Where the model comes from
Baseload is grounded in the sports-science literature on the load ↔ injury relationship — in particular the work on the acute:chronic workload ratio (ACWR, Gabbett 2016 & 2020). We say "grounded in" deliberately: this work describes population-level trends and is still debated within the scientific community. We don't claim it is "validated" or that it guarantees an individual outcome. These are solid reference points to guide a decision, not laws.
What the model can do
- Measure the effort of each session (TRIMP), across every sport, with or without heart-rate data.
- Track the balance between recent load and underlying fitness (ACWR) and warn you when it rises too fast.
- Distribute fatigue by muscle group and estimate your recovery, group by group.
- Cross-reference these signals with your HRV, your resting HR and your sleep to adjust the day's recommendation.
- Explain everything in plain language: you always see why.
What it approximates — and what we own
Muscle load is a heuristic
The distribution of effort across muscle groups relies on empirically calibrated expert coefficients, not on a measurement of your actual activation. The literature (EMG) can validate an order and a magnitude — not an exact figure per muscle. It's deliberately a good order of magnitude, refinable by your feedback, not a laboratory truth.
No age adjustment on recovery
Muscle recovery rates are the same for every athlete. Yet a master athlete, in reality, recovers more slowly. This simplification is cautious on the load side (it never shortens estimated recovery) and owned for the beta. (Maximum heart rate, on the other hand, is adjusted for age.)
Downhill cost is not modelled separately
In trail running, it's the descent — not the climb — that marks the legs the most. Baseload accounts for elevation in a conservative way (it tends to over-count rather than under-count), but it doesn't yet treat downhill braking as a distinct damage channel. It's an identified work item, not an oversight.
What it doesn't do
Baseload doesn't predict an injury and doesn't guarantee that you won't get hurt. An injury depends on factors the app can't see: technique, gear, actual sleep, stress, terrain, history, chance. Baseload reduces one known risk — load spikes — it doesn't remove the risk. It never compares you to other users, and never plays on guilt or streaks.
Not a medical device
Baseload is a training-support tool, not a medical device. It makes no diagnosis and replaces neither a doctor, nor a physiotherapist, nor a professional coach. In case of pain that persists, an injury or any doubt about your health, consult a professional. The details are in our terms of use.
Everything stays on your iPhone
On a health product, trust can't be declared — it shows in the architecture:
- Your health data stays on your device. No raw heart rate, no GPS track, no full workout is sent to a server.
- The coaching is computed locally, on your iPhone — including the weekly review.
- No account, no mandatory email. An anonymous technical identifier is enough.
- No advertising-tracking SDK, no reselling of data. Audience measurement is anonymous and off by default.
Full details in our privacy policy.
Saying what we can't do isn't a weakness — it's the condition for coaching you can trust. We prefer honest, explainable advice to a promise no one can keep.