Race-time predictor

A known race result gives an approximate equivalent performance at another distance. We use the Riegel power formula (exponent 1.06): pace generally slows as the distance increases. An equivalent assumes preparation for that distance; it does not tell us whether you have that endurance.

For a marathon estimate we combine your race result with your typical weekly running distance during preparation for the target marathon, using the Vickers–Vertosick one-race model. Use your typical distance, not your biggest week or an amount you hope to run. The study did not define a precise averaging window. More weekly distance improves the model’s number mathematically; it is not advice to increase training.

This version accepts 5 km, 5 miles, 10 km, 10 miles and half-marathon road races. It does not support walking, run/walk strategies, trails or ultras. First-marathon readiness is not modelled. Older, slower and low-volume runners were weakly represented, and marathon estimates can still be too optimistic.

When weekly distance is missing or outside our display limits, or the calculated time is outside 2:20–6:00, we show a labelled equivalent instead of a training-based estimate and range. These display limits are conservative product choices, not validated accuracy boundaries or minimum preparation advice. Inconsistent inputs also receive an explanation rather than an adjusted forecast.

Your marathon result could vary considerably from the estimate. This approximate range uses the spread of results in a recreational-runner study and gets wider for longer estimated times. It leaves more room on the slower side because some runners take considerably longer than the model predicts.

It is a guide, not a promise or a race target. Your preparation, pacing, fuelling, course and weather can put your result outside the range. The calculator does not assess whether you are ready to run a marathon, and does not directly measure all these factors.

Use a recent race you ran at full effort, at a similar level of fitness and in normal conditions. If you have a recent half-marathon result, use that; a shorter race such as a 5K tells us less about marathon endurance. We also check the faster end against the pace of your recent race before rounding the times for display. Rounding down can put the displayed faster end slightly below that exact pace check. The estimate need not be the midpoint.

The marathon point model is from Vickers and Vertosick (2016), published under CC BY 4.0. It uses a 1.07 Riegel exponent internally, distinct from our shorter-distance equivalents. We use raw race times with a normal-conditions assumption; the study also adjusted for reported course difficulty.

RGN adapted the display with an asymmetric error-scaled range, a conditional source-pace check and outward five-minute rounding. The range was explored retrospectively on the original 2014 self-reported survey data, which had already been examined repeatedly. It has no independent external validation and supplies no personalised probability or reliable subgroup guarantee. Some slower-runner ranges are very wide.

About the Riegel equivalent-performance formula. No third-party calculator code or participant records are used by this tool.