Why is my counted volume lower than the sets I actually did?

Three exclusions. Warm-ups never count. Sets rated as further from failure than the stimulus ceiling never count — they happened, but they did not buy growth. And sets with no RIR rating are reported separately rather than assumed to have counted.

What is the 12 to 20 set band?

The range the screen draws behind every bar, as a per-muscle weekly target. It comes from the volume dose-response literature, where hypertrophy rises with weekly sets across the studied range with diminishing returns. It is a band rather than a number because the useful dose varies between people and between muscles.

Why is the scale the same for every muscle?

So they can be read against each other and against the band. Per-row scaling would make eight sets of calves look identical to sixteen sets of chest, which defeats the purpose of showing them together.

Does endurance training count toward these numbers?

No, and deliberately. Weekly volume counts resistance-training sets per muscle. Endurance load is measured separately as a ramp rate on minutes, because they are different capacities in different units and pooling them would produce a number that could not answer either question.

Check Your Weekly Volume Per Muscle

Q: How does Base count weekly sets per muscle?

Fractionally. A set counts 1.0 toward the muscle doing most of the work and 0.5 toward muscles that assist. A barbell row counted as zero biceps volume understates arm work badly enough to make the weekly number useless, and counting it as a full set overstates it. Pelland and colleagues in 2024 found fractional counting was the best-fitting model of the dose-response.

The screen

Weekly hard sets per muscle on a fixed 0-24 scale with the 12-20 band drawn behind every bar. A shared scale is the point — per-row scaling would make eight sets of calves look like sixteen of chest.

One bar per muscle, a shared axis, and the target band drawn behind. A "how this is counted" card sits at the bottom, because a number this heavily processed needs its rules visible.

Fractional counting

A set is worth 1.0 to the muscle that is the primary force generator and 0.5 to muscles that assist.

Pelland et al. 2024 — 67 studies, 2,058 subjects — found the best-fitting model of the volume/hypertrophy dose-response counts sets this way. The alternatives are both clearly wrong: counting a barbell row as zero biceps volume understates arm work badly enough to make the weekly number useless, and counting it as a full set of biceps overstates it just as badly.

Barbell row × 4 sets
  lats        4 × 1.0 = 4.0
  biceps      4 × 0.5 = 2.0
  rear delts  4 × 0.5 = 2.0

Half sets are real here, so the screen shows 12.5 rather than rounding to 12.

A custom exercise you typed in gets no secondary muscles rather than a guess. Under-counting a custom lift's assistance work is recoverable — the weekly number reads slightly low. Inventing assistance that is not there is not.

The three exclusions

This is why counted volume is lower than sets performed, and all three are deliberate.

Excluded	Why
Warm-ups	They were not working sets
Sets above the RIR stimulus ceiling	They happened and produced fatigue, but did not buy adaptation
Sets with no RIR rating	Reported separately, never assumed

The third is the one that surprises people. Base will not guess whether an unrated set counted, because a volume number you cannot trust is worse than no volume number.

The practical consequence: an unrated week reads as a light week. If your volume screen looks empty and your training does not, the fix is rating your sets, not training more.

The second exclusion is the one that matters conceptually. A set taken five reps short of failure is how a programme reads as compliant while producing nothing — sixteen sets of chest, none of them hard enough to matter.

The band

12–20 sets per muscle per week, drawn behind every bar.

It is a band rather than a target because the useful dose genuinely varies between people, between muscles, and with how much recovery capacity is being spent elsewhere. For a hybrid athlete that last clause is not hypothetical — endurance volume draws on the same recovery budget, and the top of the band is harder to justify in a heavy running block.

The screen states a floor as well. Below it, a muscle is clearly under-dosed and the bar says so.

Frequency is deliberately absent from all of this. Once weekly volume is equated, the same literature finds frequency largely irrelevant, so Base says nothing about how the week is split up.

Volume is a dose, not a score

Worth stating plainly: more is not better past the band. The dose-response has diminishing returns and eventually turns, and the fatigue cost keeps rising linearly.

The screen exists to answer "is anything obviously under- or over-dosed", not to be maximised. A muscle sitting at 22 sets while another sits at 6 is a distribution problem, and the fix is to move sets rather than add them.

If you are designing rather than auditing, the split builder's review screen shows the projected volume of a week before you run it, on the same scale.

Common mistakes

Reading a low bar as under-training when the sets were unrated. Check ratings first.

Adding sets to push a bar into the band. If the band is not being reached with hard sets, the answer may be redistribution rather than addition.

Expecting endurance work to appear here. It does not. See managing your ramp rate for the endurance equivalent.

Comparing your bars to someone else's. The band is a population reference, and where you sit inside it depends on recovery, training age and what else the week contains.

Treating a custom lift's low secondary contribution as a bug. Unmapped lifts contribute direct volume only, on purpose.

Chasing the top of the band during a hard endurance block. Both halves draw on one recovery budget, which is the premise the whole app is built on.

What to do next

Open the volume screen and look for the gap between counted sets and sets you know you performed. If it is large, the cause is unrated sets — fix that in the edit grid before changing anything about your programme.