FTP vs critical power
What each number means
A best-effort point is an observation from a recording. An FTP estimate depends on a testing or estimation method. Critical power is a parameter in a power-duration model. Similar-looking watt values do not make them the same measurement. Knowing which one a training tool uses helps you interpret its targets and compare results using the same method.
In a study of twelve competitive male cyclists, the average FTP and critical-power values were similar, but individual agreement was too wide for the authors to consider them interchangeable. That small sample is useful evidence against treating the labels as synonyms, not a universal conversion table.
Chosen parameters in P = CP + W′ / t. Dashed: CP. No FTP is inferred and this curve is not fitted to an athlete.
What each number means
A best-effort point is an observation from a recording. An FTP estimate depends on a testing or estimation method. Critical power is a parameter in a power-duration model. Similar-looking watt values do not make them the same measurement. Knowing which one a training tool uses helps you interpret its targets and compare results using the same method.
In a study of twelve competitive male cyclists, the average FTP and critical-power values were similar, but individual agreement was too wide for the authors to consider them interchangeable. That small sample is useful evidence against treating the labels as synonyms, not a universal conversion table.
How the critical power model works
Our experiment uses the two-parameter equation P = CP + W′ / t, with time in seconds and W′ in joules. Change CP and the curve shifts vertically. Change W′ and the short-duration end moves more strongly.
The controls are chosen parameters, not fitted results from your body. We show the model over two to twenty minutes. Extending this simple equation towards zero seconds would predict absurdly high power; its form is not valid over every duration.
What a 20-minute test tells you
A 20-minute test can supply a repeatable field benchmark when its protocol is controlled. Multiplying it by a fixed factor is an estimation convention, not a guarantee of what every rider can sustain for an hour.
A separate study of seventeen trained cyclists and triathletes compared a multi-duration CP protocol with a 20-minute FTP test. The protocols and population matter when interpreting any reported relationship.
Comparing your results over time
Keep the method, device and conditions consistent when tracking your own results. Record whether a value was tested, inferred by software or manually entered. Do not silently substitute one for another.
Use the model to understand the shape of an estimate. Use the underlying efforts to decide whether that estimate has good evidence. This article does not prescribe a maximal testing protocol.
What critical power tries to describe
Critical power comes from the relationship between hard exercise power and the time it can be sustained. In the two-parameter model, work above CP is represented by W′. At a constant power above CP, the model predicts a shorter tolerable duration as power increases.
CP is associated with a transition in the behaviour of oxygen uptake and other physiological responses: above it, a steady metabolic state generally cannot be maintained in the same way. It is an estimated boundary from a model and protocol, not a direct measurement of one isolated system or a guarantee of an exact time to exhaustion.
What W′ means—and what it does not
W′ is measured in joules and represents the curvature of the fitted work–duration relationship. In the simplest constant-power model, a larger W′ increases the predicted amount of work above CP before the limit is reached. It helps explain why two athletes with the same modelled CP can have different short-duration predictions.
It is tempting to call W′ a battery, but that picture has limits. Fatigue involves multiple processes, and a simple two-parameter curve does not fully describe recovery between surges, sprinting or very long events. The result depends on which efforts were used for the fit.
Make a useful comparison over time
If you track an estimate, keep the protocol, meter and conditions as consistent as practical. A change in testing method or a new algorithm can change the number without the same change in ability. Keep the underlying efforts, not only the resulting threshold.
FTP may be assessed or estimated in several ways; a fixed percentage of twenty-minute power is one convention. CP is obtained from a power-duration model. They can be numerically close for one person and different for another, so use the parameter your model or training zones actually require.