The calculator shows a comparison across all ages and a separate age-adjusted comparison. It applies the selected event and sex's reference age factor to your time before finding the percentile. The highlighted row uses the same reference factor.
Reference factors use the nearest supported whole-year age and event; fractional ages do not create a new fitted population. Where an age is outside the references, the table's note identifies the limit. Some performances also fall outside the supported ranking range.
An age-adjusted percentile compares your performance in this model. It differs from an age-grade percentage against a reference best performance. This calculator does not supply a separate formal age-grade score.
No. For the same sex, age, distance and time, 50 kg and 100 kg give the same level in this version. It is not a bodyweight-class ranking.
There is no height adjustment in these standards. The selected comparison data do not provide a validated height-based ranking; adding a height correction would imply evidence this version does not have.
Oxygen cost and effective VO2 max are expressed per kilogram. Bodyweight alone does not change those estimates or the training paces.
Hills, surface, weather and treadmill conditions affect run times. There is no elevation or treadmill conversion here. Submissions do not establish identical course and start protocols; record dialogs identify the reference event and surface where known.
The submission-based standards are not a verified flat-course or treadmill cohort. Their course conditions remain unknown. The speed-duration oxygen estimate does not establish that different surfaces or gradients give equivalent rankings.
They show times at the same percentile of this version's model over other distances. Different events can represent different groups of athletes. These comparisons do not guarantee every time without event-specific preparation.
The curve estimates relative frequency from the selected model, rather than counting observed results in time bins. Faster times are to the right. Percentages describe how many performances are slower within this comparison.
Your marker keeps the submitted time and the result's age-adjusted percentile. The curve uses the same age reference as your standards row. Charts normally display the 1st to 99th percentiles and extend to include your result when it falls outside that window. That display window does not limit your rating.
Interpolated distances use neighbouring event models. Running ultra estimates extend the marathon relationship; longer swimming estimates extend the shorter-distance fit. These are not independent observed distributions at every distance.
Running Level uses a speed-duration estimate independently of the standards percentile. Speed gives estimated oxygen cost; dividing it by the estimated sustainable fraction for your finish time gives effective VO2 max.
For speed v in metres per minute and time t in minutes: cost = −4.60 + 0.182258v + 0.000104v²; fraction = 0.8 + 0.1894393e−0.012778t + 0.2989558e−0.1932605t. Effective VO2 max = cost ÷ fraction, in ml/kg/min.
This follows the performance approach associated with Jack Daniels' VDOT system. VDOT's explanation distinguishes running fitness from measured VO2 max: economy and other factors affect performance. Sprint and extrapolated ultra results do not establish laboratory VO2 max.
Training-pace estimates are shown only from 800m through the marathon, using the submitted performance rather than an age-adjusted time. Easy pace is slower than race pace in this displayed range. Use a recent best effort and adjust training to conditions; the table is not a complete programme.
| Beginner | Faster than 5% of runners. |
|---|---|
| Novice | Faster than 20% of runners. |
| Intermediate | Faster than 50% of runners. |
| Advanced | Faster than 80% of runners. |
| Elite | Faster than 95% of runners. |
Average is the Intermediate standard (50th percentile) for men and women across all ages. The age tables show the same levels for each age.
Levels describe this event's model comparison, not your training history or ability in every sport. Elite is a population percentile, not a professional qualification. Use a recent best test or race effort for a meaningful benchmark; standards are not a pace target for every workout.
Newer standards use self-reported Running Level submissions. They describe that comparison, not all people or a verified race-entry population. One person can submit multiple performances; approximate identity checks do not establish a count of verified athletes.
The selected version fixes the data and model used in your result. New data can lead to a later version. Sources identify inputs and methods without implying endorsement.
New releases follow reviewed data quality and model checks, with no fixed update interval. We do not have a verified geographic breakdown of submissions. The standards pages give the available counts and source limitations.
Available record figures open the corresponding event, athlete, age/category, performance date and source. Missing records stay unavailable. Some sources list best performances rather than a ratified record register; check categories and source selection before treating every entry as a ratified world record.
A fitted percentile can extend beyond an observed record. It estimates the comparison population and is not evidence that a record has been achieved or could be achieved under that record's rules. Age adjustments, equipment and event categories can also differ; use the record's details when comparing them.
Record sources include World Athletics, Tom Bernhard's 2025 tables, IAU and Wikipedia. Age rows show the fastest available performance in each five-year age group.
Older versions can display calculated record benchmarks.
StuWeb. The original running model was calibrated against 10 km and half-marathon performances published by StuWeb, with age adjustments and separate male and female comparisons.
RunScore age-grading tables. Age adjustments and reference benchmarks use the 2020 RunScore age-grading tables.
The ability standards are Running Level model estimates. Attribution identifies the sources and does not imply their endorsement of these standards.
Detailed submission and fitted-result counts are unavailable for this older version; current counts must not be attributed to it.