Most training software carries a hidden opinion about you. It shows up as a table: a threshold that’s 95% of something, zones at fixed percentages of that, a rider category based on how your watts per kilo compare to a chart of other people. None of it is malicious and some of it is a decent starting guess. It’s just that the person the table describes has never actually been born.
The evidence is old and blunt
The HERITAGE Family Study put 481 sedentary adults through the same supervised 20 week endurance program. Same sessions, same supervision, same duration. The average improvement in VO2max was around 400 ml/min, which is a respectable result and a completely useless description of any individual in the study. The actual spread ran from roughly no change at all to over a liter per minute. Same program. An order of magnitude apart in outcome, with heritability estimates of around 47% for the response itself.
Mann, Lamberts and Lambert reviewed this territory in 2014 and made the point that keeps getting forgotten: the group mean is a fact about the group. Very few individuals are near it, and there is no rule saying you’re one of them.
Percentages of maximum don’t mean the same thing to two people
The sharpest version of this comes from Iannetta and colleagues, who tested 100 women and men and asked a simple question. If you prescribe training as a percentage of maximum, where does that actually land physiologically?
Lactate threshold showed up anywhere between 45% and 74% of VO2max depending on the person. Maximal lactate steady state ranged from 69% to 96%. In heart rate terms, threshold landed between 60% and 90% of maximum. So a session written as “70% of max heart rate” is an easy aerobic ride for one athlete and above threshold for another. The prescription is identical. The stimulus is not remotely.
Their conclusion was that fixed-percentage methods conform poorly to intensity domains and don’t control the metabolic stimulus. Which is a polite way of saying the numbers you were handed may have been describing a stranger.
What we built instead
Allure fits a three parameter critical power model to your own best efforts and works outward from there. Your threshold is your critical power, measured, not 95% of a test protocol. Your strengths and limiters are read as deviations from your own power-duration curve rather than from a category table, so the app tells you where you’re strong compared to the rest of your engine, not compared to a hypothetical balanced rider who doesn’t exist.
The framing throughout is you against your past self. Not you against the peloton, and not you against an average.
Where we still assume
Zero constants is not achievable and we’re not going to pretend otherwise. Some things in the app are still standard-issue.
Power and heart rate zone boundaries are currently fixed percentages of your threshold, the same multipliers everyone uses. The fitness and fatigue curves use the conventional time constants rather than ones fitted to your own response history. Effort, our per-session training stress number, is calibrated on the usual convention where an hour at threshold scores 100.
Each of those is a place where a population assumption is standing in for a measurement, and each is on the list. Zones derived from your own critical power come first, because that’s the one that changes what you do on Tuesday. The genuinely individualized version of form needs longitudinal data we’re still collecting.
We’d rather tell you which parts are measured and which parts are inherited than present the whole thing as personalized and hope you don’t ask. Your training history is yours, exportable whenever you want it, and the reasoning behind the numbers should be legible too.
Sources
- Bouchard C, An P, Rice T, et al. Familial aggregation of VO2max response to exercise training: results from the HERITAGE Family Study. J Appl Physiol. 1999;87(3):1003-1008.
- Mann TN, Lamberts RP, Lambert MI. High responders and low responders: factors associated with individual variation in response to standardized training. Sports Med. 2014;44(8):1113-1124.
- Iannetta D, Inglis EC, Mattu AT, et al. A critical evaluation of current methods for exercise prescription in women and men. Med Sci Sports Exerc. 2020;52(2):466-473.
- Morton RH. A 3-parameter critical power model. Ergonomics. 1996;39(4):611-619.