Autoregulation: training by effort rather than by percentage

A percentage-based programme says: today, squat five sets of three at eighty percent. It assumes you know your true current maximum and that today is an average day.

Both assumptions fail regularly. Maxes drift between tests. Sleep, food, stress and accumulated fatigue mean the same absolute load feels dramatically different from week to week.

Autoregulation adjusts the load to the day.

The tools

Reps in reserve. How many more repetitions you could have completed with good technique. A set at two reps in reserve means you stopped with two in the tank.

Rating of perceived exertion, in the version used for lifting, is essentially the inverse: an RPE of eight corresponds to two reps in reserve, nine to one, ten to failure.

Both describe the same thing and either works. Reps in reserve is more intuitive for most people because it refers to something concrete.

How a session looks

Instead of "five sets of three at eighty percent", the instruction becomes "work up to a top set of three at eight RPE, then four sets of three at ten percent less".

On a good day the top set might be considerably heavier than eighty percent. On a bad day it might be seventy. In both cases the training stress is appropriate, which is the point.

The back-off sets scale automatically because they are derived from the top set.

The main objection, and its answer

People are bad at estimating proximity to failure. This is well documented — untrained lifters routinely believe they are at failure with four or five reps remaining, and even experienced lifters underestimate.

The estimate improves with practice, and it improves fastest if you calibrate deliberately.

Take an isolation exercise genuinely to failure occasionally and note how far off your prediction was. Use a velocity device if you have access to one — bar speed correlates well with proximity to failure and provides objective feedback. And record your estimate before the set and your assessment after, so you build a track record.

Within a few months, most people become reasonably accurate in the ranges they train regularly.

Where it works best

Experienced lifters, who have a calibrated sense of effort.

People with variable lives — shift work, young children, demanding jobs — where readiness genuinely fluctuates a lot.

Heavy compound work, where hitting a prescribed load on a bad day is where injuries and grinding failures happen.

Where it works less well

Complete beginners, who cannot yet judge effort and who benefit from the simplicity of adding weight every session.

People who consistently underestimate their capacity, who will autoregulate downward indefinitely and never progress. This is a real failure mode and it needs a floor — a minimum load below which you do not go without a reason.

People who consistently overestimate, who will grind every session at ten RPE and accumulate fatigue.

Both failure modes are solved by combining autoregulation with a planned structure rather than replacing structure with feel.

The hybrid that works

Most well-run programmes use both. A planned progression provides the structure and the expectation, and autoregulation adjusts within a range.

For example: the plan calls for a top set of five at around a given load, with an instruction to stop at eight RPE. If it comes in heavy, you use the planned load. If eight RPE arrives ten kilos below the plan, you use that and note it.

Several sessions where the autoregulated load falls well below the plan is a signal — either the progression was too aggressive or fatigue has accumulated and a deload is due. That information is only available if you are tracking both.

Autoregulating volume as well as load

The less common and equally useful application. Instead of a fixed number of sets, prescribe a range and a stopping rule.

For example: perform sets of eight at a fixed load until the reps drop below six, up to a maximum of six sets. On a good day you do six; on a poor one, three. Total volume matches capacity automatically.

This is particularly useful for accessory work, where the precise volume matters less than the accumulated effort.

The mindset shift

The hardest part is accepting that a lighter day executed correctly is a successful session rather than a failure.

People attached to the numbers in their programme will push to hit them on days when doing so is counterproductive, which is precisely the behaviour autoregulation exists to prevent.

Judge the session by whether the intended training stress was delivered, not by whether a specific weight was moved. Over a block, the numbers take care of themselves — and they usually end up higher than a rigid plan would have produced, because you did not spend three weeks recovering from a bad Tuesday.