A training plateau, in the sense Zyra Workouts uses on its dashboard, is an exercise you have trained at least three times in the last 30 days whose estimated 1RM has moved less than 2% from the first of those sessions to the most recent. Whether it is time to deload depends on a second number the app reads from the same sets: the average RIR you logged. Flat with reps in reserve is a push signal; falling with nothing in reserve is a deload signal. Here are the exact rules and what each one is asking you to do.

Target 610 @ RIR 2 · ceiling = round((6+10)/2) + 2 = 10 · step 2.5 kgReps above the ceiling (10) → progressadd one step (+2.5 kg) and start the range again at 660 kg × 11 → 62.5 kg × 6Reps from 6 to 10 → holdsame load; the rep target stays inside the range60 kg × 8 → 60 kg × 8Reps under the floor (6) → back offtake one step off (−2.5 kg) and aim for 6 again60 kg × 5 → 57.5 kg × 6No history → no recommendation. Failure sets (RIR 0): ceiling = top of the range (10).
Fig. The progression rule the app applies between sessions, for a 6–10 @ RIR 2 target on a barbell (step 2.5 kg). The ceiling is the midpoint of the rep range plus the RIR target — here round((6 + 10) ÷ 2) + 2 = 10; sets taken to failure (RIR 0) use the top of the range instead. Beat the ceiling → add a step; fall under the floor → take a step off; anything between → hold. Dumbbells and stacks step by 2 kg, bodyweight moves progress by reps. A deterministic rule, not a coaching judgement.

How Zyra Workouts detects a plateau (the exact rule)

For each exercise, the dashboard gathers the sessions from the last 30 days, keeps the working sets (warm-ups excluded), and computes per session the best estimated 1RM and the average RIR (from the RPE you logged, converted with 10 − clamp(round(RPE), 6, 10)). Then it compares the first session in the window with the most recent:

pct = (e1RM_last − e1RM_first) / e1RM_first × 100

and labels the exercise:

ConditionLabelDirection
fewer than 2 sessions in the window0%flat (not enough data)
pct ≥ +2+x% / N wksup
pct ≤ −2 and average RIR ≤ 1Deload neededdown
pct ≤ −2 and average RIR > 1−x% / N wksdown
−2 < pct < +2 and 3 or more sessionsPlateauflat
−2 < pct < +2 and exactly 2 sessions0%flat (too early to call)

N wks is the span between the first and last session, rounded, minimum 1. The exercises shown are the ones you have logged the most working sets on in the window, so a lift you did once does not crowd out your staples.

Two properties of this rule are worth holding onto. It compares endpoints, not a fitted line, so a bad first session in the window can make the next four weeks look like progress and a great first session can make them look flat — the sparkline beside the label shows the shape. And the 2% band is in *estimated 1RM*, not in weight on the bar: on a 100 kg bench, one extra rep at 8 → 9 is a 3.6% e1RM change and clears the band, while 100 → 102.5 kg at the same reps is 2.5%. Reps move this number as much as plates do.

In plain terms: three or more sessions in a month, and the best set's e1RM has not moved 2% either way — that is a plateau.

Deload or push

The plateau label on its own does not say which way to lean. The RIR does.

Flat with reps in reserve — say the label reads Plateau and your working sets have been logged at RIR 2–3. The numbers are not moving because the sets stop before they have to. The double progression ceiling for a 6–10 @ RIR 2 target is 10 reps; if you have been racking at 8 with two left, you have never given the engine a reason to add weight. The honest move is to push the reps toward the ceiling on the next session and let the rule do its job. Nothing is wrong; the effort has been below the target.

Falling with nothing in reserve — the label reads Deload needed: e1RM down 2% or more and RIR averaging 1 or less. You have been taking sets to within a rep of failure and the numbers have still gone backwards for a month. Pushing harder is not available; there is no harder. This is the pattern the back-off rule cannot fix one step at a time, and the one the fatigue model on the same dashboard is designed to surface early.

Falling with RIR above 1 — the label shows the negative percentage without the deload text. Something has changed — a new variation, a rushed warm-up, less sleep, a diet — but the effort log says there was room in the sets. Look at what changed before you change the program.

This is arithmetic on your log, not a trial result and not a diagnosis of anything. The rule cannot see a stressful month or a sore shoulder; it can only see that reps and RIR did what they did. If a pattern persists and something feels wrong beyond tired, that is a question for a coach or a qualified professional, not a dashboard.

What a deload week looks like

What the app does about a deload is deliberately small. A program has a deload position — none, first cycle, or last cycle — and the marked cycle is flagged as the lighter one. The source is explicit that this is a *marker only, no fabricated maths*: the app does not silently cut your sets or your loads. What you do in that week is yours to set, so here are the two levers that exist and what each one changes:

LeverExampleWhat stays the sameWhat the log shows
Fewer sets, same load4 × 82 × 8 at the same weightthe weight and the e1RM of each setweekly set count halves; e1RM unchanged
Higher RIR target, same sets@ RIR 1@ RIR 3 at the same set countthe set count and rep rangelighter loads or fewer reps per set; e1RM dips for the week
Both, gently4 × 8 @ RIR 13 × 8 @ RIR 2the exercisesa smaller version of both effects

A useful check after the week: the 30-day trajectory will still include the deload sessions, so expect the label to lag. The signal that the week did its job is the first full session back — the same weights at a higher logged RIR than before, or the reps past the ceiling again. If the numbers come straight back, you were carrying fatigue. If they do not move, the plateau was a programming question (range, volume, exercise choice), not a recovery one, and the deload was still cheap to run.

In plain terms: the app tells you which cycle is the light one; you decide whether light means fewer sets or more reps in reserve, and the log tells you afterwards whether it worked.

The short version

Zyra Workouts calls it a plateau when an exercise's e1RM has moved under 2% across 3+ sessions in 30 days, and calls for a deload when it has fallen 2% or more with RIR averaging 1 or less. Flat with reps to spare: push to the ceiling. Falling with nothing left: take the light cycle — fewer sets or a higher RIR target for a week — and read the first session back. The e1RM behind all of this is explained in how to track strength progress. Zyra Workouts is in development and not yet on the App Store; the rules quoted here are read from its source code and may change before release. Training guidance, never medical advice. Educational overview only — not medical advice.