Twice a week is enough: a full linear progression on two sessions

Starting the program

Two full-body sessions cover the whole body between them: Workout A is squat, bench and row; Workout B swaps in the overhead press and deadlift. Each session is a squat, a press, a pull and a hinge, which is what makes two days a week viable at all. Nothing gets skipped because there's no third day for it to live on.

There are no training maxes to work out. Linear progression runs off a working weight per lift, so the app asks what you can currently handle for the prescribed sets and starts there, 60 kg on the squat here. Start conservatively: the method depends on having room to add weight for months, and the app adds it for you.

A start date goes in the calendar and every session date flows from it. From that point the program tells you what to lift and when.

The climb: 5 kg a session, decided for you

Squat opens at 60 kg. Complete the prescribed sets and reps and the app adds 5 kg to the next session. You don't tell it you succeeded and you don't do the arithmetic. It reads the sets you logged and moves the weight.

Because the squat is in both sessions, it moves twice a week: 60, 65, 70, 75. Four sessions in and it's already up 15 kg. That's the compounding that makes novice linear progression worth running even on a two-day schedule. The frequency is low but the increment is generous, and the lower body earns 5 kg jumps rather than 2.5.

The chip under the exercise name shows the rule while you're training rather than burying it in a settings screen, and the PREV column shows what you did last time. You turn up and the weight is already on the screen.

The stall, where the climb runs out

At 80 kg it stops. The prescription asks for five reps on every working set and only three go up. That's a failed session, and it's the moment that separates a progression engine from a spreadsheet: the obvious mistake is to look at the heaviest completed set, see 80 kg on the bar, and add weight anyway.

Durata holds. The weight stays at 80 kg for the next session, and the one after that, because the target was five reps on every working set and three of them fell short. Nothing advances on a session you didn't complete, and lowering the bar to make the reps doesn't count either. What you actually lifted is what's judged.

Three of those in a row is the program's cue to act, and you can check its working before it does. The rule chip under the exercise name is a button: tap it and the app lays out the evidence from this exact slot's own history against the rule it is being judged by. Two misses recorded at 80 kg, a third pending, and the law that says three misses at a weight cuts it by 10%.

The deload: the app backs you off

Three missed sessions trips the rule and the app drops the squat by 10%, so 80 kg becomes 72.5 kg, then restarts the climb from there. You didn't decide that, calculate it, or remember to do it. The program specifies a deload after three failures and the engine applied one.

The deload doesn't cost you the block. It sets up the climb that follows. Twelve sessions across six weeks look like this: four sessions of adding weight, three where it held, one cut of 10%, and a rebuild that carries you straight back through the weight that stopped you. By session 10 you're above it. You finish at 92.5 kg, which is 12.5 kg above the ceiling that beat you and 32.5 kg above where you started.

The curl runs a different rule entirely

The barbell lifts run linear progression. The curl doesn't. It runs double progression, and both rules are live in the same session. On double progression the weight is fixed and you chase reps inside a range first. Only when every set reaches the top of the range does the weight go up, and the reps reset to the bottom.

Reps inside the range are not a failure. Getting 9 where you managed 8 last time is progress, so the app holds the weight and lets you keep climbing rather than counting it against you. It's the right rule for an isolation lift, where 5 kg jumps every session would be absurd.

Six weeks later

Twelve sessions fill the History calendar from the start date, two a week, and the squat's own Estimated 1RM chart draws the shape of the block: a climb, a flat stretch where the weight held, the dip at the deload, then a longer climb through the old ceiling. That sawtooth is what linear progression actually looks like once you train long enough to stall.

Six weeks of training on two days a week, no spreadsheet, and no session where you had to work out what to put on the bar.

Start to finish

The same lift and the same working sets, twelve sessions apart. Every step in between was decided by the engine from the sets that were logged: four increases, three sessions where the weight held, one cut of 10%, and a rebuild straight back through the ceiling.

The estimated one-rep max tells the same story from the other end. Session one's squat estimated 70 kg; the last one estimates 107.9 kg. Six weeks, two sessions a week, and nothing to work out.

Session 1

60 kg → 92.5 kg

+32.5 kg

The squat, start to finish: 60 kg on session one to 92.5 kg on session twelve, having stalled three times at 80 kg and been cut to 72.5 kg on the way.

Session 12