Proximity to Failure

Every working set ends somewhere between an easy stop and the last repetition a lifter can complete. How close a set ends to that point, counted as repetitions in reserve (RIR), is one of the variables that decide what a set does. This document sets out how close working sets end in a program, which work the rule covers, and how the load follows the effort.

What the research shows

• Muscle growth depends on it. In meta-regressions across the training literature, muscle size increased as sets were terminated closer to failure.
• Strength mostly does not. In the same analyses, strength gains were similar across a wide range of repetitions in reserve.
• Failure itself adds little. Pooled against non-failure training, sets taken to failure gave only a trivial hypertrophy advantage, and sets taken to momentary muscular failure gave no clear advantage at all. An earlier meta-analysis found no significant difference in strength or size between failure and non-failure training.
• Failure costs recovery. After sessions taken to failure, neuromuscular performance took considerably longer to return; sessions that stopped short of failure recovered faster between 24 and 48 hours.
• Lifters can estimate it. The RIR-based RPE scale tracks proximity to failure. Across studies, people underpredict the repetitions they have left by about one, predictions are more accurate closer to failure and with fewer than 12 repetitions to go, and training status does not appear to change accuracy.

The rule

Parameter	Value
Hypertrophy work	working sets of 6 or more reps below 80% of 1RM
Target effort	RPE 8–9, 1–2 reps in reserve
Never prescribed on hypertrophy work	RPE 10, failure
Unchanged	sets of 5 or fewer reps, sets at 80% of 1RM or more
Unchanged	deadlifts, back and front squats, Olympic lifts, jumps and sprints
Unchanged	warm-up sets and deload weeks
Load check	a load more than 5 points of 1RM away from its reps and RPE moves to the Epley load
Load check limit	up to 16 repetitions to failure

Hypertrophy work ends one or two repetitions short of failure. That is where the growth benefit of effort is captured without the recovery cost of failure. A set prescribed at RPE 7 is raised to RPE 8; a set prescribed at RPE 10 is lowered to RPE 9.

Heavy strength work keeps its prescription. Low-repetition and high-load sets serve strength, which responds to load and volume rather than to the last repetitions of a set. Lifts that load the whole body and the nervous system heavily, and explosive work where speed and technique are the point, also keep the effort the program gave them.

Warm-ups and deloads are not affected. A warm-up prepares the working sets, and a deload exists to lower effort.

The rule applies to every experience level. Beginners do not estimate repetitions in reserve worse than trained lifters, and the loads recommended for novices in position stands on resistance training already correspond to 8–12 repetition maximums, which is effort close to failure.

How the load follows the effort

At the same number of repetitions, fewer repetitions in reserve means a heavier load. When a set moves from one RPE to another, its percentage of one-repetition maximum moves with it, using the Epley relation between repetitions to failure and load:

• repetitions to failure = repetitions performed + repetitions in reserve;
• fraction of 1RM = 1 / (1 + repetitions to failure / 30).

Ten repetitions at RPE 7 (13 to failure) and at RPE 8 (12 to failure) differ by about 2.4% of the load, so a set written at 70% of 1RM becomes 71.5%. The same relation estimates one-repetition maximums from logged sets, so prescriptions and estimates stay on one scale.

When the load does not match the effort

A target RPE is only as good as the load that comes with it. Eight repetitions at 65% of one-repetition maximum leave four or five repetitions in reserve for most people, whatever effort is written next to them. So after the effort is set, the load of each hypertrophy set is checked against its repetitions and RPE.

• The average relation. Across repetitions-to-failure tests pooled from 269 studies, people completed about 10 repetitions at 80% of 1RM, 13 at 75% and 16 at 70%. Sex, age and training status changed the relation little; the leg press allowed more repetitions and the bench press slightly fewer.
• The individual spread. At a given load, repetitions varied between people by about 2.5 repetitions (one standard deviation) at 80% of 1RM and about 4.4 at 60%.
• The rule. A load more than 5 points of 1RM away from what its repetitions and RPE imply is moved to the Epley load. Differences within 5 points, roughly the individual spread, are left as written.
• The prudent side. At each number of repetitions, the Epley relation gives a slightly lighter load than the pooled average, so a corrected set errs on the light side.
• The limit. The check applies up to 16 repetitions to failure; beyond that the spread between people is too wide for a single number, and the load stays as written.

Common errors

• Leaving three or more repetitions in reserve on hypertrophy work. It is comfortable and repeatable, and it gives up growth the same sets could produce.
• Prescribing failure on every set. It adds recovery time without adding muscle, and it makes the next session worse.
• Raising effort without raising load. An RPE target that the prescribed load cannot reach turns into a set that stops early.
• Treating the rule as a reason to add sets. Effort and volume are separate levers; this rule changes how hard each set is, not how many there are.