Building Blocks · Training · Chapter 4 of 11 · 3 min read

Load: scoring a ride

To plan training you need to measure it. Hours are a poor measure: an hour of easy spinning and an hour at threshold are different things. Kilometres are worse indoors. BlockRide gives every ride a load, one number for how much training it was.

The formula

Load counts time, weighted by how hard each second was. Hard counts more than double:

Load of one second ∝ (Power ÷ CP)²Added up over the ride, and scaled so that one hour at threshold scores 100.

Squaring is what makes intensity expensive. An hour at 70% of threshold scores 0.7 × 0.7 × 100 = 49. Half an hour at threshold scores 50. Twenty minutes in total at 120%, split into short efforts, scores 48. Three very different rides, about the same dose. This is the same idea as the Training Stress Score that Andrew Coggan introduced for power meters, and a relative of Banister's "training impulse", which weights heart rate in a similar way. The scale of 100 for an hour at threshold is the common convention, so a BlockRide load reads much like the numbers from other training tools.

BlockRide scores each second on its own power and adds them up, instead of scoring the ride's average. A ride with surges therefore scores more than a steady ride at the same average, which matches how it feels.

Three tanks

One number still hides something. A load of 60 from a long endurance ride and a load of 60 from sprints train different systems. So each second's load goes into one of three tanks, by the power it was ridden at:

  • Cruise: up to threshold. Endurance, tempo, sweet spot.
  • Push: above threshold. The efforts that drain the battery.
  • Burst: close to your peak power. Sprints and attacks.

The split follows a long line of research. Stephen Seiler's studies of how successful endurance athletes train use three intensity zones divided at the two physiological thresholds, and find that most of their training time sits in the lowest one, with a smaller, deliberate share of hard work. Trials such as Stöggl and Sperlich's show that the mix matters, not only the total. Tanks let BlockRide plan and check the mix.

In the game the three tanks in the battery tile fill as you ride. With a plan, each tank has today's target: so much Cruise, so much Push, so much Burst.

Rides without a power meter, and other sports

Not everything you do is a BlockRide ride. Sessions can come in from Strava, Garmin or a FIT file, and they count too:

  • With power, load is worked out exactly as above.
  • Without power, and if you have allowed health data, load is estimated from heart rate.
  • With neither, it is estimated from the time and the sport.

A run or a ski session makes you tired in full, but only part of it carries over to cycling fitness. BlockRide counts other sports at one share towards fitness and a larger share towards fatigue. The shares are part of BlockRide's tuning.

What load is not

Load measures the dose, not the effect. Two riders can do the same load and respond differently, and the same rider responds differently to the same load at different times. Thierry Busso's modelling work showed exactly that. This is why BlockRide never treats load as a score to maximise. It is the input for the next chapter, where doses become fitness and fatigue, and for the plan, which decides how much is enough.

Garmin Connect and Body Battery are trademarks of Garmin Ltd. or its subsidiaries.