Pull-up progression: log the total, not just the belt

An 82 kg lifter doing five pull-ups with 20 kg hanging from a dip belt is moving 102 kg, and at RPE 8 that set is worth an estimated 125.8 kg for one. Log only the 20 kg and the number means nothing: no estimate, no sensible back-off, no next weight. Pull-up progression gets simple once the log holds the total, body weight plus added load, because then the pull-up behaves like any barbell lift.

Every figure below is compute.mjs output, the script that runs the app's own formulas, or carries a citation to the study it came from. You can check the estimates against the one-rep max calculator.

What the pull-up trains

In the app's exercise library the Pull-Up is a compound vertical pull: lats and upper back primary, biceps and forearms secondary. Chin-Up, Neutral-Grip Pull-Up, Wide-Grip Pull-Up and Weighted Pull-Up sit beside it and log the same way.

Grip changes less than the gym folklore says. A study of 19 strength-trained men compared the overhand, underhand, neutral and rope versions and found broadly similar activity across the muscles measured. The clearest difference was more middle trapezius activity with the overhand grip than the neutral one [1]. Pick the grip your elbows and shoulders like and keep it, because a grip change resets the comparison you are trying to make.

Three working sets run through the app's volume code give the back 3 sets and the biceps and forearms 1.5 each. The lats and upper back fold into one back total, and secondary muscles count half a set. A warm-up set adds nothing anywhere, as the post on whether warm-up sets count as volume explains.

Log the total, not the belt

The pull-up is a bodyweight plus load lift. The set card has an Added field for what hangs from the belt and a body weight beside it, and the app stores the sum as the set's load. It also keeps the body weight you had that day with each set, so a later weigh-in never rewrites an old session.

That one choice is what makes the rest work. The estimated 1RM, tonnage, PRs and stall detection all read the total, so a pull-up gets the same treatment as a squat with no special rules. A plain bodyweight set is logged with nothing added, and its load is simply your body weight.

Assistance goes in the same field as a negative number. A band or a machine that takes weight off you lowers the total, which is honest: you moved less than your body weight, and the estimate should say so.

Pull-ups are not per-side, so there is one line per set. Rate every working set. RPE 10 means no reps left, RPE 9 means one, RPE 8 means two [2]. A set logged without RPE is treated as RPE 10, the assumption every classic formula makes, so a blank rating on a set you stopped early reads as a max you did not hit.

Treating the pull-up as a loaded lift with a one-rep max is not a logging trick. A study of 52 men tested a pull-up 1RM, then found bar speed tracked percentage of that max closely enough to estimate reps left in a set from the first rep [3].

What 102 kg for five at RPE 8 is worth

The model reads the set through the published reps by RPE chart: five reps with two in reserve is a seven-rep effort, and 102 kg sits at 81.1 per cent of the estimated max. Classic formulas that ignore RPE give lower numbers for the same set, 119 kg by Epley and 114.8 kg by Brzycki, because they assume the set was a max.

The ladder below is the same estimate read across rep counts, every row at RPE 10. The load is the total. Subtract 82 kg for what goes on the belt.

Body weight alone, 82 kg, sits between this lifter's 12RM of 85.5 kg and their 15RM of 79.8 kg. So the same lifter who can grind five with 20 kg should manage somewhere in the low teens unloaded. If that is wildly off, one of the two sets was logged with the wrong RPE.

Both ends of that ladder build muscle. A meta-analysis of 21 studies found similar growth from light and heavy loads when sets went to failure, with heavy loads, above 60 per cent of 1RM, giving more one-rep strength [4]. As a lifter gets stronger, body weight becomes a smaller share of their max, and unloaded sets drift towards the light end. That is the argument for the belt.

Why there is no warm-up ramp

On a barbell lift the app builds a warm-up from the working weight. Run the same ramp on a 102 kg pull-up and it opens with a 20 kg bar and then a 50 kg set. Nobody who weighs 82 kg can hang at 50 kg, so the app does not generate a ramp for bodyweight plus load lifts.

Warm up by added load instead. A set or two of bodyweight reps stopped well short of failure, then a short set with part of the belt load, is enough for most lifters. Log them with the warm-up role so they stay out of your weekly sets.

How does the next weight move?

Two of the app's seven schemes suit weighted pull-ups. Top set + back-off fits a lifter who rates every set and wants one heavy set a session. Double fits a lifter working a rep range at one load. The post on how to choose a progression scheme covers the rest.

With Top set + back-off, the back-off is 88 per cent of the top set by default, and it is 88 per cent of the total. For the 102 kg top set that rounds to 90 kg, which is 8 kg on the belt. Taking 88 per cent of the 20 kg added load instead would barely lighten the set at all, and the back-off would stop being one.

At 125.8 kg the model puts eight reps at RPE 8 at 93 kg and at RPE 7 at 88.9 kg, so eight reps at 90 kg land between the two. Logged at RPE 8, those eight reps estimate 121.8 kg. Logged at RPE 7, 127.3 kg. The full method is in the guide to top sets and back-off sets.

The app's step for bodyweight lifts is 1 kg, not the 2.5 kg of a barbell. That matters, because the next top set has two honest ways up:

Both are progress. If your gym has no 1 kg plate, a rep at the same load is the cleaner step, which is exactly how double progression works.

Three sessions without progress flags a stall by default. A deload week never emits an increase, and more than 30 days off the lift resets it.

When your body weight moves

On a pull-up your body weight is part of the load, so a cut changes the lift even when the belt does not. Say the same lifter drops to 79.5 kg and still does five with 20 kg at RPE 8. The total is now 99.5 kg and the estimate falls to 122.7 kg.

That drop is real in absolute terms and misleading in relative ones. The lifter pulls the same added load at a lighter body weight, which is a gain in strength per kilo of body weight. A load increase never leaks through on a cut, so the app will not push the belt heavier while you are losing weight. Hold the added load, chase reps, and read the trend against your body weight. The post on building muscle on a cut covers the rest of that trade.

The reverse happens on a bulk. Gain a few kilos, keep the belt the same, and your total rises with no change in effort. That is a genuine gain in absolute load, so count it, but do not mistake it for a new technique.

Where pull-ups go wrong

The commonest mistake is logging the belt as the load. A 20 kg set reads as a light lift, the back-off comes out nearly the same weight, and every total is 82 kg too low.

Next comes letting the range shrink as the belt gets heavier. Half reps with more weight look like progress in the log and are not. Set a standard, chin over the bar and arms straight at the bottom, and keep it.

A stale body weight is the third. If the number on the card is from a month ago, every total is off by whatever changed. Update it when the app asks.

Kipping on sets you rate is the fourth. Momentum lets you finish reps your back did not do, and RPE stops meaning anything.

What changes when you swap it

A chin-up or neutral-grip pull-up logs exactly the same way, so the history carries the same meaning even if the numbers differ. Keep each grip as its own exercise rather than mixing them under one name.

A lat pulldown is different. It is a machine lift logged by the weight on the stack, a plain load lift, so the warm-up ramp is available and it steps like any machine. The numbers do not translate: a pulldown stack is not your body weight, and a pulldown 1RM tells you little about your pull-up. If you swap for a block, start a fresh history for the pulldown rather than reading it against the pull-up.

For weekly volume, every working set of pull-ups credits the back a full set. The post on how many sets per week per muscle gives the ranges from the research.

Questions

Should I log assisted pull-ups as negative weight?

Yes. Put the assistance in the Added field as a negative number, so an assisted-pull-up machine taking 20 kg off an 82 kg lifter logs a total of 62 kg. The estimate then reflects what you actually pulled. As the assistance comes down, the total rises and the log shows the progress a plain rep count would hide.

How many bodyweight reps before adding weight?

There is no fixed number, but the ladder gives a guide. Once you can do sets in the low teens at your body weight, added load is the simplest way to keep sets in a range you can rate honestly. Start the belt light and progress it in 1 kg steps, or a rep at a time.