Two Sets Is Enough

Most training advice is a volume argument. Ten sets a week, twenty sets a week, more is better until it isn't. The arguments are conducted entirely in the currency of sets, and almost nobody stops to ask whether a set is the right unit at all.

It isn't. Here's the unit that works.

#Two conditions, simultaneously

A muscle fibre grows when two things are true at the same instant.

It is recruited. Motor units fire in strict ascending order of size — Henneman's size principle. Small, low-threshold units driving fatigue-resistant type I fibres go first; the large, high-threshold units driving the type II fibres with the most growth potential go last. Recruitment tracks effort, not load: full recruitment happens around 90% of maximal unfatigued isometric force, roughly a 5RM — but you reach the same recruitment with a light weight by taking the set close enough to failure that fatigue forces the nervous system to call up the rest.

It is under high mechanical tension. An activated fibre that isn't loaded hard doesn't grow. Force in a sarcomere is a function of how many myosin heads are bound to actin at once. During fast shortening, heads detach before completing their power stroke — few crossbridges engaged, low per-fibre force. During slow shortening they attach, stay attached, and accumulate. That's the force-velocity relationship, and it's why a grinding rep transmits far more tension per fibre than an explosive one despite feeling like less work.

Both conditions are only met in the last handful of reps before failure. Everything before that is a tax you pay to get there.

#What this does to set counting

If roughly the last five reps of any set to failure are the stimulating ones, then:

• A set of 20 to failure contains ~5 stimulating reps.
• A set of 6 to failure contains ~5 stimulating reps.
• The other fourteen reps in the first set bought nothing except fatigue.

This is why the load barely matters between about 5RM and 30RM, provided you go close enough to failure — a result that looks bizarre if you're counting volume-load (sets × reps × weight) and obvious the moment you count stimulating reps instead.

#And why the fourth set is nearly wasted

Both conditions degrade within a session, for different reasons. Peripheral fatigue — accumulating inorganic phosphate and hydrogen ions inside the fibre — reduces the force each crossbridge can produce, so per-fibre tension falls even at identical recruitment. Central fatigue reduces achievable recruitment, so the high-threshold units you were training for stop getting reached.

By the fourth set you're performing the movement without delivering either half of the stimulus.

#The frequency argument

Here's the part that actually changes a programme. Recovery scales with session volume: trained lifters need more than four days to recover from 6–8 sets to failure, but roughly 48 hours from 1–2.

So low volume per session isn't a compromise on stimulus. It's what buys the frequency. Three sets spread across three sessions beat the same three sets stacked into one, because each of those three first sets is delivered by a recovered nervous system into an unfatigued muscle.

That argues for something like: two hard sets of 5–8 reps per muscle, taken to within a rep of failure, three times a week, training every other day.

#Where I'm going beyond the evidence

Two honest caveats, because this is exactly the kind of claim that gets oversold.

First: Chris Beardsley — whose model this is — puts the ceiling at around five sets per muscle per session, not two. Between five and ten sets the curve plateaus without obvious harm. The two-set prescription is a minimum-effective-dose bet chosen for adherence and time, not something his own reading demands. If progress stalls, the room is in adding sets toward five, not adding days.

Second: this is a synthesis he argues for, not settled consensus. There's a substantial dose-response literature supporting higher weekly set counts, particularly in advanced trainees, and no study has cleanly separated the two camps. What isn't seriously contested: proximity to failure matters, the last reps do the work, and junk volume performed in a fatigued state is worse than useless because it consumes recovery.

#One more free variable

Load the stretched position wherever the exercise allows. At long muscle lengths, sarcomeres sit past optimal actin-myosin overlap and the deficit is carried by passive tension in titin — which is not just a spring but a mechanosensor, with strain-dependent unfolding exposing binding sites for signalling proteins. Hypertrophy at long lengths also biases toward adding sarcomeres in series.

The applied result, from meta-analysis: lengthened partials match or beat full range of motion. Overhead triceps extensions over pushdowns. Incline curls over preacher curls. Seated leg curls over lying.

It costs nothing and most people leave it on the table.

---

This is one lever out of a larger system. The full protocol — every mechanism, every dose, every source tier, and an append-only record of everything it has gotten wrong — is at /protocol.