Working 1RM suggestions

When the 1RM behind your %1RM loads can come down, and why it only ever asks.

Two numbers per lift

Your all-time best is the estimated 1RM from your heaviest set. It sets your PRs and never goes down.

Your working 1RM is the number %1RM loads are worked out from. It starts at your best. A training max you type in always wins and is never touched. See Equations.

When a suggestion appears

Barbell Bench PressLower your 1RM for now?1RM now263 lbSuggested240 lbYour last 3 hard sessions (Oct 16, 23 and 29) point to about 233 to 240 lb. Using 240 lb makes your %1RM weights match what you can lift right now.

Only for lifts whose loads come from the working 1RM, and only in two cases:

Three hard sessions in a row came in at least 7% under it, within 6 weeks. A hard set is 10 reps or fewer at RPE 7 or harder. Warm-ups, drop sets, Off-day workouts and deload weeks do not count. The suggestion is the highest of the three estimates, never more than 10% lower in one step, never under 80% of your best, and at most one per 6 weeks.
A break of 4 weeks or more from a lift you trained at least 3 times in the 12 weeks before. 4 to 8 weeks off suggests 5% lower; 8 weeks or more, 10% lower.

You decide

Nothing changes until you tap Use. Keep puts the suggestion away for 8 weeks (a break suggestion, for that break). Every change is listed in the 1RM history on the exercise page, with Undo.

Any later session that beats the lowered number raises it straight back. No question asked.

Why it says so

Tap Why on a card for the sessions behind it: date, hardest set, estimate. The one the number comes from is marked.

Where to find them

Stats: a Suggested changes section under the period tabs. Swipe between cards.
Home: "You have new suggestions" until you have seen them.
The exercise page: the same card.

In Progression settings, Working 1RM suggestions turns them off. Lower 1RM without asking applies them for you, with history and Undo.

Why the rules are cautious

1RM test and retest vary by about 4% (Grgic et al., 2020), so 7% clears ordinary day-to-day noise.
Three sessions in a row rules out one or two bad days.
RPE is most accurate near failure (Zourdos et al., 2016; Helms et al., 2017), hence RPE 7 or harder.
Rep-based 1RM estimates loosen past 10 reps (Reynolds, Gordon and Robergs, 2006).
Strength holds for about 3 to 4 weeks off, then fades (McMaster et al., 2013; Bosquet et al., 2013).