How a cycling power curve is calculated
What a power curve shows
A power trace answers a question about time: what were you producing at this moment in the ride? A power curve asks a different question: what was your highest average power over a particular duration? The same recording can answer both.
Scroll on. The chart below pins to the screen and builds the curve as you go.
The ride
A 30-minute ride, recorded once per second. Scroll to start.
Best 1 s: 898.0 W from 13:15 to 13:16.
What a power curve shows
A power trace answers a question about time: what were you producing at this moment in the ride? A power curve asks a different question: what was your highest average power over a particular duration? The same recording can answer both.
Finding your best five minutes
Choose five minutes. The highlighted window now contains 300 consecutive one-second samples. Add their watts and divide by 300. Move it forward by one second, calculate again, and repeat until its right edge reaches the end of the ride.
The largest of those averages becomes the five-minute result. It is not the five highest minutes collected from different parts of the session. A recovery inside the winning window remains part of its average. Try moving the window away from the best effort and watch the number change.
Plotting the result
The winning average moves from its position in the ride to its duration on the second chart. Watts still run vertically. The horizontal axis now measures effort length, not clock time.
Repeat the search at five seconds, one minute and twenty minutes. Each duration can have a different winning start. The curve joins these independently calculated best averages. Our plot uses a logarithmic duration axis so both seconds and minutes remain visible. Equal horizontal distances represent equal ratios of duration.
Finding the effort behind each point
The full curve joins the best averages for each duration. Select a duration and its source interval appears on the ride trace. This connection is the most useful way to interrogate an unusual point: inspect the recording that produced it.
Switch on the added sprint. Short-duration results respond strongly, while longer windows dilute those extra seconds across a much larger denominator. They can still change: a sprint contributes to every longer interval containing it. Switch it off to reverse the experiment.
What your power curve can tell you
Your power curve helps you see what you have actually achieved across different effort lengths. Look at short efforts to understand your sprint, a few minutes for a hard climb, or longer efforts for sustained riding. Use the duration that resembles your event or goal to choose a realistic starting point for pacing and training.
Compare similar periods to see where your best efforts have improved and where you might focus next. For example, stronger five-minute power may help on a short climb; a higher 20-minute best can help you judge a longer effort. These are records of what you did, not guarantees of what you can do today.
The curve only reflects the rides and efforts included. A quiet month with no hard five-minute efforts can produce a lower record without any loss of fitness. Different points on a season curve may also come from different rides, so the whole curve is not one continuous effort.
How to check your own data
Pick a duration relevant to the effort you actually attempted. Open its winning interval and check whether the start, finish and data look plausible. Compare equivalent periods and devices before interpreting a small change.
The experiment calculates from the full one-second data; only the visual trace is simplified. Zero watts are included. There are no missing samples or paused periods in this example. Download the CSV to repeat the calculation in a spreadsheet.
Why the best power changes with duration
A five-second sprint and a twenty-minute effort ask different things of the body. Very short efforts can draw rapidly on limited stored energy and produce high power; as the effort continues, the rate at which energy can be supplied and fatigue develops constrain what can be sustained. Aerobic and anaerobic processes contribute together, not as switches that turn on at particular seconds.
That is why a power-duration curve usually slopes down: the longer the averaging window, the harder it is to maintain the high power reached briefly. The exact curve reflects both physiology and what you actually attempted. It cannot identify a muscle fibre type or energy-system contribution from a single point.
Use it to choose a pacing target
Imagine a climb that normally takes you around eight minutes. Find your best comparable eight-minute efforts, then inspect when and where they happened. A recent, well-recorded climb is a more relevant starting point than extrapolating from a five-second sprint.
A record is an upper observation from that day, not a target you must hold on every climb. Wind, altitude, temperature, accumulated fatigue and whether you need to ride afterwards all affect a sensible pacing decision. Start conservatively when the cost of fading is high.
When a surprising record needs investigation
A one-second dropout or implausible power spike can alter short-duration bests; a gap or paused period can affect longer windows. Check the original trace, timestamps and meter before treating a sudden jump as an improvement.
For season comparisons, use similar date ranges and ask whether both periods included hard attempts at the duration of interest. A curve built from easy commuting and one built from races are different samples of your riding, even if they cover the same number of weeks.