Two hours on the bike are not all equal. Two easy hours along the valley, two hours of intervals, two hours up a mountain pass: same duration, very different legs the next morning. Training load is the number that tries to say what an activity actually cost you, by combining how long it lasted with how hard it was.
Everything else is built on it. Your fitness, your fatigue and your freshness are all computed from it, day after day (Fitness, fatigue, freshness).
Duration × intensity, on a single scale
The idea goes back to Eric Banister in the 1970s: sum up each session as one “training impulse”, the TRIMP, which multiplies duration by a measure of intensity. Every method in use today descends from it.
What makes them useful is their scale. The one cycling settled on comes from Andrew Coggan: one hour at your threshold scores 100. Your threshold is your FTP, the highest power you can hold for about an hour (What FTP is, and how to test it). An easier or shorter activity scores less, a longer or harder one more. And because everything is measured against your threshold, the load tells you what the activity cost you, not an average rider.
In Komeak, it is simply your load. Elsewhere, you’ll find it under other names, TSS® among them.
With a power meter
This is the most direct measure: the power meter records what you produced, second by second. The calculation takes three steps.
1. Equivalent power. On rolling terrain or during intervals, average power lies: freewheeling descents and stops drag it down, while every surge costs a lot. Equivalent power — NP® elsewhere — gives hard moments more weight than easy ones. Above an easy pace, every extra watt costs your body more and more: a surge of a few seconds weighs far more than its length suggests, and a minute of freewheeling doesn’t pay back the climb before it. So it gives the steady power that would have cost as much as your actual effort. For a steady ride it sits close to the average; for a stop-and-go one, well above it. Why it is computed this way, and what variability says: Equivalent power: why average power lies.
2. Intensity, as a share of FTP: equivalent power divided by FTP, known elsewhere as IF®. At 100%, it is the intensity of an hour at threshold.
3. Load: intensity² × duration in hours × 100.
One hour at 100% of FTP does score 100. Two hours at 75% score 0.75 × 0.75 × 2 × 100 ≈ 113. Intensity is squared: riding a bit harder costs a lot more — your legs knew that already.
With a heart-rate strap only
Without a power meter, heart rate tells the intensity. Banister’s method weighs each minute by the share of your heart-rate reserve it used, between your resting and your maximum heart rate. That weight grows exponentially with intensity, as blood lactate does. To speak the same language as power, the result is brought onto the same scale: one hour at your threshold heart rate scores 100.
It needs three markers: resting heart rate, threshold heart rate and maximum heart rate. Its limits are well known:
- heart rate lags: a fifteen-second sprint is over before it has risen;
- it drifts with heat, dehydration and fatigue, for the same effort.
It remains a good measure of steady efforts: Measuring training load with heart rate goes into the method and its limits.
With no sensor at all: how it felt
After the session, a rating from 1 to 10: the session RPE method, proposed by Carl Foster in 2001. It uses Borg’s CR-10 scale, whose labels keep ratings comparable from one session to the next: 1 “very, very easy”, 3 “moderate”, 5 “hard”, 7 “very hard”, 10 “maximal”. The rating covers the whole session, once it is over.
Foster compared this rating, multiplied by duration, with a load computed from heart rate: on the bike, the two moved together from one session to the next. A two-second rating is enough to follow your training when you have no sensor (Perceived effort).
Perceived effort is worth giving even with a sensor. Power measures what you produced; perceived effort, what it cost you. An ordinary activity that felt very hard says a lot about how you are doing that day.
Which method, how precise?
| Method | What it measures | Its limit |
|---|---|---|
| Power | What you produced | Needs an accurate FTP |
| Heart rate | How your heart responded | Lags on short efforts, drifts with heat and fatigue |
| Perceived effort | What the effort cost you | Subjective, personal to you |
| Duration only | Nothing of intensity | An estimate, until something better comes |
Because they all share the same scale, their loads add up: a week of activities with and without sensors keeps a consistent load. What differs is how precise each one is.
What load doesn’t tell you
- What kind of effort it was. Two hours of endurance and one hour of intervals can carry the same load, and have nothing like the same effect.
- What happens off the bike. Sleep, work, an illness coming on: load doesn’t see them, your perceived effort does.
- A wrong threshold. Everything is measured against your FTP or threshold heart rate. With an FTP 10% too low, every intensity is overestimated by as much, and every load by about 23% (1 ÷ 0.9², squared). An up-to-date threshold matters more than anything else.
In Komeak
Every activity you add gets its load, computed from the best measure Komeak has: power, if the file holds it and your FTP is known; otherwise heart rate, if your heart-rate thresholds are; otherwise your rating; otherwise, until you rate it, an estimate from its duration.
- The method and its precision are shown on the activity’s page, on four levels: excellent for power, good for heart rate, fair for perceived effort, low for the estimate from duration.
- A sensor that measured too little is set aside: below half of the moving time, the next method takes over.
- What’s missing is said. An activity that holds your power while your FTP is unknown says so, and offers to let you enter it. The same goes for your heart-rate thresholds.
- Rating takes two seconds, from 1 to 10 with Borg’s labels. Komeak asks for it on the past week’s activities; the activity’s menu lets you give or change it at any time. With a sensor, the Komeak analysis compares it with what you usually rate for that intensity and duration, and says whether you rated it harder, or easier, than usual — without venturing a cause.
- Your methods and thresholds are set in Profile, Cycling: you say whether you ride with a power meter, with a heart-rate strap, and enter your thresholds there.
The load from perceived effort follows the same scale as the others: an hour rated 8, between “very hard” and “maximal”, is worth about an hour at threshold.
Sources
- Allen H., Coggan A., McGregor S., Training and Racing with a Power Meter, 3rd ed., VeloPress, 2019.
- Banister E. W., “Modeling elite athletic performance”, in MacDougall J. D., Wenger H. A., Green H. J. (eds), Physiological Testing of the High-Performance Athlete, Human Kinetics, 1991.
- Morton R. H., Fitz-Clarke J. R., Banister E. W., “Modeling human performance in running”, Journal of Applied Physiology, 69(3), 1990, pp. 1171–1177.
- Foster C. et al., “A new approach to monitoring exercise training”, Journal of Strength and Conditioning Research, 15(1), 2001, pp. 109–115.
- Borg G., Borg’s Perceived Exertion and Pain Scales, Human Kinetics, 1998.