One-Rep Max Calculator (Epley/Brzycki) mechanism

Model scope

The two named 1RM formulas are correct under their normal assumptions. Verdict is NEEDS_REVIEW because the exact table computation is underspecified: the spec must name the selected 1RM basis, define whether output is per-set load or total volume, and specify rounding. The honest-label text is appropriate but should be surfaced prominently, especially for Brzycki estimates above roughly 10–12 repetitions.

• The Epley formula, 1RM = weight × (1 + reps/30), is standard for sub-maximal rep estimates when weight and reps are valid.
• The Brzycki formula, 1RM = weight × 36/(37 − reps), is standard, but its usable rep range must be explicitly constrained; it is undefined at 37 reps and unsuitable for high-repetition estimates.
• The %1RM training-table calculation is not explicitly specified. The standard implied calculation is training load = estimated 1RM × percentage, but the spec does not state which selected 1RM estimate, percentage representation, or rounding policy is used.
• The meaning of an entry such as 5×5 at 80% is ambiguous: it should specify whether the output is the per-set load, the total session volume, or both.

Units and conventions

• Weight is a mass value in one consistent unit; the estimated 1RM and training loads use that same unit.
• Reps are a nonnegative integer count of completed repetitions.
• Percentages are interpreted as fractions for arithmetic: 80% = 0.80, 75% = 0.75, and 90% = 0.90.
• Test-vector table loads assume training load = selected estimated 1RM × percentage, with no rounding unless stated.
• A 5×5 notation is interpreted as 5 sets of 5 repetitions; the calculated load is the per-set load, not total volume.
• Kilograms and pounds must not be mixed without explicit conversion.

Inspectable vector

One of 4 source-bound review vectors:

{
  "inputs": {
    "weight": 100,
    "weight_unit": "kg",
    "reps": 5,
    "formula_choice": "Epley",
    "table_percentage": 0.8,
    "sets": 5,
    "reps_per_set": 5
  },
  "expected_outputs": {
    "estimated_1RM": 116.6666667,
    "estimated_1RM_unit": "kg",
    "table": {
      "scheme": "5x5",
      "percentage": 0.8,
      "load_per_set": 93.3333333,
      "load_unit": "kg"
    }
  },
  "working": "1RM = 100 × (1 + 5/30) = 100 × (1 + 0.1666667) = 100 × 1.1666667 = 116.6666667 kg. Training load = 116.6666667 × 0.80 = 93.3333333 kg per set."
}

Failure boundary

BREAK_CLAIMED · SELF_FLAGGED

The Brzycki formula has an unhandled division-by-zero input at exactly 37 reps. The calculator must reject reps = 37 rather than emit Infinity, NaN, or an invalid training table. Negative-weight interpretation is also ambiguous unless input validation or an absolute-magnitude convention is specified.

Other records

Sources and evidence

These pointers bind the page to evidence. They do not grant authority or collapse verification into deployment.

• VERIFIED_LEDGER.json
• evidence-tranche4/tranche4-results.json
• BOUNDARY_HAZARD_CATALOG.json