Programming
Training to failure: when it helps and when it just costs you
Failure is treated as a virtue in gym culture and as a mistake in most coaching. The evidence supports something in between, and it depends heavily on the exercise.
The question of how close to failure to train sits underneath nearly every programming decision, and it has a reasonably clear answer that neither camp fully accepts.
What the evidence broadly shows
For hypertrophy, sets taken close to failure — within roughly one to three reps — produce similar growth to sets taken to failure, provided volume is equated. Going all the way to failure does not appear to add much, and it substantially increases fatigue and recovery cost.
For strength, training to failure appears if anything slightly counterproductive. Maximal strength responds to heavy practice performed well, and grinding reps at the end of a set degrade technique and add fatigue without adding useful practice.
The general conclusion most coaches have arrived at: train close to failure most of the time, go to failure occasionally and selectively, and almost never on the main barbell lifts.
The concept of reps in reserve
Reps in reserve — how many more repetitions you could have completed — is the practical tool for managing this. A set at two reps in reserve means you stopped with two good reps left in the tank.
Most productive training sits somewhere between zero and four reps in reserve, with the bulk around one to three. Sets stopped further from failure than that contribute less per set, though they still contribute.
The catch: people are bad at estimating this, and untrained lifters are particularly bad — they routinely believe they are at failure when they have four or five reps left. The estimate improves with experience, and the fastest way to calibrate it is to occasionally take an isolation exercise genuinely to failure and see how far off you were.
Where failure is appropriate
Isolation exercises with low systemic cost. Curls, lateral raises, leg extensions, cable work. Failure here is safe, easy to recover from, and gives you a clear reference point for what failure feels like.
Machine-based work where technique cannot degrade dangerously and there is no risk of being pinned.
The last set of an exercise, when it will not compromise subsequent work.
Occasionally, to test and recalibrate your sense of proximity to failure.
Where it is not
Heavy compound barbell lifts. A squat taken to failure means either failing under load or grinding a rep with a collapsing position. Both are poor practice and both carry injury risk. Deadlifts to failure are worse — the lower back position degrades and the fatigue cost is enormous.
Early in a session, where it wrecks everything after it.
High-frequency programmes. If you train a pattern three times a week, failure on Monday means Wednesday is compromised. Frequency and failure trade against each other.
When technique is still being learned. A beginner's last few reps before failure are the worst reps they will perform, and repeated practice of poor movement is precisely what you do not want to be grooving.
The fatigue cost, which is the crux
The final rep or two of a set to failure contributes marginally more stimulus and considerably more fatigue — neurological, muscular and psychological. The stimulus-to-fatigue ratio gets worse the closer you push.
Since total weekly volume drives hypertrophy, and fatigue limits how much volume you can perform and recover from, training every set to failure typically reduces the total productive work you can do in a week.
Two sets to failure may be less effective than four sets at two reps in reserve, despite feeling much harder.
The psychological trap
Failure feels productive. There is a satisfying certainty to having genuinely emptied the tank, and stopping with reps left feels like leaving something on the table.
Which is exactly why it needs to be a programmed decision rather than a mood. A lifter who trains by feel will drift toward failure on their good days, which are the days when they should be accumulating quality volume instead.
How to apply it
Write the target proximity into the programme, alongside the sets and reps. "Four sets of eight at two reps in reserve" is a complete instruction; "four sets of eight" is not.
Log the actual proximity you achieved. Over time, comparing your estimate to your performance on the next set tells you how accurate your sense of it is.
Push closer to failure late in a training block, when you are about to deload anyway and the accumulated fatigue has somewhere to go. Stay further from it early in a block, when you are building volume.
And on the big lifts, leave reps in the tank as a default. The strongest lifters in the world spend most of their training well short of failure, which should tell you something.
