J Metrics, explained
Review recorded load and missing data without confusing an estimate with measured fitness.
Explore →AdvanzedRacing · JasMiamiMethod
Your race. Your evidence. Scenarios you can inspect.
Bring your race, athlete references, course and conditions into one workspace. Change an assumption and see what the calculation depends on.
1. Give the scenario a real race
Choose one of your saved races or enter an event, date, location and distance. Keep an official event link for your own review. A goal time stays a goal; the planner does not treat it as evidence of your current capacity.
Open the scenario workspace →2. Inspect the evidence behind the numbers
Review available athlete references with their source and measurement date. Add missing observations explicitly. Running, swimming and cycling need appropriate sport-specific evidence; an undated number or a value from another sport is not silently substituted.
Review race evidence →3. Make the course visible
Import a GPX file or enter a distance–elevation profile. Inspect the profile, climbs and descents before calculating a scenario. Course data is an input to review, and GPS or elevation errors can affect the result.
Review a course →4. Change assumptions, compare scenarios
Explore entered temperature, weather and execution assumptions. Request available event-time weather or use clearly labeled manual conditions. Save dated scenario snapshots and compare their inputs and calculated outputs. These calculations have not been validated as individual race predictions.
Explore conditions →5. Keep execution tied to observations
Inspect the pace or power outputs the available evidence can support. Heart-rate references require explicit observed inputs. Missing evidence stays visible, and unsupported outputs remain unavailable. A scenario does not replace medical advice, course rules or a coach's review.
Review execution assumptions →6. Keep fueling practice separate
Record practiced carbohydrate intake, when you tried it and whether you tolerated it. Review this context alongside the scenario. A calculated finish time does not establish your tolerance or automatically authorize a higher intake. The separate HYROX tool remains available for run and station assumptions.
Read the fueling guide →How to read a scenario
A scenario shows the result of explicit assumptions. Individual JMM accuracy has not been measured on independent races. Comparing conditions does not produce a statistical prediction interval or establish what will happen on race day.
| Input | Use in the scenario | What remains unknown |
|---|---|---|
| Athlete references / Referencias del atleta | Dated sport-specific observations, with their source and context. Undated or unsuitable records remain visible as gaps. | Whether a test represents sustained race effort, recent training tolerance and event-specific durability. |
| Course / Recorrido | A reviewed GPX or manually entered distance–elevation profile, including signed climbing and descending segments. | GPS/elevation error, surface, turns, traffic, technical handling and event rules. |
| Conditions / Condiciones | Provider weather when available for the selected event time, or explicit manual assumptions. Temperature and other conditions can be edited. | Local exposure, individual heat response and changes after the forecast was retrieved. |
| Execution / Ejecución | Inspectable scenario arithmetic from supported inputs, with missing-data reasons. Pace, power and heart rate are distinct quantities. | Individual prediction accuracy, race-day readiness and the sustainable intensity of the chosen scenario. |
| Fueling / Nutrición | Practiced carbohydrate intake and dated tolerance context, kept separate from performance calculations. | Race-specific tolerance, products, access to aid stations and individual clinical requirements. |
Validation still needed
The workspace saves scenarios with their inputs, timestamp and calculation version. Saving a scenario does not validate the model. Later evaluation must use only information available before the race, compare against actual results and keep model-development athletes and events separate from evaluation data.
Publish error in minutes and percentage, bias and interval coverage by sport, distance and population. Show sample size, missing inputs and validation date. Compare against simple baselines. Do not promise a universal accuracy percentage or infer faster race times from HRV or JStress without validation.
AI should explain evidence, assumptions and options; calculations and revisions should be reproducible. Future integration needs authorized data, provenance and provider limits, not simply more data.
More of JMM worth exploring
Evidence you can inspect
Explore sources, applicable populations and limitations. A published principle is not proof of your exact training dose.
Explore →A coach and an athlete, working together
Review profiles, discuss changes and track results through authorized coaching access. Device connection and workout delivery are separate checks.
Explore →JMM is free during the limited-account pilot. External services may have their own fees and limits. Account access is required for athlete tools. No guaranteed performance, injury prevention or watch delivery is claimed.