What this means in real training
Use a buffer before using failure
A productive set usually needs to be close enough to failure to challenge the target muscles, but close enough is not the same as missed reps.
RIR simply means how many good reps you probably could have done before failing.
For many normal hypertrophy sets, a practical default is to stop when one to three clean reps are still likely available.
A simple 2-4 week test is to keep most compound lifts around 1-3 RIR, then compare reps, load, joint tolerance, soreness, and next-session performance before adding more true failure.
If you cannot tell whether a set had one rep left or five, spend the next block practicing consistent reps, rest periods, and logbook notes before making failure the target.
That does not mean every set must end with form breakdown, missed reps, or recovery debt.
Recent meta-analyses and meta-regressions suggest proximity to failure is a continuum, not an all-or-nothing switch, and strength and hypertrophy do not respond identically.
Where failure can fit
Failure is more reasonable on stable exercises, isolation movements, machine work, or final sets where the cost is lower.
It is riskier and often less useful when fatigue makes technique collapse on heavy compounds, loaded hinges, Olympic lifts, or movements where missed reps are expensive.
If the session only works when every set is a grind, the plan is probably too fatiguing to repeat well.
A simple default is RPE 7-9 for most compound work, occasional RPE 10 on safer accessories, and no load increase unless the planned reps were completed with the same technique, range, and rest standard.
For example, keep squats, presses, rows, and hinges 1-3 reps short most of the time, then use 0-1 RIR on one final machine curl, leg extension, lateral raise, or cable movement if the rest of the workout still recovers.
A practical week might include zero true-failure sets on heavy compounds, one or two close-to-failure accessory slots per muscle, and a note to remove the newest failure set first if the next session gets worse.
If you want to test failure, make it a small experiment: one final machine curl, leg extension, lateral raise, or cable movement, not every squat, row, and press in the week.
Why failure is not free
Training to failure tends to create more acute fatigue than stopping a bit short, which can affect later sets and recovery.
That extra fatigue cost is one reason “every set” is a bad default even if failure can still be useful sometimes.
The cost is not only soreness. It can show up as fewer quality sets later in the workout, slower load progression, more joint irritation, or a plan that needs constant deloads to survive.
Deload or pull back when two or more sessions in a row show worse reps at the same load, rising joint irritation, sleep-disrupting soreness, or technique that breaks before the target muscle is trained.
When that happens, make the first fix boring and measurable: cut one failure slot, stop accessories 1-2 reps earlier, or trim 10-20% of optional volume for two weeks while keeping the main exercises stable.
Before adding more failure work, check the boring variables first: weekly set volume, rest intervals, exercise selection, sleep, calories, and whether the same lifts are actually progressing.