Enter one finish time and this tool does three things with it. First, the basic conversion — pace, speed, and the reverse calculation, so entering a target pace instead works just as well. Second, it places your time on the real population distribution of finishers at that exact distance, modelled as a log-normal curve (the standard, well-documented shape of race-time distributions) calibrated against published average and beginner benchmarks. Third, it projects your current level to every other standard distance using an endurance fatigue exponent — 1.06 for running, the well-validated Riegel constant published in 1977 and confirmed across tens of millions of race results since; a gentler, less rigorously established estimate for cycling and swimming, disclosed as such.
A 25:00 5K run projects, via the Riegel formula, to almost exactly 52:00 for 10K, 1:55:00 for a half marathon, and 4:00:00 for a marathon — matching the commonly-cited reference conversion for that exact 5K time. The same run sits at roughly the 75th percentile of all 5K finishers — faster than about three-quarters of everyone who lines up.
It was derived from — and validated against — running race results specifically. Cycling's aerodynamic drag dominates its energy cost in a way running's air resistance doesn't, and swimming's water resistance scales differently again, so this tool uses gentler, less rigorously sourced fatigue estimates for those two sports rather than borrowing running's constant wholesale.
It's the estimated fraction of all finishers at that distance who are slower than your time — a 75th percentile 5K means you'd finish ahead of roughly 75% of a typical race field, based on a statistical model of the population, not a specific race's actual results.
Published "average" race times include the entire finisher population — first-timers, charity participants and run-walkers alongside serious amateurs — which pulls the average well below what a regularly-training recreational athlete would post.
Riegel's formula (and its cycling/swimming analogues here) assumes equivalent training for both distances — a fast 5K only predicts a fast marathon if the underlying endurance base actually supports it, which is why race-prediction tools are most accurate between distances that aren't too far apart.