Six weeks of 6-12-25, logged perfectly, and no idea if you got stronger.

Six weeks of 'completed' and no answer

Six weeks into a 6-12-25 block, the log is perfect. Every session is there. Every set has a weight and a rep count, and every workout carries the same 'completed' mark.

Ask it the only question that matters, did you get stronger, and it goes quiet.

The tonnage line wobbles. Estimated 1RM on the 25-rep leg extension produces a number nobody would ever test. The heavy six went up ten pounds in week three, but the set of twelve right after it was a grinder, and the log has no way of saying so. You did the work. You wrote it down. And the record still can't tell a productive block from a six-week wash.

The one-set-one-number assumption

Call it the one-set-one-number assumption. Nearly every logging app models a set as a single unit: one load, one rep count, maybe one RPE. Sets stack into sessions, sessions sum into volume, and the chart draws a line through the totals. For 5x5, that model holds up. For most straight-set programming it's fine too.

Mixed-rep protocols shatter it. A 6-12-25 tri-set hits one muscle with a heavy six, a moderate twelve and a light twenty-five, back to back. 1-6 contrast work alternates a heavy single with a set of six. Rest-pause clusters chop one working set into mini-sets separated by a few breaths. Every one of these produces sets that don't mean the same thing, and the moment a tracker flattens them into one number, the progress thread snaps.

The log is complete. It just isn't comparable.

Why tonnage and e1RM go blind on mixed-rep blocks

The standard advice for reading a log is to watch volume climb and watch estimated 1RM climb with it. On a mixed-rep block, both mislead.

Tonnage is dominated by whichever range moves the most total weight, and in a 6-12-25 that's rarely the heavy six. Twelve reps at a moderate load, or twenty-five on a machine stack, can easily outweigh it. So the week you push the six harder, leave less in reserve, and have to shave ten pounds off the twenty-fives can show up as a volume drop. The chart punishes your best session.

Estimated 1RM does no better. Rep-max formulas were built around low-rep sets and lose their footing far from a max, so whatever a 25-rep set produces is noise. And a single session RPE blends a near-limit six with a burning twenty-five into one figure that describes neither.

The problem isn't missing data. It's missing context.

What the RPE research actually hands you

Effort ratings in lifting were designed at the set level from the start. The RPE scale most serious lifters use ties each rating to reps in reserve, and the version studied by Zourdos and colleagues in the Journal of Strength and Conditioning Research (2016) was built around rating individual sets. The tool assumes you'll attach it to the set it describes.

There's a second wrinkle. Research on how well lifters predict their remaining reps suggests accuracy shifts with rep count and with proximity to failure Prediction accuracy of repetitions to task failure slightly improved when predictions were made closer to set failure (β=-0.025). ([source](https://www.researchgate.net/publication/351730507_Accuracy_in_predicting_repetitions_to_task_failure_in_resistance_exercise_a_scoping_review_and_exploratory_meta-_analysis)). If a heavy six and a long set of twenty-five don't carry equally reliable ratings, blending them into one number throws away the part you could trust.

And lifters are clearly running these protocols. 'The 6-12-25 rule gym' appears as a People Also Ask query in current Google results, alongside other numbered-rule methods. The demand is there. The logs weren't built for it.

One block, two readings

Take an illustrative leg day: a lifter running 6-12-25 for quads, with hack squat for the six, leg press for the twelve and leg extension for the twenty-five. Week 1 and week 4 are in the table below. The numbers are a worked example, not measured data.

Read it the way most trackers do and week 4 is a small regression. Total tonnage fell 130 pounds. Both sessions are marked done. Nothing flags.

Now read it by range, with effort attached. The hack squat went up 20 pounds at the same RPE 8. That's a strength gain, and it's the one the block was supposed to produce. The leg press held its load and dropped from RPE 9 to 8, so identical work got easier. The leg extension lost 10 pounds, but it was taken closer to failure after two harder front-end sets.

Same log. Opposite conclusion. The only thing separating them is effort recorded at each range.

Same log. Opposite conclusion.

What a readable mixed-rep log has to hold

If the thesis holds, the fix is small and specific. Every range gets its own RPE. Not the session, not the exercise: the range. A 6-12-25 round is three effort ratings. A rest-pause cluster is one per mini-set, or at minimum one on the first and one on the last.

Then the comparison changes shape. You stop asking whether the session total went up and start asking three narrower questions. Did load at a given RPE rise in the heavy range? Did the middle range get easier at the same load? Did the long set hold up after harder front-end work? Each of those is answerable across a block. A session total isn't.

It also changes what a stall looks like. When the heavy six sits at the same load and RPE for three weeks while tonnage keeps creeping up on the back sets, that's a plateau hiding under volume, and any chart built on totals will miss it.

Where the instrument comes in

Platepusher records effort at the set level Per-set RPE entry is not available in the current build. The app previously supported RPE-based auto-weight prescriptions on plan sets, but that UI was deliberately removed (data fields kept for future use), so there's no way today to log or view an RPE value per set, including within tri-sets or rest-pause clusters., so a 6-12-25 round lands in the log as three sets with three ratings instead of one smeared session. The heavy six is the natural top set to watch, and plateau detection is built around exactly that: it flags the week your top set stops moving.

The larger point goes back to the argument in Your training history is the experiment. A multi-year log is the N=1 study you've already run on yourself. Mixed-rep blocks are some of the most informative weeks in it, because they push effort in ways straight sets never do. They only pay off in retrospect if the record can be read.

Without effort context at each range, six weeks of intensification is a list of sessions. With it, you have a dataset.