Deadlift progression for lifters who log every pull

You pull 180 kg for three, and the third rep leaves the floor slower than the first two but locks out clean. You could have done one more. That is RPE 9, and it is the only number a deadlift progression needs from you each week. Logged that way, the triple is worth an estimated 201.8 kg for one, computed with the model behind the one-rep max calculator, and the app's next suggestion is 185 kg. Every number below comes from compute.mjs, the code that runs the app's formulas, or carries a citation.

The plan is one heavy top set a week, rated by reps in reserve, with back-off sets behind it and load added only when the rating says the bar is moving.

What the deadlift trains, in the app's terms

In the app's exercise library the conventional deadlift is a compound hinge. Hamstrings, glutes and spinal erectors are the primary muscles, and lats and traps are secondary.

The muscle research is less tidy than any library entry. A systematic review of 19 deadlift studies found the erector spinae and the quadriceps reported greater activation than the gluteus maximus and the biceps femoris, although the studies used methods too different to pool [1]. So the lift loads more of the leg than the label says. The app credits the deadlift's working sets to the posterior chain, which is where most lifters expect to feel it, and you should not expect a heavy pull to cover your quad sets for the week.

How to log a deadlift

A deadlift is a load lift and it is not per-side, so one set is one line with the weight on the bar. Give every set a role. Sets on the way up are warm-ups and never count toward volume. The heaviest set of the day is the top set. Anything after it at a lower load is a back-off.

Rate the top set by reps in reserve. On this scale RPE 10 means no reps left, 9 means one, and 8 means two [2]. A set logged without RPE is treated as RPE 10, the assumption every classic one-rep max formula makes, so the same 180 kg triple logged blank comes out at 195.2 kg instead of 201.8 kg. The gap feeds every suggestion that follows.

Trust the count, not the bar speed. Fifteen powerlifters working up to a true max rated the deadlift 9.6 out of 10 on average, and the bar on that last pull moved at 0.14 metres a second, slower than their squat max at 0.23 [3]. In the same lifters the rating tracked the percentage of max closely on all three lifts, with correlations of 0.88 to 0.91 [3]. A slow lockout is normal for a heavy pull. Ask yourself whether one more rep would have come off the floor.

The session's credit works like this. Run five warm-up sets and four working sets through the app's volume code and the hamstrings and glutes each get four sets, and the traps get two as a secondary muscle. The five warm-ups add nothing. The post on whether warm-up sets count as volume explains why the app draws the line there.

The warm-up ramp for a 180 kg top set

The app builds the ramp from the planned weight of the first working set. For 180 kg, compute.mjs warmup --kg 180 returns five sets, because the 85 and 90 per cent rungs switch on once the working weight passes 80 kg and 120 kg.

The top set is three 25s and a 5 on each side, from the plate calculator. Each ramp load rounds to the nearest 2.5 kg, and anything heavier than 95 per cent of the working weight is dropped. Skipping the empty-bar set on a deadlift is fine: pull from plate height and start at 90 kg. The single at 162.5 kg matters more: it tells you how the 180 kg triple will feel before you commit to it.

What 180 kg for three at RPE 9 is worth

Three at RPE 9 is one rep in reserve, so 180 kg was a 4-rep load that day. The model reads the reps-by-RPE chart along its diagonal, reps plus reps in reserve, and lands on 89.2 per cent of 1RM: an estimated max of 201.8 kg. Epley says 198 kg and Brzycki 190.6 kg, because both assume the third rep was the last one available. The estimated 1RM guide covers how the chart is read and where the Epley tail takes over.

The ladder below, from compute.mjs ladder --kg 180 --reps 3 --rpe 9, is the heaviest load for each rep count at RPE 10 on that day.

A true triple that day was worth 186.1 kg, so next week's 185 kg triple at RPE 9 asks for a little more strength than week one showed. The rule bets on that, and the rating settles the bet.

How the top set + back-off scheme moves the bar

The app has seven progression schemes: Auto, Double, Top set + back-off, Load, Reps, RIR and Manual. For a once-a-week heavy pull, Top set + back-off fits best. It reads the top set's RPE against a target and sets the back-offs at a fixed percentage of the top set, 88 per cent by default. The post on what RPE a top set should be covers how to choose that target.

With a target of RPE 9 and a 180 kg triple, the app's next top set depends on the rating you log.

The step on a barbell lift is 2.5 kg, but on a 180 kg pull the on-target adjustment rounds to a 5 kg jump. That is coarser than the bench sees, and it suits the deadlift, where a plate pair is a small fraction of the bar. A grinder holds the load instead of cutting the back-offs off a bad day.

The back-offs should land a few reps short of failure. 157.5 kg for five at RPE 7 is worth 200.4 kg against the top set's 201.8 kg, so a back-off rated that way agrees with the triple. If the back-offs rate much harder than that, the top set was rated too low.

Here is the same lifter over three weeks, each rating honest.

If week two had been 185 kg at RPE 10 instead, the estimate would read 200.7 kg and the app would hold at 185 kg. That is the point of rating the pull rather than adding a plate on a schedule. Across every scheme, a planned deload week never emits an increase, a load increase never leaks through on a cut, and three sessions without progress flags a stall by default. More than 30 days off resets the lift: the same 180 kg triple after 35 days away comes back as a 170 kg suggestion.

How often to deadlift heavy

Once a week is enough for the heavy pull, and the evidence on frequency says why. A meta-analysis of 22 studies found strength gains grew with sessions per week, with effect sizes of 0.74 for one and 1.08 for four or more, but the effect disappeared when weekly volume was equated [4]. The frequency effect for lower-body strength alone did not reach significance [4]. Frequency is mostly a way to fit in more sets.

The app's recovery read agrees with a weekly heavy pull. Readiness comes from two rolling windows, hard sets in the last 48 hours and recent load in the last 72 hours, never a calendar week. A heavy deadlift on Monday counts against the hamstrings and glutes until Thursday. A second lower-body day on Friday then starts fresh, which leaves room for squats or a lighter hinge without stacking fatigue on the pull.

If you want more hinge practice, add a lighter variation on another day, such as a Romanian deadlift. It keeps its own history, and the weekly top set stays the one number that steers the pull.

Mistakes that stall a deadlift

The most common mistake is logging the top set without RPE. Every pull then counts as a max, the estimate reads low, and suggestions start from the wrong place.

Another is changing the setup mid-block. A hexagonal bar let 19 powerlifters lift a heavier 1RM than a straight bar, 265 kg against 245 kg, and it shifted load from the lower back and hip toward the knee [5]. A trap-bar triple and a straight-bar triple are different lifts with different histories. Conventional and sumo are too.

Substitutions and what changes when you swap

A trap bar deadlift moves more weight with less demand on the lower back. The review of deadlift studies reported strong thigh activation and reduced erector spinae activation on the hexagonal bar [1]. A Romanian deadlift shifts the work toward the hamstrings, with lower erector spinae activation than biceps femoris and semitendinosus in the one study that measured it [1]. Both suit Double progression better than a top set, because they are usually run for sets of 6 to 10. Whatever you swap in starts its own history in the app. A 201.8 kg conventional estimate says nothing about a trap bar.

Questions

How much weight should I add to my deadlift each week?

Add load only when the top set is rated at or under its target. With Top set + back-off and a target of RPE 9, the app turns a 180 kg triple at RPE 9 into a 185 kg suggestion, a 5 kg step after rounding. A triple at RPE 10 holds the load. Expect weeks without a jump.

Should deadlift warm-ups count toward weekly sets?

No. The app never counts warm-ups toward volume. Five warm-ups and four working sets credit the hamstrings and glutes with four sets each, not nine. The warm-ups rehearse the pull and tell you how the top set will feel, which is their job. Log them with the warm-up role so the weekly total stays honest.