What if every workout you completed could be read as a 10-dimension physiological profile? WOD Analyzer is a free, open-source pedagogical tool that turns any CrossFit, Hyrox, or weightlifting WOD into a structured physiological signature across 3 energy systems and 7 neuromuscular capacities. No artificial intelligence — purely deterministic parsing and physiological rules calibrated from the exercise science and programming literature.
The pitch: making the physiological profile of a WOD explicit
Every workout carries an implicit physiological signature. Fran (21-15-9 thrusters + pull-ups) is a predominantly glycolytic workout with a significant explosive power component. Murph (1-mile run, 100 pull-ups, 200 push-ups, 300 air squats, 1-mile run) is overwhelmingly oxidative with a sustained strength-endurance demand. Most athletes program their weeks without ever quantifying these dimensions — and imbalances accumulate silently.
WOD Analyzer parses the raw text of any WOD (free-format, French or English), detects movements via a catalogue of 94 movements (weightlifting, powerlifting with squat and deadlift variants, CrossFit gymnastics and advanced strict gymnastics, mono-structural cardio, Hyrox, hardstyle/sport kettlebell, and an expanded strongman section of 14 movements: atlas stone, yoke walk, log clean & press, farmer handles, tire flip, heavy sled, axle deadlift, sandbag carry, Husafell carry, Conan’s wheel, keg clean & press, sandbag over yoke, duck walk, frame carry), identifies the format (AMRAP, EMOM, EXMOM, For Time, Chipper, Tabata, RFT, Strength, Hyrox, Death by), and computes in milliseconds a 10-dimension profile displayed as a Recharts radar, horizontal bar charts, and an auto-generated text synthesis.
How it works: the hybrid model
The calculation engine is built on a hybrid energy systems × neuromuscular capacities model, with each movement carrying its own parameters in the catalogue.
The 3 energy systems
The model describes the energetic contribution of each session across three physiological systems:
| Energy system | Time window | Typical profile |
|---|---|---|
| ATP-PCr (phosphagen) | 0–10 s, maximal effort | Snatch, clean & jerk, box jump, sprint < 10 s |
| Anaerobic glycolytic | 10 s – 2 min, lactate “burn” | Thrusters, burpees, double-unders at high cadence |
| Oxidative (aerobic) | > 2 min, VO2max and endurance | Row 2,000 m, 1-mile run, ski erg Hyrox |
The 7 neuromuscular capacities
Beyond energy systems, the model quantifies seven neuromuscular dimensions:
- Maximal strength — loads > 85 % 1RM, high CNS demand (heavy deadlift, 1–3 rep squat)
- Explosive power — ballistic, Olympic, and plyometric movements (clean, snatch, box jump)
- Strength endurance — moderate loads, high volume (thrusters, wall balls, pull-up sets)
- VO2max / cardiorespiratory endurance — efforts > 3 min at high intensity (row 2 km, repeated 400 m)
- Lactate capacity — 30 s – 2 min maximal intervals (power EMOM, Tabata)
- Core stability — anti-rotation, proprioception (GHD sit-ups, L-sit, plank)
- Technical skill / coordination — complex movements requiring fine motor control (muscle-up, pistol squat, double-under)
The movement catalogue
Each entry in the catalogue pairs a movement with: its dominant energy system, secondary capacities, energy system distribution (three percentages summing to 100), typical load, average time per repetition, and French/English regex aliases for automatic detection. The catalogue covers Olympic weightlifting, powerlifting (paused squat, pin squat, RDL, deficit deadlift, good morning), CrossFit gymnastics and advanced strict gymnastics (strict pull-up, strict HSPU, ring dip, bar dip, assisted muscle-up, chest-to-bar, toes-to-bar, knees-to-elbows), mono-structural cardio (rowing, ski erg, assault bike, echo bike, air bike, running), Hyrox-specific movements (farmers carry, sled push/pull, sandbag lunges, wall walk, broad jump, lateral lunge, wall balls…), kettlebell (swing, clean, snatch, clean & jerk, thruster, lunge, push press) and an expanded strongman section of 14 movements (atlas stone, yoke walk, log clean & press, farmer handles, tire flip, heavy sled, axle deadlift, sandbag carry/over-shoulder, Husafell carry, Conan’s wheel, keg clean & press, sandbag over yoke, duck walk, frame carry).
Scoring
Each dimension score is computed from the cumulative time per movement (reps × time per rep) weighted by format intensity. A 20-minute AMRAP amplifies the oxidative component; a power EMOM (10 rounds × 10 s maximal effort) amplifies the ATP-PCr system and explosive power. The text synthesis describes in natural language the dominant profile, secondary capacities, and any dimensions that are absent or under-represented.
Special formats: Death by and EXMOM
Beyond the classic formats (AMRAP, EMOM, For Time, Chipper, Tabata, RFT, Strength, Hyrox), WOD Analyzer now recognises two advanced formats with very specific physiological signatures.
Death by — progressive EMOM to failure
The Death by format is a progressive EMOM: minute 1 requires 1 rep, minute 2 requires 2 reps, minute 3 requires 3 reps, and so on until the athlete can no longer complete the quota within the allotted time. The physiological signature depends on the breakdown minute: an average athlete typically holds 10–12 minutes on burpees, 7–9 on heavy squat clean, and up to 14 on light air-squat-type movements.
WOD Analyzer automatically estimates the average-athlete breakdown point based on the movement profile (light cardio, basic gymnastics, kettlebell, heavy weightlifting, strongman…) and computes the total volume with the triangular sum formula: total reps = cap × (cap + 1) / 2. A 3-to-25-minute slider lets you adjust the breakdown point manually to personalise the scoring. The format strongly drives lactate capacity and strength endurance.
EXMOM — Every X Minutes On the Minute
The EXMOM format generalises EMOM: instead of a fixed one-minute window, the athlete has X minutes to complete the prescribed work, using the remaining time as recovery. An E2MOM clean & jerk, for example, requires 3 reps in 2 minutes — about 30 s of effort and 90 s of recovery, an ATP-PCr / strength-power dominant signature.
Three syntaxes are recognised: E2MOM, E3MOM, Every 2 minutes, Toutes les 3 minutes… (minutes only, X ≥ 2). If the total duration or number of rounds is not specified, WOD Analyzer applies 10 rounds by default. A 3-to-20-round slider lets you adjust the volume. Scoring favours maximal strength, explosive power, and the ATP-PCr system, with attenuation of glycolytic and oxidative components thanks to the long recovery windows.
Two reading levels
WOD Analyzer is designed to serve two distinct use cases:
🏋 CrossFit / Hyrox athlete
Paste your WOD, get its profile in seconds. Use the session history to accumulate multiple sessions, compare radars, and detect imbalances: “12 glycolytic sessions this month, 0 pure maximal strength.” The A4 PDF export lets you share your profile with your coach or add it to your training journal.
🎯 Coach / Sports professional
Use the 42 benchmark WODs (18 Girls, 16 Heroes, 8 Open) to calibrate your programming against established references. The radar comparison “My WOD vs benchmark” objectively quantifies relative intensity. The weekly programming module (5–6 session sliders, 48 h maximal strength / intense glycolytic separation) produces an exportable landscape A4 PDF plan.
Beyond single-session analysis
WOD Analyzer integrates several features that go beyond simple profile calculation:
Recovery recommendations
A deterministic recovery block is generated from an energy system × intensity table, with specific bonuses for core stability demand and CNS loading in sessions with a high maximal strength component. Recommendations cover the optimal recovery window (12 h – 72 h depending on profile), nutritional priorities, and active recovery modalities.
42 benchmark WODs and radar comparison
The 42-benchmark database includes the 18 Girls (Fran, Grace, Helen, Isabel…), 16 Heroes (Murph, DT, Randy…), and 8 CrossFit Open workouts. Each benchmark has been pre-analysed and profiled. The radar comparison superimposes your WOD on these references, making it possible to verify, for instance, whether your “home Fran” is genuinely comparable to the original in terms of glycolytic demand.
Weekly programming
The programming module provides sliders from 5 to 6 sessions per week. An algorithm enforces fundamental physiological constraints: a minimum 48 h separation between sessions dominated by maximal strength and intense glycolytic blocks. For each time slot, 2–3 candidate WODs from the benchmark catalogue are proposed with their profiles displayed. The plan is exportable as a landscape A4 PDF for printing or sharing.
Session history and JSON export
The session history stores in-memory all WODs analysed during the same browser session. Multiple selection (up to 4 sessions) enables a multi-session comparative radar and aggregated statistics that automatically flag imbalances. JSON export/import provides persistence across browser sessions — export before closing the tab, import at the next use.
For sports professionals and STAPS educators
WOD Analyzer is designed as an educational support tool for coaching and teaching contexts. Use it in athlete discussions to show the physiological signature of their monthly training load, or in exercise physiology courses to illustrate energy system concepts interactively. The dual reading level — athlete-friendly output alongside full physiological decomposition — makes it equally suitable for individual athlete feedback and for classroom or workshop settings.
Important: WOD Analyzer analyses the prescription, not the execution. The profile generated corresponds to a standard Rx-level athlete. It does not account for individual variability, scaling choices, or actual execution tempo. It is a pedagogical tool, not a substitute for the eye and experience of a qualified coach.
Key takeaways
- Hybrid model: 3 energy systems × 7 neuromuscular capacities — 10 dimensions computed from a 94-movement catalogue (weightlifting, powerlifting, CrossFit and advanced strict gymnastics, cardio, Hyrox, kettlebell, strongman expanded to 14 movements) with full physiological parameters.
- Automatic format detection — AMRAP, EMOM, EXMOM, For Time, Chipper, Tabata, RFT, Strength, Hyrox, Death by, each weighted differently in the scoring.
- Special formats Death by and EXMOM — Death by (progressive EMOM to failure, 3–25 min breakdown slider) and EXMOM (Every X Minutes On the Minute, 3–20 round slider, default 10 rounds).
- 42 benchmark WODs built-in — 18 Girls, 16 Heroes, 8 Open. Radar comparison available for each benchmark.
- Weekly programming with PDF export — algorithm enforcing 48 h strength/glycolytic recovery constraints.
- 100 % frontend, open-source — no backend, no data collection, no account required. Runs in any modern browser.
Mini French → English glossary for in-app terms
The WOD Analyzer interface is in French. Here is a quick reference for every term you will encounter on screen:
| French (in-app) | English equivalent | Context |
|---|---|---|
| Filière dominante | Dominant energy system | The primary energy system solicited (> 50 % of total energy demand) |
| Filière ATP-PCr / Phosphagène | ATP-PCr / Phosphagen system | 0–10 s maximal effort; explosive and Olympic movements |
| Filière Glycolytique | Anaerobic glycolytic system | 10 s – 2 min; lactate “burn” characteristic of MetCons |
| Filière Oxydative / Aérobie | Oxidative / Aerobic system | > 2 min; VO2max and endurance efforts |
| Force maximale | Maximal strength | Loads > 85 % 1RM; high CNS demand |
| Puissance explosive | Explosive power | Ballistic, Olympic, plyometric movements |
| Endurance de force | Strength endurance | Moderate load, high volume (thrusters, wall balls) |
| Capacité lactique | Lactate capacity | Maximal intervals 30 s – 2 min |
| Gainage / Stabilité | Core stability | Anti-rotation, proprioception (GHD, L-sit, plank) |
| Skill / Coordination | Technical skill / Coordination | Complex movements: muscle-up, pistol squat, double-under |
| Récupération recommandée | Recommended recovery | Optimal rest window before next training session |
| Benchmark | Benchmark WOD | Girls, Heroes, Open workouts used as reference comparisons |
| Historique | Session history | In-memory log of all WODs analysed in the current session |
| Export PDF | PDF export | A4 printable report including radar, tables, synthesis, and recovery |
| Programmation hebdo | Weekly programming | 5–6 session weekly plan with physiological constraint enforcement |
Limitations and scientific honesty
WOD Analyzer is a didactic model, not a digital twin of your physiology. Several deliberate simplifications deserve explicit naming:
- The model analyses the prescription, not the execution. The same 21-15-9 thrusters and pull-ups performed by an elite athlete and a beginner mobilise different energy systems — the tool ignores this individual variability and assumes a standard Rx level.
- Actual intensity depends on execution. Pacing, transition management, and scaling choices radically alter the real physiological profile. WOD Analyzer only perceives the written prescription.
- 94 movements still do not cover every case. Atypical WODs (rare movements, unrecognised abbreviations) may be partially or incorrectly parsed. The interface displays which movements were detected, providing full transparency.
- Parameters are calibrated from the literature, not from direct measurement. Time per repetition, energy system distributions, and typical loads come from a synthesis of the programming and exercise physiology literature — they represent population tendencies, not individual values.
- Educational tool, not a decision-making system. WOD Analyzer does not replace the eye and experience of a qualified coach or a formal physiological assessment.
What makes the tool useful is precisely its transparency about these limitations: detected movements are listed, scoring is decomposed movement by movement, and nothing is a black box.
Open source — contribute and build on it
WOD Analyzer is built with React + Vite + TypeScript, Tailwind CSS + shadcn/ui, Recharts for radar charts, and jsPDF + html2canvas for PDF exports. The application is fully open-source at github.com/JETPOD/wod-analyzer. 100 % frontend, no backend. The only data stored locally (localStorage) is your bodyweight, used to compute the theoretical %1RM — never transmitted anywhere. Pull requests, bug reports, and movement catalogue additions are all welcome.
If you are a developer, coach, or researcher interested in extending the movement catalogue, adding new benchmark WODs, contributing a translation, or adapting the scoring model for a specific discipline, the repository is your starting point.
v7 update — bodyweight-relative load scaling
Version 7 addresses a limitation flagged by users: a 60 kg back squat does not produce the same physiological demand for a 55 kg athlete as it does for an 85 kg athlete. WOD Analyzer now factors in bodyweight to estimate relative load and to recalibrate the maximal strength scoring.
Global bodyweight field
A new « Mon poids (kg) » (“My bodyweight (kg)”) field appears above the WOD input box. Entered once, it is persisted locally in the browser (localStorage) and automatically feeds all subsequent calculations. No server transmission — the value stays on your device.
Theoretical %1RM via reference ratios
Each loaded movement in the catalogue (49 entries: weightlifting, powerlifting, kettlebell, strongman) is assigned a reference ×BW ratio derived from the strength-sport literature: 1.5 × BW for back squat, 1.8 × BW for deadlift, 0.85 × BW for thruster, 1.8 × BW for atlas stone, 2.5 × BW for yoke walk, 0.45 × BW for kettlebell swing, and so on.
Theoretical %1RM is then computed with a three-tier priority:
- Explicit %1RM — if you write “85 % 1RM” in the WOD, that value is used as-is;
- Direct %BW — syntaxes
@1.2BW,1.2xBW,80 % BWare recognised automatically and divided by the movement’s reference ratio; - kg + bodyweight + ratio — a 90 kg back squat for a 70 kg athlete yields 90 / (70 × 1.5) = ~86 % 1RM, classified as maximal strength; the same 90 kg for an 85 kg athlete drops to 71 % 1RM, classified as strength endurance.
Maximal-strength scoring recalibrated on 5 tiers
The bonus applied to the maximal strength dimension is now modulated across five tiers based on the estimated %1RM:
| Estimated %1RM | Scoring impact |
|---|---|
| ≥ 95 % | Max strength ×1.4 + skill bonus (maximal load, demanding execution) |
| ≥ 85 % | Max strength ×1.0 (classical neuromuscular zone) |
| ≥ 70 % | Max strength ×0.5 + strength endurance ×0.2 (mixed zone) |
| ≥ 50 % | Strength endurance ×0.3 (moderate volume) |
| < 40 % | Strength endurance ×0.5 (light load, high reps) |
Rep-counting bug fix
The format detector now recognises the variants N rounds, N tours, N séries, N sets without requiring the “for time” suffix. Previously, “4 rounds of 5 squats” could be mis-parsed and produce an incorrect rep count. Counting is now robust: 4 rounds × 5 reps = 20 reps, as expected.
Enriched per-movement display
Each loaded movement now displays the triple kg · %BW · ~%1RM: for example “5 back squat · 90kg · 129%BW · ~86%RM · ~25s”. The tilde signals that this is a theoretical estimate derived from the reference ratio, not an individual measurement.
🇫🇷 Lire cet article en français : WOD Analyzer : décodez la physiologie de vos entraînements
NutriCellScience, Mark DOWN — EN edition
Laisser un commentaire