Most training apps default to a 4:1 periodisation structure. Three weeks of progressive loading, one week of deload. Rinse and repeat.
It’s the standard recommendation. It’s in every beginner periodisation guide. It’s what your Garmin coach plan does. It’s what TrainingPeaks default templates assume. It’s the structure Strava fitness and freshness charts are visually designed to reinforce.
It’s also not supported by the strongest evidence in the research. And it’s probably why your deload weeks don’t feel restorative, your peaks never land cleanly, and you enter every important training block carrying invisible fatigue from the last one.
Where the 3:1 Structure Came From
The 3 hard weeks plus 1 easy week pattern has a specific origin. It comes out of bodybuilding and powerlifting traditions where the training stimulus was heavy compound lifts and the recovery cost was primarily neural.
For that specific kind of training, in that specific population, 3:1 works reasonably well. Three weeks is long enough to drive strength adaptation. One week is enough to clear most of the accumulated neural fatigue. It became the default because it worked well enough for the dominant training culture of the era.
When endurance coaches adopted periodisation, many of them imported the 3:1 structure without examining whether it fit endurance physiology. Training apps then digitised the convention. Consumer platforms inherited the default. The pattern replicated itself through the industry without anyone checking the foundation.
The research on endurance training periodisation actually supports something more varied, more athlete-specific, and generally involves more frequent, shorter recovery windows than 3:1 provides.
What the Research Actually Suggests
Several reasonable conclusions emerge from modern endurance periodisation research.
The optimal recovery frequency depends on training intensity distribution. Athletes doing predominantly polarised training (lots of easy volume, small amounts of very hard work) can sustain longer loading blocks than athletes doing threshold-heavy pyramidal training. A 4 or 5 week build can work for polarised programs. A threshold heavy block often needs a touch every 2 to 3 weeks.
Deload magnitude matters more than frequency. A true deload, where volume drops by 40 to 60 percent and intensity by 20 to 30 percent, can restore adaptation capacity within 5 to 7 days. A soft deload, where volume drops only 15 to 20 percent and intensity is unchanged, is not a deload. It’s just a slightly lighter week. It delays fatigue without clearing it.
Individual variation is massive. Heart rate variability based periodisation research suggests different athletes need recovery windows at very different frequencies. Some athletes show accumulated fatigue signals at 10 days. Others can push to 5 weeks. The 3 week fixed interval is wrong for most of them in one direction or the other.
Age, lifestyle load, and training history modulate the answer. A 45 year old working a demanding job and training 12 hours a week has different recovery needs than a 25 year old athlete whose only stressor is training. The 3:1 pattern ignores all of that.
None of this is controversial in sports science. It just hasn’t filtered into consumer training apps, which continue to recommend calendar-driven deloads that ignore individual response.
How to Tell If Your Deloads Are Working
Here’s a simple set of checks you can run on your own data.
The Monday after deload should feel meaningfully different. If you’ve been deloading for 7 days and the Monday following feels the same as the Monday before, your deload didn’t do enough. Volume reduction wasn’t significant enough or duration wasn’t long enough.
Heart rate at submaximal paces should drop. Run or ride the same route at the same perceived effort before and after deload. Your heart rate should be noticeably lower after. If it isn’t, the deload hasn’t restored the physiological capacity that training stress had reduced.
Resting heart rate and HRV should normalise toward pre-block baseline. Not surpass baseline. Just return to it. If you were running a training block with HRV suppressed 8 percent below baseline and RHR elevated 4 beats, deload should bring both back to within 2 percent and 1 beat of baseline. If they don’t fully normalise, one of two things is happening. Deload wasn’t deep enough, or the training block had accumulated more fatigue than 7 days can clear.
Subjective motivation should rebound. The tell-tale sign of an underpowered deload is entering the next loading block still dragging. Feeling like you should want to train hard but the desire isn’t there. This is your central nervous system telling you it needed more time than the calendar gave it.
The Deeper Problem with Calendar-Driven Recovery
The assumption underneath 3:1 is that accumulated fatigue is predictable and timeboxes neatly. Three weeks of load equals roughly 7 days of needed recovery.
This is not how physiology works.
Fatigue accumulates non-linearly with training load. A week of 9 hours of training at 80 percent threshold costs differently than 9 hours at 65 percent threshold. Life stress, poor sleep, and environmental factors add invisible load that the training log doesn’t capture. The same written block produces different actual fatigue in different weeks.
Calendar-driven deloads assume the fatigue arrived on schedule. Real fatigue often doesn’t. You can be flatly overreached after 2 weeks or still fresh after 5. The calendar is agnostic to this. Your body is not.
What should drive deload timing is data. Specifically, trend data. Not individual readings.
Suppressed HRV for 5 consecutive days. Independent of the calendar, this is a deload signal. If your HRV is sitting 10 percent below baseline for five straight mornings, you are in accumulated fatigue territory. Whether it’s week 2 or week 4 of your block, the body is telling you to back off.
Rising resting heart rate with no explanation. A 4 beat upward shift in RHR over a week, not attributable to illness or environmental factors, is your cardiovascular system signalling recovery debt. This is a deload signal, not a keep pushing signal.
Decoupling of heart rate and pace. When you run the same pace at 7 beats higher heart rate than last week, something is shifting. Sometimes heat. Sometimes illness. Sometimes cumulative fatigue that a calendar hasn’t yet told you about.
What a Responsive Deload Structure Looks Like
The approach that works better for most endurance athletes I’ve talked to isn’t a fixed interval. It’s a responsive one.
Loading blocks of 2 to 5 weeks, driven by accumulated fatigue markers. When the data says fatigue is mounting, that’s the block boundary. Not the calendar.
Deloads of 4 to 10 days, depending on how deep the accumulated fatigue was. Deeper fatigue needs longer recovery. 7 days isn’t a universal answer.
Intensity preserved, volume reduced. The most effective deload structure for endurance athletes appears to be preserving short high quality efforts while cutting volume dramatically. Keep the sharpness, lose the fatigue. Opposite of what most apps suggest, which is often to maintain volume and cut intensity.
Strength work during deload, if running the kind of program that uses it. Many athletes benefit from maintaining some strength stimulus even during endurance deloads, because strength decay is faster than you think and the recovery demand is low if the volume is cut appropriately.
The Platform Problem
Consumer training platforms have not caught up to any of this.
They recommend 3:1 because it’s simple, universal, and hard to argue with at the level of a marketing tagline. They don’t recommend responsive periodisation because that requires actually understanding the athlete’s individual fatigue markers and connecting them to training decisions. That’s a hard problem. Most platforms haven’t solved it.
Which is exactly the gap that intelligent training analysis should fill. Not the question of how hard to train. The question of when you should stop.
Green score. Destroyed legs. There are 6 blind spots in your wearable data. We wrote a free guide covering every one of them.
Green score. Destroyed legs. There are 6 blind spots in your wearable data. We wrote a free guide covering every one of them.
Download the Free Guide