Enter one or more recent race times to get an estimated marathon range. It's an estimate, not a promise.
Enter at least one race time.
Riegel, P. S. (1981), “Athletic records and human endurance”, American Scientist 69(3). Time scales with distance raised to an exponent; 1.06 is Riegel's value, and 1.1 gives a more cautious estimate for the marathon. Multiple inputs are averaged.
It uses Riegel's formula, T2 = T1 × (D2 ÷ D1)^k. The faster end of the range uses Riegel's published exponent (1.06); the slower end uses 1.1, closer to what many recreational runners actually run. With several races entered, the predictions are averaged.
Shorter races mostly test speed; the marathon tests endurance and fuelling. Low weekly mileage, few long runs, heat, hills or a first marathon all push the result towards — or past — the slow end of the range.
A recent half marathon, because it is closest in distance. 5K predictions are the least reliable for the marathon.