If You Run 5/3/1: Which Parts Sterka Keeps, and What a Spreadsheet Cannot Do

If you have run 5/3/1, you already know most of how Sterka’s Get Stronger goal works. Jim Wendler’s program has been the default answer to “what should I run” for well over a decade, and the reason is structural: it picks a working number below your real max, loads percentages off that number, and puts one rep-out set per cycle in charge of calibrating it. That is a good design. Sterka’s strength engine evolved from it rather than around it.

So this is a map, not a pitch. Here is each piece of the template, what the engine does with it, and the four things it reads that a spreadsheet has no way to see.

What is a training max, and does Sterka use the same one?

Yes, and it is seeded the same way: 90% of your one-rep max, floored to the nearest plate. The difference is what it takes 90% of. The classical approach takes 90% of a max you tested. Sterka takes 90% of an estimate read off submaximal sets you left two reps short of failure, which under-reads a true max by roughly 3 to 7% on its own. Two discounts stack instead of one, so a freshly seeded anchor sits further below your real ceiling than a tested-max seed would. That gap is widest at the seed and narrows as your rep-outs earn the number up. The full arithmetic is its own post, and you can run your own number in the training max calculator.

The vocabulary carries over unchanged. Training max, percentages off it, a rep-out that drives the next cycle, a reset when the number stops being true. Nothing here needs relearning.

How does the four-week wave work?

Three weeks that build, then a deload. The length is fixed at four weeks, same as the template.

Where it diverges is the rep emphasis. In 5/3/1 the fives, threes and the single all live inside one cycle. Sterka gives each its own block and alternates them, so a full turn is two waves, roughly eight weeks. An intensity block runs triples, then doubles, then the single. A volume block runs fives throughout and peaks on a submaximal 3+ set at 88% of the training max instead of a near-max one. Block one is a volume block, block two is intensity, and it alternates from there.

An intensity block’s three loading weeks look like this, as fractions of your training max:

The 1+ is one prescribed rep as the floor, then as many clean reps as you have, stopping one short of failure. Underneath it sit two sets of five at 70%: the First Set Last pattern, volume banked under the heavy work rather than a second heavy effort.

Week four is where the map stops being a translation. In the template, the deload is another row of percentages. In Sterka the wave stands down entirely for that week: no wave percentages are authored at all, and the ordinary load author runs instead under a deload ceiling of about 54% of your estimated one-rep max, which is where a 60%-of-training-max week lands once the 90% seed is accounted for. Volume halves at the same time, and the cap reaches every loadable compound rather than only the barbell work, so the pulldown comes down with the squat. More on why that week is the plan working.

What about Boring But Big?

BBB is the most-run supplemental in the template: five sets of ten at 50 to 60% of your training max, bolted under the main work, with the percentage moved across cycles and other assistance cut back to pay for it. Sterka does not ship it as a named template, but it does bank supplemental volume under the heavy work for the same reason BBB exists.

Be clear about the size of the difference, because it is not small. BBB is around fifty supplemental reps. An intensity block’s First Set Last back-off is two sets of five at 70%, which is ten. Classical FSL is itself 5 × 5; Sterka ships two sets, not five. If BBB is what has been driving your size on top of the main lifts, an intensity block delivers a fraction of that dose and you should expect to feel the difference.

The reason is what the two blocks are for. Sterka splits accumulation and expression into separate blocks rather than running them in the same four weeks, so the volume block is where the fives live and the intensity block deliberately banks only what fits under a near-max single. On Build Muscle, where growth is the goal rather than a supplemental to it, weekly volume per muscle is what the engine steers on directly.

The AMRAP set drives the progression

AMRAP is as-many-reps-as-possible, the + on the last set of a 5/3/1 week. It is the piece the template gets most right, and Sterka runs it for the same reason: the rep-out is not a test bolted onto the program, it is the feedback loop. Two divergences are worth knowing, and both cut against the template’s cadence rather than for it.

The increment is conditional here. In 5/3/1 the training max goes up every cycle as a matter of course, and the rep-out informs the reset rather than gating the rise. Sterka makes you earn it: beat the prescribed rep by two or more and the training max climbs, by 5 kg on a lower-body lift and 2.5 kg on an upper-body one, never past what the set actually demonstrated. Hit the prescribed rep without a surplus and it holds. Miss it and the training max steps down to 90% of itself, the classical reset, so the next wave resumes at a weight you can move.

The rep-out is a week-three event, not a weekly one. If you are used to a + set every week, weeks one and two here are prescribed reps at prescribed loads. Do the arithmetic and it is slower on paper, so here it is. Across one turn, classical 5/3/1 puts 5 kg on a lower-body training max twice over, unconditionally: 10 kg. Sterka’s ceiling for the same eight weeks is 5 kg earned on the intensity block plus 2.5 kg earned on the volume block, whose 3+ at 88% feeds a deliberately junior nudge of its own. That is 7.5 kg, and only if both are earned. The trade is the one the template already admits to: 5/3/1’s increment rises whether or not you earned it, which is exactly why it needs a ×0.9 reset to catch back up with you.

Worked example: a set of 102.5 kg × 5

Weeks one to three are the same numbers the wave planner runs. Week four is the one place the two can part company: the planner draws 60% of your training max, while the engine caps the week against your estimated max instead, so once a demonstrated training max has climbed the cap is the lower of the two.

A set of 102.5 kg for five reps estimates a one-rep max of 115.3 kg. 90% of that is 103.8, floored to the nearest plate for a training max of 102.5 kg. Off that anchor the block runs 82.5, 87.5 and 90 kg in week one, then 85, 90 and 95 in week two. Week three asks for the 1+ at 97.5 kg, then two sets of five at 72.5.

That top single sits around 80% of a true max the program never asked you to test, which is the floor of the zone current ACSM guidance treats as productive for strength. The stop rule is one rep in reserve, and the reps you leave are what make the next block’s number accurate.

One thing that arithmetic does not say: 80% is where the number starts, not where it stays. A freshly seeded training max carries the widest discount it will ever carry. As your rep-outs demonstrate the load, the training max climbs toward a ceiling of 92% of the estimate that single itself implies, and the week-three weight climbs with it. Block four asks for a genuinely heavier single than block two does, against a smaller reserve. Report your reps in reserve honestly and the number tracks you. This is the one input the engine cannot check for you: it reads what you logged, not what you had left. Log a set as RIR 1 that was really a grind and you are handing the engine a training max your next peak week has to cash.

What a spreadsheet cannot do

Everything above a spreadsheet can hold, and plenty of lifters run 5/3/1 out of one for years. Four things it cannot, because they need the log read back:

Price fatigue per muscle, and act on it. Sterka books every hard set as fatigue on the muscles that did the work, full cost to a primary mover, half to a secondary, and half again more when an isolation lift has a single target and nowhere to hide. How close to failure the set went and how heavy the axial loading was both scale it. It decays on a half-life that differs by muscle, from 36 hours for side delts to 72 for the lower back. When a day’s anchor muscle is too far down, the session gets rebuilt rather than printed. A cell in a spreadsheet has no way to refuse.

Read the week that went badly. A run-down week, a stretch of poor sleep, a block where the working sets stopped moving: these show up in what you logged before they show up in how you feel about it. The engine has triggers that read those logs and can overrule the plan, including deloads the wave did not schedule. A running block normally holds those back on the reasoning that its own week four is coming. Stop advancing through it and that reasoning expires: the block stops suppressing the evidence, and the back-off can arrive mid-block.

Handle ten days off without starting over. Come back from a layoff and a template resumes exactly where the cursor was. Sterka reads the gap, eases the load back and lets the rep-out re-establish the anchor rather than assuming the old number survived the break.

Grade progress across blocks, not just within one. The reset that matters most is the slow one. When your best real lift stays below the heaviest working weight a block guarantees to prescribe, graded a rep short and with the rep-out itself left out of the reckoning, for two consecutive blocks, the training max comes down 10%. That is not a missed rep on one day. It is a number that stopped being true, and noticing it takes a memory longer than a cycle.

Where the engine is more cautious than the template

Three places, and they do not all work the same way.

The health screen comes first, and it is not lift-specific. Certain answers on the intake screen mean the engine prescribes no near-failure work at all, from any block. That is not only a rep-out swapped for a straight single: on the session’s main lift the load itself comes down into the back-off band, so the heaviest axial set of your week is a back-off rather than a peak. What the app tells you names no condition and no body part: it points you at your screening and your clinician, not at a diagnosis.

The rep-out itself is withheld on spine-loading free-weight lifts inside your first twelve logged strength weeks. Week three still arrives as a fixed heavy single at the same weight, so the loading is unchanged and only the open-ended part is gone. A gym without safety bars is stricter than that: on a lift where a failed rep lands on an unspotted bar, every set holds two reps in reserve, and the week-three peak can be routed to a variant the gym can fail safely.

Losing the rep-out to your gym or to your first twelve logged weeks does not freeze your number: finishing the set shows the weight was makeable without showing what you had left, so the block’s heaviest set, rated, earns the junior step rather than the full one — 2.5 kg on a lower-body lift, 1.25 on an upper — and only where nothing else is holding the lift. Leave it unrated and it earns nothing, because a fixed set that stops at the prescribed rep cannot tell “that was everything” from “I racked it with five in the tank”.

Where your health screen or your injury notes is what holds the test, the number does freeze: that hold is a judgement about you rather than about your gym or your logged weeks, so the training max stays where it is until near-max work is back on your plan. It outranks the other two rather than sitting beside them — a lifter under the twelve-week floor who has also flagged something at screening earns nothing for a clean set. And it is not the only reason that freezes it: any day the engine was already easing this lift, coming back from a layoff or an illness, reads the same way.

And the wave itself is held back for a self-described beginner’s first 28 days, which for most readers of this post will not apply. If you set up as intermediate or advanced, you are on a wave as soon as there is logged training to build one from. The month is a starting lifter’s carve-out, not a universal waiting period.

None of that makes the template wrong. 5/3/1 works because its structure is sound, and most of that structure is intact here. What changes is the number of inputs the structure is allowed to read: not just the reps you hit on the top set, but every set behind it.

References

The load zone cited above, and nothing else on this page, rests on:

• American College of Sports Medicine position stand. Resistance training prescription for muscle function, hypertrophy, and physical performance in healthy adults: An overview of reviews. Medicine & Science in Sports & Exercise, 2026.