Race time predictor
Take a recent race and project it onto other distances, using the Riegel formula — the same one behind most online predictors, just with the fine print attached.
| Distance | Projected time | Pace |
|---|
Where it bends: the formula assumes you're as prepared for the target distance as for the one you raced. The bigger the jump, the more flattering it gets — a 10K time projects a marathon you can only run if you've actually done the marathon training. Treat the long projections as a best case, not a promise.
Under the hood
Take a time you have actually run, and the tool assumes your times over other distances follow a smooth curve from it. Double the distance and you do not double the time — you go a bit slower per kilometre. How much slower is one number, and that number carries the whole tool.
The exponent is the interesting part. At exactly 1.0 you would hold the same pace forever; the amount above 1.0 is how fast you fade as the distance grows.
The numbers that go in
| Number | What it is | Where it comes from |
|---|---|---|
| 1.07732 | the fade exponent | Riegel (1981), Table 1, "Running, men". His figure for women is 1.08283. |
| 230 min | the outer edge of what he fitted | Riegel again — above this, he writes, the curve starts bending upward. |
You will see 1.06 in most online predictors. That is not Riegel's running figure: in his own table 1.05–1.06 is what he found for men aged 40 to 70. We used 1.06 too, until we read the paper. Using his running value makes this tool slightly more conservative, which given what follows is the right direction.
Someone checked it on people like us. Vickers and Vertosick (2016) ran it against 2,303 recreational runners: up to and including the half marathon it holds up well. At the marathon it does not — for half of those runners the prediction was at least ten minutes too fast, and for a quarter of them more than twenty. So: trust it to the half, treat the marathon line as a best case that assumes you have done the marathon training.
Want the next one when it lands?
One email when a new tool or write-up is finished, and not otherwise.
No spam, never shared, one click to leave.
Why a lab percentage is not a finishing-time percentage — read the shoe check.