calculating the convolution of two random variables is really about understanding how uncertainty combines. In the discrete independent case, the recipe is straightforward: multiply probabilities for each pair of outcomes, group by common sums, and add. That gives the full distribution of the total. Once you have that distribution, you can answer practical questions about likely outcomes, tail probabilities, averages, variability, and decision risk. Use the calculator above to experiment with custom inputs and build intuition from the chart and PMF table.