How the race time predictor works
The Riegel formula
Peter Riegel's 1981 endurance equation: T₂ = T₁ × (D₂ / D₁)^1.06 — double the distance, and time grows by 2^1.06 ≈ 2.09×. The exponent captures the endurance decay every distance runner feels: speed falls as fuel systems, fatigue resistance, and economy are taxed longer.
Why the exponent matters more than the formula
At 1.06, a 50:00 10K predicts a 3:42:46 marathon. At 1.08 — a runner with speed but thin endurance — the same 10K predicts 3:49. That seven-minute spread is the honest uncertainty of extrapolating 2–4× beyond your proven distance. The rule of thumb: predictions trust themselves in proportion to how close the target is to what you've trained for. A 10K PR predicts a half well if you've been running 40+ miles weekly; it flatters everyone's marathon.
Reading the result
- Target within 2× of your race: treat the prediction as a real target and race to it.
- First marathon, predicted from a 5K/10K: aim for the 1.07–1.08 output, bank 5 minutes of buffer in the first half, and let the last 10K decide the final time.
- Target shorter than your race: the prediction is conservative — short-distance speed needs dedicated training (strides, intervals) that long runs do not provide.
What breaks predictions
Heat (pace slows ~1.5–3% per 5°C above 15°C), hills, wind, crowding at the start, and inadequate fueling past 90 minutes. Predictions assume neutral conditions and comparable training. The formula has no idea if your longest run was 8 miles — you do, and that is why the exponent selector exists.