One result of yours, and the other distances follow from it. Two results, and the textbook exponent of 1.06 gives way to your own — what doubling the distance costs you personally. The marathon, and anything that leaves the range the formula was tested on, is flagged right in the table. It all runs in your browser: the numbers you type go nowhere.
Miles, decimals are fine: 7.5
H:MM:SS or MM:SS. Bare digits work too: 5000 → 50:00
Second result — for a personal exponent
It has to be a different distance: one point gives you no slope. The farther it sits from the first, the steadier the exponent
Take races from one season: an old result describes a different shape
The fields hold an example — put your own numbers in. The moment you touch the first result, the example in the second clears itself: your exponent cannot be worked out from our race.
| Distance | Prediction | Pace |
|---|---|---|
| 1500 m | 06:42 | 7:11 /mi |
| 1 mile | 07:13 | 7:13 /mi |
| 3 km | 13:57 | 7:29 /mi |
| 5 km | 23:59 | 7:43 /mi |
| 10 km · Your result | 50:00 | 8:03 /mi |
| 15 km | 1:16:51 | 8:15 /mi |
| Half marathon | 1:50:19 | 8:25 /mi |
| Marathon | 3:50:01 † optimistic | 8:46 /mi |
† optimistic — the marathon row, worked out from a shorter result. In a study of 2303 recreational runners, that prediction came out 10 minutes or more too fast for about half of them. It is a reference point, not a plan.
✱ outside the range — the time leaves the range the formula was tested on: efforts of roughly 3.5 to 230 minutes. Past those limits it simply has not been checked.
Adds a row of its own to the table and puts it into the prediction above: 10, 20, 31 — whatever you are training for.
The things no formula will take — how much you ran this week and whether you recovered — Fizkult sees on its own, with nothing to type in.
The formula this page stands on looks like this: time₂ = time₁ × (distance₂ / distance₁)k. Peter Riegel derived the exponent 1.06 from world records, and his 1981 paper gives the range he tested it on: efforts of roughly 3.5 to 230 minutes. Neither a sprint nor a 24-hour run is described by it. And the main thing is where the number itself came from: records at different distances are set by different people, each at their own best one. In applying a single exponent to a single person, we assume that person is equally prepared at every distance. A living runner has an exponent of their own — and that is what the two-result mode works out.
At the marathon the formula runs fast, and that has been measured. Vickers and Vertosick collected the results of 2303 recreational runners: up to the half marathon Riegel's formula matched the facts reasonably well, while for the marathon it predicted too fast — by 10 minutes or more for about half of them. The mean squared error of their own models was 40–45 % lower, and the best of them looked at two past races and at weekly mileage. Our two-result mode is the same move without the mileage, which means it comes with a caveat: the races are in it, the mileage is not. Mileage and recovery are not things you type into a calculator — they are visible to whoever is watching you every day.
How do I work out a marathon time from a 10K result?
The arithmetic goes like this: the time at the longer distance equals the time at the shorter one multiplied by the ratio of the distances raised to the power of 1.06. A 10K of 50:00 gives a marathon of 3:50:01 and a half of 1:50:19. The Fizkult race time predictor does this for you and flags the marathon row on the spot: in a study of 2303 recreational runners, that prediction came out 10 minutes or more too fast for about half of them. Read the marathon row as an optimistic reference point, not as a plan. To bring the prediction closer to you, add a second result of your own at another distance — then the exponent is worked out from you instead of being taken from a table of records.
What is the endurance exponent and why is it 1.06?
It is the power in the formula — the cost of doubling the distance. At an exponent of 1.00 the pace would not change at all: double the distance, exactly double the time. A value of 1.06 means every doubling adds about 4.2 % to the pace. The number comes from Peter Riegel's 1981 paper, where he derived it from world records. Records at different distances are set by different people, each at their own best one — and in applying 1.06 to a single person, we assume that person is equally prepared at all of them. A living runner has an exponent of their own, and the Fizkult race time predictor works it out from two of your results.
Why does Riegel's formula overpredict the marathon?
The formula works from time and distance, and there is nothing else in it — no mileage, no recovery. In 2016 Vickers and Vertosick collected the results of 2303 recreational runners and tested Riegel's formula on them: up to the half marathon it matched the facts reasonably well, while for the marathon it predicted too fast — by 10 minutes or more for about half of them. Why the gap shows up at the marathon in particular, the authors do not examine. That is why the Fizkult race time predictor flags the marathon row separately instead of serving it up alongside the rest.
Can I predict a half marathon from a 5K if I have never run one?
You can work it out, and the calculator will: up to the half marathon the formula matched the facts reasonably well in that same study. But it answers one question — what happens if you are as ready for the longer distance as you are for the shorter one. The farther the prediction travels from what you have actually run, the more of that assumption is in it.
What makes the prediction more accurate?
Three things. First, a second result: two of your own races at different distances give a personal endurance exponent instead of a textbook one — the exponent becomes yours rather than everybody's. What a second race on its own does without mileage, the authors did not test: in their best model, weekly mileage stood next to the two races. Second, freshness: an old result describes the runner you were, not the one you are. Third, similar conditions — heat, wind and elevation gain can move the time more than the difference between exponents of 1.05 and 1.07 does (at the marathon off a 10K that is about 3 %, nearly seven minutes). Everything outside the arithmetic — how much you ran this week, how you slept, whether you recovered — is invisible to the calculator, and that part is a coach's job.
1. Riegel P. Athletic Records and Human Endurance. American Scientist, 1981, 69(3), 285–290. The exponent 1.06 and the range it was tested on — efforts of roughly 3.5 to 230 minutes — come from this paper.
2. Vickers A. J., Vertosick E. A. An empirical study of race times in recreational endurance runners. BMC Sports Science, Medicine and Rehabilitation, 2016, 8:26. DOI 10.1186/s13102-016-0052-y. Full text: pmc.ncbi.nlm.nih.gov/articles/PMC5000509. The 2303 runners, the formula's fit up to the half marathon, the "10 minutes or more for about half of runners" and the models' mean squared errors — 227.6 and 208.3 against Riegel's 380.7 — all come from there.
In that same study, the best prediction came from past races paired with weekly mileage. The races you have just typed in by hand. The mileage you will not have to: Fizkult adds it up itself from your workouts — the week's distance per sport sits in your weekly review.
From there it reads what the watch records — sleep, heart rate, load — and holds it against your own baseline. After a session a breakdown arrives: what that pace cost you in heart rate, and what follows from it for the next few days. Every morning, a short letter: sleep shorter than usual, yesterday was hard work, here is what to watch today.
The watch connects once. First week free, everything included.
Connect your watch in Telegram →This is a coach's read, not a medical one. If you are not recovering, see a doctor.
Pace calculator — to spread the time you have just found across the distance. Heart rate zone calculator — what that pace costs in heart rate. BMI calculator — and the second measure muscle doesn't break.
← Home