Average Percentage Calculator

Adding percentages together and dividing by the count, the simple mean $\bar{P} = \frac{\sum P_i}{n}$, is only safe when every value describes a group of the same size. Three test scores of 70%, 80% and 90%, each out of the same total, average cleanly to 80%, because each score carries as much information as the others. Once the group sizes differ, the shortcut starts to mislead. A 90% result from a quiz taken by 30 students tells you more about the class than an 80% result from 20 students, and treating the two as equals distorts the summary. The same trap catches anyone blending conversion rates from pages with very different traffic, or combining exam papers with different point totals. This tool covers both situations: leave the weight fields empty and it returns a simple mean, or enter a weight for each row and it switches to a weighted average that reflects how much data each percentage covers.

A weighted average multiplies each percentage by its weight, sums the products, and divides by the total weight. $$\bar{P} = \frac{\sum P_i \cdot w_i}{\sum w_i}$$ The weights are usually sample sizes: students who sat a test, visitors who saw a page, items in a batch. Suppose a class scores 80% on a quiz taken by 20 students and 90% on one taken by 30. The products are 1,600 and 2,700, the total weight is 50, and dividing 4,300 by 50 gives 86%. A simple mean would report 85% and quietly underrate the larger quiz. The calculator shows each row's contribution to the final figure, so you can see that the second quiz supplies 2,700 of the 4,300 points, and it displays the naive unweighted mean next to the weighted one whenever the two disagree, which is exactly when the weighting matters.

Percentages of different-sized groups can behave in ways that feel impossible. Imagine a site that converts 10% of 100 mobile visitors and 2% of 1,000 desktop visitors in January, a combined 30 sales from 1,100 visits, about 2.7%. In February both rates drop, to 8% on mobile and 1% on desktop, yet the traffic mix flips so mobile now brings 1,000 visitors and desktop only 100. The combined result is 81 sales from 1,100 visits, roughly 7.4%, far higher than before. Every individual rate fell while the overall rate nearly tripled, purely because the mix moved toward the stronger segment. Statisticians call this Simpson's paradox, and it is the strongest argument for weighting: the percentage change calculator can quantify each segment's move, but only a weighted view explains the combined figure.

The most common real case is combining exam results with different point totals. A student who scores 45/50 on one test and 60/80 on another has earned 105 of 130 possible points, an overall 80.8%. Averaging the two percentages, 90% and 75%, would report 82.5% and overstate the result; the marks percentage calculator applies the same sum-first logic to full report cards. Marketers meet the identical problem when blending conversion rates: 5% of 1,000 visitors is 50 conversions and 20% of 50 visitors adds 10 more, so the true combined rate is 60 out of 1,050, about 5.7%, nowhere near the naive 12.5%. Portfolio returns follow the same rule, with the money invested in each position acting as the weight for that position's percentage return.

How many decimals you report should match how much precision your data can support. A weighted average built from a few dozen students, such as the 86% quiz example, does not deserve four decimal places; one decimal is usually honest, and whole numbers are often enough for grades. Conversion rates earn more precision only when the sample is large, since with 1,050 visitors the gap between 5.7% and 5.71% is a fraction of one visitor and means nothing. A good habit is to keep full precision while you calculate and round only the final figure once, because rounding each percentage before averaging can shift the result. The calculator follows that rule, carrying exact values through every step and rounding only in the display. When you need a plain one-off percentage rather than an average of several, the general percentage calculator handles those single computations.