The taper

How to arrange the final two to three weeks before a goal event: how much volume comes off, what stays, and what the last week actually contains.

REF-107 (A, meta-analysis of 27 studies in competitive athletes, outcome = competition or field-based performance) is the closest thing to a settled answer in this base, and it separates the four variables cleanly:

Read the effect sizes, not just the recommendations: the volume cut is the biggest lever (0.72) and the one athletes under-do, while intensity and frequency matter mainly because dropping them costs you (inference from the pattern of pooled optima; the meta-analysis reports which settings worked, not the cost of each deviation).

REF-536 (B, 1996, uncontrolled cohort, n=8 highly trained swimmers) corroborates only that a taper works in a different modality: after 12 weeks of intense training, a 4-week taper improved competition performance by 2.32% (p < 0.01), with no significant change in plasma hormones.

Its abstract does not describe the taper's content, so it says nothing about how much volume was cut or whether intensity was kept; the "volume −40-60%, 1-3 weeks" figures once in its entry are the base's own practice, not the study's.

REF-3492 (A, 2023 systematic review with meta-analysis, 14 studies; 174 athletes and ages 17-32 are full text) reproduces the four parameters from an independent literature sixteen years later, and adds three things this card did not have.

First, a ceiling: every taper duration band up to 21 days improved the time trial, and the ≥22-day band was the smallest and not significant, with a significant between-subgroup difference (P = 0.02, full text). Second, it puts numbers on the intensity rule — with intensity maintained the pooled effect was SMD −0.55 [−0.79, −0.31]; with intensity reduced it was 0.25 [−0.67, 1.17], not significant, the point estimate having flipped sign — though the between-subgroup test itself was P = 0.10 (full text), so this is a strong pattern across studies rather than a proven contrast.

Frequency behaved the same way (maintained −0.53 [−0.82, −0.25]; reduced −0.32 [−0.76, 0.13], ns; between-subgroup P = 0.42), and only the 41-60% volume cut reached significance (−0.77 [−1.23, −0.30]; the ≤20%, 21-40% and ≥60% bands did not). Third, a hard overload block immediately before the taper beat a conventional taper on time trial (P < 0.05).

It also found no improvement in VO2max or movement economy — the taper buys a race, not a test.

REF-110 (B+, narrative review from the leading taper group) states it directly: high-intensity training before the taper plays a key role in inducing maximal adaptations, high-intensity training during the taper can maintain or further enhance them, training volume can be markedly reduced without a negative impact — therefore the load should not be reduced at the expense of intensity. Not from a study; the base's own practice: the "detraining sets in within two weeks" figure is not in that abstract; the review names the risk of detraining with reduced load, not a time to onset.

So the athlete's instinct ("race is close, take it easy") attacks exactly the wrong variable. The correct taper feels strange: the sessions stay sharp and frequent, the hours collapse. Say this out loud when you prescribe it, because it will not feel like resting.

REF-587 (B+, narrative review) generalises the same rule outside the taper: in general populations endurance performance was maintained for up to 15 weeks with frequency cut to 2 sessions a week or volume cut 33-66%, as long as exercising intensity was maintained — intensity is the key maintenance variable; the authors state the data are insufficient for athlete-specific recommendations.

And REF-695 (B, narrative review) reports that in trained athletes near-maximal speed-endurance training maintained oxidative capacity and improved 30 s-4 min performance despite a reduction in training volume.

REF-098 (B+, parallel RCT, 12 weeks load-matched, n=14 block vs n=16 traditional, trained cyclists, VO2peak 58): no difference in performance between block and traditional periodization — both improved 5-min and 40-min TT power by ~9% and ~8%, Wmax and 4-mmol power similarly; only haematological and capillary adaptations differed. REF-103 (B+, systematic review of seven studies in trained road cyclists) reaches the same place from the other side — no evidence currently favours a specific periodization model over 8-12 weeks; the base's reading: choose the structure the athlete can execute consistently.

The practical consequence for a coach: the model is not where the performance is. Do not spend the athlete's attention on whether their season was "blocked" correctly; spend it on execution and on the taper parameters above, which are measured.

Not a template, a set of constraints. Not from a study; the base's own practice — a two-step reading of the pooled "2 weeks, −41-60% exponential" optimum in REF-107; the meta-analysis gives no week-by-week schedule:

No entry tests a taper for an event of this length; this section is the base's own extension of the four parameters.