Free race time predictor

How the race predictor works

The calculator uses Pete Riegel's endurance model, first published in Runner's World in 1981: T2 = T1 × (D2 / D1)^1.06. You supply one honest performance, T1 over distance D1, and the model scales it to the distance D2 you want to predict.

The exponent of 1.06 is where the interesting part happens. If pace were perfectly constant, doubling the distance would double the time and the exponent would be 1.0. Because runners slow as distance grows — glycogen depletes, heat accumulates, muscular fatigue builds — the exponent sits slightly above 1. That single number is what stops the calculator from telling you that a 20-minute 5K predicts a 2-hour marathon.

A worked example

Take a 20:00 5K road race. Doubling the distance does not double the time:

That progression — 10 seconds per kilometre slower at 10K, 22 at the half, 33 at the marathon — is the shape most club runners actually experience. If your own training says otherwise, the gap between prediction and reality is the most useful number on this page.

What the formula cannot see

The predictor is a pace model, not a coaching model. Four things move real race results away from the prediction:

• Endurance volume. Marathon predictions assume you are running enough long runs to hold form for over three hours. On 40 km weeks, most runners fall 5–10 minutes short of the number.
• Course and surface. Trail, gravel and hilly road courses cost you time the formula knows nothing about. A course with 300 m of climbing typically adds three to five per cent.
• Weather. Warm, humid or windy days slow everyone. A race at 25°C is commonly two to four per cent slower than the same effort at 12°C.
• Execution. Going out 10 seconds per kilometre too fast in the first 5 km costs far more than 10 seconds by the finish. Pacing errors are the biggest single cause of a missed prediction.

How to use the prediction properly

Treat the output as a set of equivalent performances. If you have run 41:42 for 10K, then 1:32 is a fair reflection of your current fitness for a half — not a promise. Three practical uses:

1. Choose a realistic goal race. Pick the distance where your prediction already sits near a time you would be proud of, then build a block towards it.
2. Set training paces. Your predicted race pace is roughly threshold effort. Easy running belongs 60–90 seconds per kilometre slower; interval sessions 10–20 seconds per kilometre faster.
3. Track drift. Re-run the calculation every four to six weeks with a fresh time trial. Predictions that improve steadily mean the block is working; predictions that stall while mileage climbs usually mean you are tired, not fitter.

When the predictor is the wrong tool

Skip it for distances under 1500 m, for ultras, and for trail races with significant climbing. Also be cautious with a marathon debut: the formula assumes you have already absorbed marathon-specific training. For a first marathon, most coaches would rather you picked a goal from long-run pace and recent half marathon form than from a fast 5K.

Where this calculator stops and your own data starts

A single race time is one data point. Inside CollinghamCo Active, the same maths is applied to your rolling 90-day training — volume, pace and consistency — so the prediction reflects what you have actually been doing rather than one good day. Members also get the ACWR calculator alongside it to check whether the training needed to hit that time is being built at a safe rate.