Race Time Prediction
Estimate how long it will take to run a specified distance, based on two previous race results.
This type of predictive analysis is pretty accurate for distances from 800 metres to the marathon. It can also be used in swimming or other sports.
Enter the distance of a previous race in the "Distance Run 1" category.
Then enter the time for that race in the input boxes on the right. Leave all "Pace" fields blank.
Repeat the same for "Distance Run 2". Then enter a third distance in the "New Distance" category and press "Calculate".
Metric: Enter all distances in km or miles. Pace will be calculated in minutes/km or minutes/mile, respectively.
The algorithm
Pete Riegel developed a formula in 1977 that states that
t2 = t1 * (d2 / d1)1.06
where t1 equals the initial time, d1 equals the initial distance, d2 equals the new distance being calculated for,
and t2 equals the predicted time for the new distance. The exponent 1.06 means that, for n times the distance, most people need n1.06 times the time.
However, the exponent 1.06 doesn't fit every runner. In reality, it depends on the type of the runner. 1.06 seems to be a good value for 5000m runners. Ultra-distance runners often have a decay exponent of 1.04 or even 1.03.
At the other side of the spectrum, sprinters and decathletes have exponents of up to 1.30. A high exponent means that you slow down rapidly with increasing distance. My own decay exponent is about 1.18, which might mean that I'm approximately a 200m/400m runner.
So, in order to get a better estimate, I use a two-step approach:
- By using two initial distances d1 and d2, calculate the decay exponent:
decay = log (t2 / t1) / log (d2 / d1).
- Then, calculate the time for the third distance as t3 = t2 * (d3 / d2)decay.
| Runner | Decay exponent |
| Sabrina Mockenhaupt | 1.05 |
| Kenenisa Bekele | 1.06 |
| Michael Johnson | 1.16 |
Back to Index