Glossary
Race predictor
A model that turns one race result into a prediction at another distance. From a 25:00 5K, 3 published models put the marathon about 6 minutes apart.
What it is
A formula that maps a result at one distance onto another. The oldest is Riegel's, from 1981. Daniels and Gilbert's VDOT tables and Cameron's factor are the other two in common use.
Watches do something different again. They estimate your aerobic fitness from recent training, then predict from that.
| Source of the prediction | What it actually knows |
|---|---|
| Riegel, VDOT, Cameron | one race result |
| a watch | recent pace at recent heart rates |
| a training app | your logged sessions |
What it is not
It is not a forecast of your race. None of these models knows your weekly volume, your longest run, the course or the weather.
That is why two apps disagree so loudly. One is reading your speed, the other your training, and neither is reading your endurance. We took the disagreement apart here.
What moves it
- How far you extrapolate. A marathon from a 5K is the least reliable case.
- Weekly mileage. Vickers and Vertosick found it improves prediction.
- The freshness of the input race. A tired race gives a slow answer.
- The exponent chosen. Riegel at 1.06 and 1.07 differ by minutes.
When to ignore it
Ignore any single number quoted to the second. The honest output is a range.
Ignore a marathon prediction built from a 5K if you have not run past 25 km (15.5 mi) in training.
Sources
- Riegel PS. Athletic records and human endurance. American Scientist 69(3):285-290, 1981 primary
- Vickers AJ, Vertosick EA. An empirical study of race times in recreational endurance runners. BMC Sports Sci Med Rehabil 8:26, 2016 primary
- Keogh A et al. Prediction equations for marathon performance: a systematic review. Int J Sports Physiol Perform 14(9):1159-1169, 2019 primary