Training taper

Why does resting before a competition make an athlete stronger than training through it?

Two weeks before the race, a swimmer who has been covering huge distances all season cuts her training roughly in half. Nothing is added. She simply does less — and swims faster than she has all year.

This sits badly with the obvious model, in which training is deposits and fitness is the balance. On that model, doing less can only mean having less. Yet if it were true, the right taper would be no taper, and every coach in every endurance sport would have found that out long ago. So the obvious model must be missing something. What?

Look again at what a hard training week does. It does two things at once, not one. It provokes adaptation — the slow rebuilding of enzymes, mitochondria, capillaries, tendon and bone that makes the athlete genuinely more capable. And it leaves damage and depletion — torn-up muscle, spent glycogen, disturbed sleep, blunted neural drive, an immune and hormonal system under load.

Both come from the same session, and this is the key: what you can do on a given day is not your adaptation. It is your adaptation minus your current burden. An athlete in a heavy block carries a great deal of both, and the second is masking the first.

Now the question sharpens. Stop the load and both start to fade — neither is permanent — so whether stopping helps depends entirely on which fades faster. And the answer is not symmetric. Muscle repairs and glycogen refills over days; the structural adaptations that took months to build do not unwind in a week. The burden decays quickly, the capacity slowly, and in the gap between those two rates sits the whole phenomenon.

That gives a shape rather than a rule. Remove the load and the burden falls away fast while the capacity barely moves, so measured performance climbs. Keep removing it and the capacity erodes too — plasma volume drops, enzyme activity and neuromuscular sharpness decline — while there is little burden left to shed. Performance turns over. There is a peak, and it sits at a particular moment, not at "as much rest as possible."

Can we say when? The literature converges on a window of roughly one to three weeks, with about two weeks a common optimum for endurance events. It says something more useful about how. The reductions that help are in volume — commonly on the order of a half, and in some studies more — while intensity is held near race pace and the number of sessions per week is barely reduced. That asymmetry is not arbitrary once you see the mechanism: volume generates the burden, whereas intensity and frequency keep the adaptation and the movement pattern sharp. Cut the wrong one and you shed fitness while keeping the fatigue.

The payoff is real but modest, and worth stating honestly: meta-analytic estimates of the gain from a well-designed taper cluster around a couple of percent, with individual studies ranging from essentially nothing to several percent. In a race decided by a fraction of a second, a couple of percent is enormous. It is not a transformation.

Two clarifications keep this honest.

The first is that "supercompensation" is easily overread. It does not mean the body builds extra capacity as a reward for resting. Most of what a taper delivers is the unmasking of fitness already there, plus genuine but bounded recovery — glycogen restored, muscle repaired, blood volume and hormonal balance normalised, neuromuscular force returned. Calling it "rebound above baseline" makes the rest itself sound productive. The training was productive; the rest lets you see it.

The second concerns the model doing the work. Two components decaying at different rates comes from impulse-response models of training, in which fitness and fatigue are each an exponential response to load, fatigue given a larger immediate effect but a shorter time constant. That captures the shape athletes experience, which is why taper theory looks tidy. But its parameters are fitted after the fact, it predicts individuals poorly, and it lumps many physiological processes into two curves. Treat it as a description of the shape, not a mechanism you can compute a taper from — which is why prescriptions come as ranges and coaches still adjust by feel.

Two corollaries follow. Complete rest is worse than reduced training, because it drops intensity and frequency along with volume. And a taper cannot manufacture fitness that was never built, only reveal it.

• Why the same taper length suits a sprinter and a marathon runner differently, given how their fatigue is generated.
• How overreaching deliberately deepens fatigue before a taper, and when that tips into overtraining.

Every term the collection defines is gathered in the glossary.