Race Time Predictor

Last updated: 2026-05-22

Predicting a Race From Another Race

You ran a 10K in 45 minutes. What's a realistic marathon target? You set a 20-minute 5K — could you run a sub-90 half? Race-time prediction answers these questions by extrapolating from a known performance to a different distance. Enter one recent race result and this calculator returns a headline prediction for your target distance plus a table of predictions for every standard distance from 1500 m to the marathon, using either of two established models.

The Riegel Model

The most widely used formula is Riegel's (Pete Riegel, 1981). It observes that race time scales with distance raised to a power slightly above 1:

$t_2 = t_1 \cdot \left(\frac{d_2}{d_1}\right)^{e}$

Take logs and it linearizes:

$\log t_2 = \log t_1 + e \cdot \log\!\left(\frac{d_2}{d_1}\right)$

The exponent $e$ — standard value 1.06 — captures the empirical fact that you slow down a little for each doubling of distance. At $e = 1.06$, doubling the distance costs about 4.3% on pace ($2^{1.06} \approx 2.085$). An exponent of 1.0 would imply you can hold any pace for any distance (impossible); 1.20 would imply a much steeper drop than trained runners actually show.

Because the exponent is an empirical best-fit rather than a law of physiology, this calculator lets you adjust it. Lower it toward 1.03 if you are a speed-oriented runner whose pace holds up well over distance; raise it toward 1.10 if your endurance fades faster than average — research by Tanda and others suggests the marathon in particular is often better fit by an exponent above 1.06.

The Cameron Model

Riegel uses one exponent for all distances. Cameron's formula (David Cameron, 1998) instead uses a distance-dependent fatigue function, fitted to world-class results from 400 m to 50 miles:

$t_2 = t_1 \cdot \frac{d_2}{d_1} \cdot \frac{f(d_1)}{f(d_2)}, \qquad f(x) = 13.49681 - 0.000030363\,x + \frac{835.7114}{x^{0.7905}}$

with $x$ in metres. Because the fatigue term changes shape across distances, Cameron is often slightly more realistic at the long end, predicting marathon times a little slower than Riegel. Switch between the two models and compare: if they agree closely, the prediction is robust; if they diverge, the truth is usually somewhere between.

Examples

The two models track closely at nearby distances and fan apart as the gap from the known race grows — exactly where a prediction is least certain anyway.

Practical Scenarios

1. Setting a Marathon Goal

A common rookie mistake is targeting a marathon time by simply doubling a recent half. Both models warn you off it: a 45-minute 10K predicts a ~3:27–3:30 marathon, not 3:00 (which would need a 37:30 10K). Many would-be sub-3 marathoners have a 10K time that quietly tells them the goal needs more training first.

2. Choosing a Pace Group

Race-day pace groups are offered at fixed intervals (3:30, 3:40, 3:50 marathon, etc.). Read your predicted time off the table and pick the group that matches it — not your aspirational time. When Riegel and Cameron straddle a boundary, the more cautious group is the safer start.

3. Comparing Distances Across Runners

Two friends ran different races: a 19:30 5K and a 1:30 half marathon. Who is faster? The all-distance table converts any result to a common basis — read the 5K runner's predicted half (or the half runner's predicted 5K) and compare directly.

4. Pacing a Track Event

The formulas also extrapolate down: a 16:00 5K predicts roughly a 4:28 1500 ma 16:00 5K predicts roughly a 4:49 mile. Track athletes use this to check that their endurance translates into race-distance speed.

Where Prediction Falls Short

• It assumes consistent training across distances. A runner who only trains 5Ks will under-perform the marathon prediction; long-run endurance is a separate adaptation.
• The models are empirical. Riegel's exponent and Cameron's fatigue function are regression fits, not derivations. Real marathon performance often lags both predictions for runners not specifically marathon-trained.
• They break at the extremes. Sub-1500 m sprintsSub-mile sprints and ultra-distances (50K+) follow different physiology and neither model extrapolates cleanly there.
• They cannot account for course profile, weather, or pacing. Predictions assume similar conditions; a flat fast 10K and a hilly windy marathon are not directly comparable.

For amateur runners targeting standard distances under similar conditions, race prediction is a useful planning tool — better than guessing — but treat the numbers as a ballpark, not a guarantee.