Geometric Mean Percentile

Multiplies your per-match percentiles and takes the Nth root, so a single near-zero match pulls the whole score down hard.

(p1 × p2 × ... × pN) ^ (1/N)

Best at

Groups who believe one catastrophe should count for more than one triumph.

Where it misleads

Brutal on an off day, and the maths is unfamiliar enough that people distrust the number.

How it works

Geometric Mean Percentile takes the geometric mean of per-match percentiles instead of the arithmetic mean. This severely penalizes any match with a very low percentile.

How it works: Multiply all match percentiles together and take the Nth root (where N is the number of matches). Unlike the regular average, a single last-place finish (near 0%) will dramatically lower the score.

When to use it: When you believe consistency is paramount and a single terrible performance should be heavily penalized. It creates a "no bad games allowed" philosophy.

Watch out for: Can be harsh on players who have one off day. The mathematical concept is less familiar to casual players. Best for analytically-minded groups.

Works well with

Chess Terraforming Mars Spirit Island

Also in Average & Percentile

Average PlacementA balanced, instantly readable reading when your matches are roughly the same size.Median PlacementMeasuring someone's typical evening rather than their worst one.Average PercentileMixed field sizes and uneven attendance; the fairest reading in this family, and comparable enough between groups to travel to a global board.Best N of MLong seasons where you want to forgive the occasional disaster and reward a peak.Trimmed MeanA balanced reading that neither a lucky win nor an unlucky collapse can swing.