Training load

Training load is a single number for how much training a workout was. It puts a hill session, a long run, and a bike ride on one scale, so training of different kinds can be added together and compared week to week.

The number measures the physiological cost of a session rather than the work performed. Ten kilometres at threshold and ten kilometres easy are the same distance and different training, so distance alone cannot be added across sessions. Load is measured from the body's response instead, which is what makes different kinds of training comparable Halson, 2014.

Session load#

Every completed workout carries a session load: one number for how hard the session was and how long it lasted.

It is calculated from your heart rate, moment by moment. Each moment is priced by where your heart rate sits between your resting and your maximum, your heart rate reserve Karvonen et al., 1957, so the scale means the same thing for you as for a runner whose maximum is twenty beats higher.

The scale is anchored so that an hour of easy running is about 60. Intensity counts as well as time, which is why a 45-minute interval session can carry more load than a 90-minute easy run.

Hover the Load value on the workout page and two Training Effect scores sit beside it, each on a 0 to 5.9 scale:

The two load models#

Two models convert heart rate into load. Settings → Measurement → Training load method selects the one in use. See Athlete settings.

Both are calibrated to the same anchor, so changing method rescales the whole history rather than creating a step in it. The exponential weighting uses different coefficients for men and women, which is why gender is part of the athlete profile.

EPOC produces steadier values across sessions of different shapes. The exponential method separates high-intensity work from easy work more sharply. Neither is more accurate than the other, and the acute, chronic, and fitness curves assume the whole history was calculated the same way, so the method is best changed rarely.

Load for a workout you logged by hand#

A workout you typed in has no heart rate trace, so its load is estimated. Runtool uses the best input you gave it:

RPE is your rating of perceived exertion, how hard the session felt on a 1 to 10 scale Borg, 1982. A rated effort multiplied by its duration is an established training-load measure in its own right Foster et al., 2001, which is why the RPE estimate lands on the same scale as the heart rate models.

Your rating is priced at the zone it names. An RPE of 7 is threshold work and costs what an hour of Z4 costs. The full table, and the research behind it, is on the Intensity page.

Every estimate is labelled. Hover the Load value on the workout page or the calendar and it names which of the three produced the number: "From your zones", "From RPE", or "From duration".

The three land on the same scale, so a logged run and a synced one are comparable, and connecting a device later will not make your load jump.

Two things improve an estimate, in this order:

1. Set your pace zones. With zones, the distance and time you entered price the run directly. This is the largest single improvement.
2. Add an RPE. Without zones, the effort rating is what separates a hard hour from an easy one.

Projected load#

A workout you have not done yet carries a projected load: the same number, on the same scale, worked out from the workout's structure and your zones. It shows the load of a week before you run it, and it feeds the forward part of your ACWR.

Projected load is a forecast, and it is replaced by the measured value once you complete the workout. Where there is less information to work from, it estimates low rather than high.

What it is based on depends on the workout:

Some workouts show no projected load, and the workout states why when you hover it:

• Runs, when you have no pace zones set. A run's load is priced from your zones. Without them there is no basis for the estimate, so no value is shown. Set your zones and it fills in.
• Workouts created from an imported activity. These already happened, so they carry their measured load. There is nothing to forecast.

Acute and chronic load#

Session loads roll up into two running values:

• Acute load: recent training, weighted towards the last few days. Your current stress, over a 7-day reference window.
• Chronic load: the baseline built over weeks. What you are adapted to, over a 28-day window.

Both are exponentially weighted rather than flat averages, so a session's contribution fades day by day instead of dropping out of the window all at once Williams et al., 2017. A rest day counts as a rest day, not as a gap in the data.

Fitness is built by raising acute load above chronic load repeatedly, with recovery between. Both curves are in the load widget on the Overview.

ACWR#

The acute:chronic workload ratio (ACWR) divides acute load by chronic load. Around 1.0, recent training matches what you are adapted to. Values well above it mean training is increasing faster than the baseline it sits on. The bands, the projected part of the line, and what the ratio does not measure are on the ACWR page.

Fitness, fatigue, and form#

The fitness lens (beta) reads the same load history three ways:

• Fitness: the slow-moving product of chronic training, over a 42-day window.
• Fatigue: the fast-moving cost of recent work, over 7 days.
• Form: fitness minus fatigue, your readiness to perform right now.

This is the impulse-response model: each session leaves both a fitness effect and a fatigue effect, and the two decay at different rates (Morton et al., 1990; Clarke & Skiba, 2013). Fatigue rises faster than fitness and also fades faster, which is what a taper uses: over a taper fitness is largely retained while fatigue falls, and form rises.

Form is expressed as a percentage of your own fitness, so the bands mean the same thing at any load level.

Limitations#

Every figure on this page comes from a model. The limits are:

• The scale is individual. It is derived from your heart rate reserve and your zones, so a load of 400 in your week and 400 in another athlete's week are not the same training. Load is comparable for one athlete over time, not between athletes.
• Heart rate reflects the day, not only the session. Heat, altitude, illness, and short sleep raise heart rate at a given pace, so an easy run in those conditions is priced higher than the same run in normal conditions.
• Sessions without heart rate are priced approximately. Strength and cross-training are estimated from duration and category, so they contribute to the totals with less precision than a recorded run.
• Training Effect labels come from group data. They are population findings and are not calibrated to an individual. The same applies to the ACWR bands.

Where load shows up#

• Workout page: the session or projected load, the model that produced it, and its Training Effect scores.
• Calendar: the load of each workout, and the week's total in the week header.
• Overview: the acute and chronic curves, the ACWR, and the fitness lens.
• Analyse: total training load over longer spans, across every sport.

FAQ#

Does load count non-running training?

Yes. Strength and cross-training workouts carry load like any session, so your chronic load reflects everything you actually do.

Can I compare my training load with another runner's?

No. The scale is derived from your heart rate reserve and your zones, so the same number represents different training for different athletes. Load is comparable for one athlete over time.

Which load method should I choose?

Either. EPOC produces steadier values across sessions of different shapes; the exponential method draws a sharper line between hard and easy days. The more important point is to stay with one, because the acute, chronic, and fitness curves assume the whole history was calculated the same way.

What are the TE and An numbers next to my load?

Training Effect, on a 0 to 5.9 scale. TE is the aerobic stimulus of the session; An is how much of the work happened above your LT2. A steady long run scores high on TE and near zero on An; short hard reps raise An.

Why did a workout's load change after I edited my zones?

Load is calculated relative to your zones. Reprocessing a workout against updated zones re-grades its intensity, so the value changes. The recorded activity itself is unchanged.

References#

• Borg, G.A.V. (1982). Psychophysical bases of perceived exertion. Medicine & Science in Sports & Exercise 14(5), 377–381. 10.1249/00005768-198205000-00012
• Børsheim, E. & Bahr, R. (2003). Effect of exercise intensity, duration and mode on post-exercise oxygen consumption. Sports Medicine 33(14), 1037–1060. 10.2165/00007256-200333140-00002
• Clarke, D.C. & Skiba, P.F. (2013). Rationale and resources for teaching the mathematical modeling of athletic training and performance. Advances in Physiology Education 37(2), 134–152. 10.1152/advan.00078.2011
• Foster, C., Florhaug, J.A., Franklin, J., Gottschall, L., Hrovatin, L.A., Parker, S., Doleshal, P. & Dodge, C. (2001). A new approach to monitoring exercise training. Journal of Strength and Conditioning Research 15(1), 109–115. 10.1519/00124278-200102000-00019
• Halson, S.L. (2014). Monitoring training load to understand fatigue in athletes. Sports Medicine 44(S2), 139–147. 10.1007/s40279-014-0253-z
• Karvonen, M.J., Kentala, E. & Mustala, O. (1957). The effects of training on heart rate: a longitudinal study. Annales Medicinae Experimentalis et Biologiae Fenniae 35(3), 307–315. PMID 13470504
• Morton, R.H., Fitz-Clarke, J.R. & Banister, E.W. (1990). Modeling human performance in running. Journal of Applied Physiology 69(3), 1171–1177. 10.1152/jappl.1990.69.3.1171
• Williams, S., West, S., Cross, M.J. & Stokes, K.A. (2017). Better way to determine the acute:chronic workload ratio? British Journal of Sports Medicine 51(3), 209–210. 10.1136/bjsports-2016-096589