Why does Base show every lift separately instead of one strength score?

Because pooling hides the case that matters. Three lifts climbing while one has been flat for a month is a specific, fixable problem, and an average renders it as fine. Every lift gets its own row and its own trend for exactly that reason.

What counts as a stalled lift?

A flat or falling estimated one-rep max trend over several weeks while you are training it consistently. One session down is noise. Four weeks with no upward movement is a signal, and the usual causes are volume that is too low, recovery that is insufficient, or a rep range that has drifted.

Where do I see the strength trend?

Progressive Overload, from Settings or from the Workout tab's progress tiles. The Strength screen inside it gives every lift its own ninety-day estimated one-rep max trend, and the Workout tab shows tiles for your most frequently trained lifts.

Track Whether You Are Actually Getting Stronger

Q: How does Base estimate my one-rep max?

The Epley formula: weight multiplied by one plus reps divided by thirty. It is only meaningful up to about twelve reps, so above that Base returns nothing rather than a number it does not believe. A set of twenty at a light weight tells you about endurance, not maximal strength, and the formula badly overestimates there.

Q: Why is estimated one-rep max better than total volume for tracking strength?

Because it does not move when you add a set. Tonnage rises trivially by doing more sets of a lighter weight, which is a dose increase rather than a performance improvement. Estimated one-rep max only moves when you lift more weight or get more reps at the same weight.

Why estimated one-rep max

Epley:  e1RM = weight × (1 + reps ÷ 30)

225 × 5   →  225 × 1.167 = 262 lb
225 × 8   →  225 × 1.267 = 285 lb
185 × 12  →  185 × 1.400 = 259 lb

Above about twelve reps the formula overestimates badly, so Base returns nothing rather than a number it does not believe. A set of twenty tells you about muscular endurance; it does not tell you about a maximum.

The reason to prefer this over tonnage is that it does not move when you add a set. Total volume load rises trivially by doing more sets of a lighter weight, which is a dose increase rather than a performance improvement — useful for checking weekly volume, useless for answering "am I stronger".

The screens

Progressive Overload. The four cards are in causal order rather than menu order — strength is the leading indicator, volume the input that drives it, recovery the limit on usable volume, and energy what pays for all of it.

Progressive Overload is the hub, and its ordering is an argument rather than a list:

Strength — the leading indicator
Weekly volume — the input that drives it
Recovery — what limits how much volume is usable
Where calories go — what pays for all of it

If strength is flat, that chain is the order to check things in.

Strength gives every lift its own row with a ninety-day trend and a percentage change per week.

The Workout tab also carries tiles for your most frequently trained lifts — "main" meaning most logged, since nothing in the data marks a lift as an anchor and frequency is the closest honest signal for the lift a programme is built around.

Reading a trend

Pattern	Reading
Rising steadily	Working. Change nothing.
One session down	Noise. Sleep, timing, a bad day.
Flat for 3–4 weeks with consistent training	A signal worth acting on
Falling while bodyweight rises	Usually recovery or technique rather than the programme
Sawtooth around a level	Load being pushed past what the rep range supports

The thing to resist is acting on one session. Estimated one-rep max is far less noisy than tonnage but it is not silent — it moves with sleep, with time of day, and with how honestly the last set was rated.

What to check when a lift stalls

In order, because they are ordered by how often they are the cause:

Is it being trained hard enough? Open weekly volume and look at that muscle. A lift stalling at six counted sets a week is under-dosed rather than plateaued.

Are the sets rated? Unrated sets do not count toward volume and give the progression engine nothing to work with, so the prescription stops adapting.

Is recovery keeping up? A flat trend during a heavy endurance block is the recovery budget being spent elsewhere, which is the entire premise of tracking both halves together.

Is the rep range right? A lift living at the top of a 12–20 range is being trained for something other than maximal strength, and its estimated one-rep max will reflect that honestly.

Is it actually a new lift? A lift that forked into a duplicate — typed as a custom entry when the library already had it — starts from zero with no history. That looks exactly like a stall and is not one.

Rate of change

The trend carries a percentage per week, and calibration matters here.

Strength gains slow enormously with training age. A novice can add several percent a week; someone five years in is looking at a fraction of a percent, and a flat month is unremarkable rather than alarming. The general tile averages across lifts and reports how many are rising, which is a more robust read than any single lift's slope.

Common mistakes

Reading one session as a trend. Four weeks is the shortest window worth a decision.

Comparing your rate of change to a novice's. Training age dominates everything.

Chasing estimated one-rep max with singles. The formula is a projection from sets in a normal working range. Grinding heavy singles to move the number is optimising the measurement rather than the thing.

Treating a deload dip as a stall. Load drops during a deload. The trend picks back up.

Ignoring a stall because the average looks fine. This is exactly what the per-lift view exists to prevent.

Forking a lift into a duplicate. Check the suggestions when adding an exercise; a lift typed fresh starts with no history and no progression state.

What to do next

Open the Strength screen and find your flattest lift. Check its counted weekly volume before changing anything else — under-dosing is the most common cause of a stall and the easiest to fix.