Race Predictor Calculator

It fits a power law between distance and time: T₂ = T₁ × (D₂/D₁)^k. The exponent k describes how much your pace fades as distance grows. If k were 1.0 your pace would never slow, so you would run a marathon at 5K pace. Real values sit a little above 1.

Peter Riegel's power law with the exponent fixed at 1.06, published in Runner's World in 1977 and refined in 1981. He derived 1.06 from competitive results between 800 m and the marathon. It is the most widely used race predictor in the world and is a reasonable average — but an average is not you.

Because the error compounds with the extrapolation. Two runners with identical 5K times but exponents of 1.04 and 1.10 are separated by well over twenty minutes at the marathon. Your exponent is measurable from two races you have already run, so assuming it is a choice, not a necessity.

There is no good or bad, only a profile. Elite distance runners cluster near 1.05, recreational runners near 1.06, and runners with limited aerobic base at 1.08 or above. A low exponent means you hold pace well over distance; a high one means speed is ahead of endurance.

Two recent, genuinely maximal efforts from the same period of fitness, at clearly different distances. A 5K and a 10K, or a 10K and a half. The further apart the distances, the more reliable the exponent — but only if both were truly raced.

Yes. Clear the second time and the page falls back to Riegel's 1.06 and says so. You get a projection, but it is the textbook assumption rather than a measurement of you.

A 1998 refinement in which the exponent falls as the anchor race lengthens: 1.06 − 0.0026 × the anchor distance in miles. It exists because a fixed 1.06 over-predicts for elite runners and under-predicts for slower ones over long extrapolations. It is shown here as a third opinion.

Because the longer race was run at a faster pace than the shorter one, which no physiology supports in two maximal efforts. Almost always the short race was a training effort, a hilly course, or from a different year. The page refuses to project from such a pair.

Not very. Projections are most reliable between roughly half and twice the anchor distance. A marathon is over eight times a 5K, so the arithmetic is precise but the assumption behind it is doing enormous work. Use a half marathon as the anchor if you have one.

No, and that is its central weakness. The curve knows two results and nothing else. A runner who has never exceeded 25 km in training will not run the marathon the curve projects, regardless of how good their 10K is, because the limiter is durability rather than pace.

Because it shortens the extrapolation. Projecting a marathon from a half is a factor of two; projecting it from a 5K is a factor of eight. The engine anchors on the longer of your two results for exactly that reason and says so in the output.

Substantially, and the model cannot see either. A hilly or hot race produces a slow time that inflates your apparent exponent and makes every longer projection pessimistic. Use results from comparable conditions.

As a starting point, not a commitment. When the three columns disagree, plan against the slower one. A marathon started two per cent too fast does not cost two per cent at the finish; it costs far more, because the collapse is non-linear.

After every race that is genuinely raced. Your exponent shifts as your training shifts: a winter of long runs lowers it, a block of track work raises it. Recomputing after each race turns the exponent into a training feedback signal.

Because they assume different exponents, or use table lookups such as Daniels' VDOT rather than a power law. They agree on the shape of the curve and disagree on how fast pace fades. That disagreement is real uncertainty, not a bug in one of them.

No. VDOT converts a race result into an oxygen-cost figure and reads equivalent performances off tables derived from Daniels and Gilbert's work. A power law is simpler and, once your own exponent is solved, personalised in a way generic VDOT tables are not. They usually agree closely at nearby distances.

Not reliably. Beyond the marathon the power law breaks down: terrain, night running, fuelling and sleep dominate, and the exponent required to describe a 100-mile result bears little relation to the one that describes road racing.

Indirectly. Masters runners often show a lower exponent, holding pace over distance well while losing top-end speed. That is a real shift in profile and it is exactly the kind of thing a fixed 1.06 conceals.

Then the exponent describes a runner who no longer exists. Fitness moves. Use results from the same training block, ideally within six to eight weeks of each other, or the number mixes two different athletes.

Most likely because they used a fixed 1.06 and you solved your own. If your exponent is above 1.06, every generic calculator on the internet will flatter you. The uncomfortable number is usually the useful one.