Estimated 1RM and rep bands

You cannot compare 185 pounds for 8 reps with 225 pounds for 5 reps by eye. Logex converts every set to the same unit first: the heaviest single you could probably lift, estimated from the set you actually did. That estimate is the e1RM, and it is the common currency of the app.

In plain terms. Each rep you could do beyond one is worth about a thirtieth of the weight on the bar. Five reps at 225 is roughly the same effort as a single at 262.5.

The formula

Logex uses the Epley formula:

e1RM = weight × (1 + reps ÷ 30)

It lives in calculateEstimatedORM and is repeated, with the same constants, in the Live Set Coach (epley), the exercise-history rules and the Progress hub, so every screen agrees.

Worked example

Set	Calculation	e1RM
225 lb × 5	225 × (1 + 5 ÷ 30) = 225 × 1.1667	262.5
185 lb × 8	185 × (1 + 8 ÷ 30) = 185 × 1.2667	234.3
135 lb × 15	135 × (1 + 15 ÷ 30) = 135 × 1.5	202.5
300 lb × 1	300 × (1 + 1 ÷ 30) = 300 × 1.0333	310.0

Two things to notice. First, a single is estimated slightly above the bar weight (a factor of 1.0333), so a “true” 1RM and the e1RM are never exactly equal. Second, because every set is converted the same way, the 185 × 8 set scores lower than the 225 × 5 set even though it has more reps. That is the point.

Rep bands

Logex also buckets sets by how many reps they have. The bands are used to keep baselines honest, because your estimated strength from sets of 12 is not the same as from sets of 3.

Band	Reps
Low	1 to 5
Mid	6 to 10
High	11 to 15 (and anything above 15 is also High)

A set with no valid rep count has no band. The band is stored on each set (rep_band), and per-band baselines are described in Set quality, baselines and badges.

A stabilised session number

One lucky set should not define a session. Alongside the best-set e1RM, Logex stores a stabilised session e1RM: a weighted average of every eligible set’s e1RM, where each set is weighted by its reps, capped at 10.

sessionE1RM = Σ(e1RM_i × min(reps_i, 10)) ÷ Σ min(reps_i, 10)

Heavy low-rep sets and light high-rep sets both contribute, but a 15-rep set counts no more than a 10-rep set.

Worked example

Sets: 225 × 5, 225 × 5, 205 × 8, 135 × 15.

Set	e1RM	Weight (min(reps, 10))
225 × 5	262.5	5
225 × 5	262.5	5
205 × 8	259.7	8
135 × 15	202.5	10

Weighted sum is 262.5×5 + 262.5×5 + 259.67×8 + 202.5×10 = 6,727.3. Total weight is 28. The stabilised e1RM is 6,727.3 ÷ 28 = 240.3, well below the best set (262.5), which is exactly what stabilising is for.

Which sets are allowed in

Which sets feed an e1RM depends on what the number is used for.

• Records, PRs, rep records and the Live Set Coach: any complete set that is not a Warmup, Warmdown or Dropset.
• The saved per-session e1RM and quality score: only Working, Volume, AMRAP and 1RM sets. The app’s own documentation gives the reason: if a PR Attempt or Backoff set were counted there, it would raise the baseline for later sessions and lower their quality. Example from the docs: 200 × 5 followed by a 245 × 1 PR Attempt saves 233.3, not 253.2.

A last “Working” set that is much lighter than the earlier ones is probably a manual back-off. If the final eligible set is typed as Working and its e1RM is at or below 72% of the peak of the earlier eligible sets, it is dropped from the quality calculation, the same as if you had labelled it Warmdown.

Rep records (best weight for each exact rep count, shown up to 12 reps) use the same eligibility as PRs.

Effort-aware variants

The headline e1RM never uses RPE, because self-reported effort should not be able to create a record. But planning code wants a better answer to “what could I do today?”, so it uses versions that add the reps you had left in reserve. RPE is stored from 1 to 10 in tenths (8.5 is valid, 8.25 is not), and reps in reserve is taken as 10 − RPE.

Where	How the estimate is built
Live Set Coach capacity	weight × (1 + (reps + RIR) ÷ 30), RIR = 10 − RPE, set to 8 RPE if blank
Warm-up learning	same Epley, but RIR is limited to 0 to 3
Daily strength and Peak planning	reps plus RIR (limited to 0 to 3, total reps capped at 10), then the mean of Epley and Brzycki
Set-to-set fatigue learning	Epley with RIR limited to 0 to 3

Brzycki is weight × 36 ÷ (37 − reps). The planner averages it with Epley and ignores any set above 10 reps when predicting a peak, because both formulas drift at high reps. When the effective reps come out to exactly 1, the actual load is used as is.

Worked example

You did 200 lb × 5 at RPE 8, so RIR = 2 and effective reps = 7.

Epley   = 200 × (1 + 7 ÷ 30)      = 246.7
Brzycki = 200 × 36 ÷ (37 − 7)     = 240.0
planner = (246.7 + 240.0) ÷ 2     = 243.3

Limits and edge cases

• Epley is a population average. It is more reliable between about 3 and 10 reps. A set of 20 will overstate your max, which is why Peak planning ignores sets above 10 reps and the stabilised e1RM caps each set’s weight at 10 reps.
• Two rep-band definitions exist. The stored rep bands above (1 to 5, 6 to 10, 11 to 15) feed baselines and quality. The planner uses its own coarser split when it picks “comparable” past sessions: up to 6 reps is low, up to 12 is medium, and above that is high.
• Bodyweight and assisted lifts use the load actually moved, not the number you typed: a share of your bodyweight plus any added weight, or minus the assistance. At 180 lb, a pull-up with 25 lb added moves 0.95 × 180 + 25 = 196 lb, so 8 reps is an e1RM of 248.3. Without a bodyweight no e1RM is made up, and progress is tracked by the added weight, the assistance or the reps instead (see The exercise catalogue, bodyweight and assisted loads).
• Machine lifts are treated like any other weight and reps. The estimate says nothing about how a machine’s cam compares with a barbell.
• RPE is only as good as your honesty. That is why effort-aware numbers only ever shape suggestions, never badges or records.