The science library Intensity distribution

How hard, how often: intensity distribution

Polarized, pyramidal and threshold distributions, and where the 80/20 pattern came from. Two people can measure the same week and disagree about it; this covers why.

4 sources cited Jump to references ↓

Three zones, three shapes

Distribution work almost always uses a three-zone model: easy work below the first threshold, moderate work between the two thresholds, and hard work above the second ¹. The shape of a training week is just how the total is split across those three.

Polarized: a large easy base, a meaningful slice of genuinely hard work, and very little in the middle.
Pyramidal: a large easy base, a moderate middle, and a smaller hard tip, each zone larger than the one above it.
Threshold: a heavy concentration in the middle zone, with less at either end.

The labels describe an observed split, not a prescription. A single week can look pyramidal and the block it sits in can look polarized.

Where 80/20 comes from

Descriptive studies of trained and elite endurance athletes across several sports converge on a similar pattern: roughly 80% of sessions are performed at low intensity, with about 20% at high intensity and comparatively little deliberate time in the middle ¹.

That is an observation of what successful athletes do, at high training volumes, over long careers. It is not by itself evidence that the split causes the performance, and it comes from populations training far more than most people reading it.

The grey-zone argument

The critique of the middle zone is narrower than it usually gets repeated as. The claim is that moderate work carries a real autonomic and metabolic cost while delivering less aerobic stimulus per unit of fatigue than easy volume, and less of the high-intensity stimulus than genuinely hard work. A diet built around it accumulates fatigue that limits both ends ¹.

What the argument does not say is that tempo is useless, or that time between the thresholds should be zero. Race-specific work for long events lives there, and so does a large share of the training in pyramidal distributions used successfully at elite level.

What the trials actually compared

In a controlled comparison of four distributions in well-trained endurance athletes, the polarized group improved more on key endurance markers than threshold, high-intensity or high-volume groups over the same period ². Meta-analysis of the randomised trials points the same direction, favouring polarized over threshold distributions, though the pooled evidence rests on a small number of short studies ³.

The comparisons that exist mostly pit polarized against threshold-heavy training. Pyramidal distributions are far less often the comparator, so the evidence is thinner on the question people most often want answered. The honest position is that polarized and pyramidal have not been separated cleanly ³.

Sessions or minutes

A distribution is only defined once you say what is being counted. Counting sessions asks how many workouts had a hard intent; counting time asks how many minutes were spent where, either across whole sessions or strictly time-in-zone ¹. The three methods do not agree, and the gap widens the more interval work you do.

Where zone boundaries are unavailable or unreliable, session RPE multiplied by duration classifies sessions well enough to track the pattern over time ⁴. Pick one method and stay with it; a distribution that changes because the counting changed tells you nothing.

References

Numbered in order of first appearance. Every link goes to the original work.