PACER's training methodology: turning running science into an executable plan
PACER uses training science to create personalized race plans. Daniels VDOT, Hansons cumulative fatigue and 80/20 intensity distribution inform its calculations and planning alongside your real training data, goals and available training time.
PACER combines real training data and sports-science calculations. Plans can change through conversation and weekly reviews; a data sync does not rebuild the plan each time.
1. Daniels VDOT and E/M/T/I/R pacing
VDOT, introduced by running coach Jack Daniels, is a measure of your current effective running fitness derived from a recent race or all-out test. With a VDOT value, training intensity can be translated into five precise pace zones instead of vague instructions like "take it easy" or "go faster":
| Zone | Name | Purpose |
|---|---|---|
| E | Easy | Aerobic base, capillary and mitochondrial development, recovery |
| M | Marathon pace | Adaptation to goal race pace |
| T | Threshold | Improve lactate clearance, delay fatigue |
| I | Interval | Develop maximal oxygen uptake (VO₂max) |
| R | Repetition | Improve running economy and speed |
PACER estimates your VDOT from multiple sources in priority order (manual anchor > real all-out test/race > device race prediction > pace-plus-heart-rate estimate > VO₂max fallback), and maps each training day's target pace directly onto E/M/T/I/R.
2. Hansons cumulative fatigue
The core idea of the Hansons method is "cumulative fatigue": rather than relying on single monster long runs, you keep training on legs that are never fully recovered, so you can resist fatigue in the late stages of a race. In practice this means a comparatively restrained long-run cap, volume spread across the week, and careful control of single-session intensity. PACER applies this constraint when designing weekly load and long-run share — avoiding the classic mistake of "blow up on a weekend long run, do nothing all week".
3. 80/20 polarized training
A large body of endurance research and elite practice points to the same distribution: about 80% of training volume at low intensity (easy aerobic) and only about 20% at moderate-to-high intensity. The low-intensity base builds your aerobic engine; a small dose of high-quality stimulus drives the breakthroughs. The most common amateur mistake is exactly "too much moderate intensity" — not easy enough to recover, not hard enough to improve. PACER applies this ratio across periodization and keeps hard days from being adjacent where possible, leaving room to recover.
4. Supporting models: fatigue quantification, progression and nutrition
The core science stack: Daniels VDOT (E/M/T/I/R pace zones), Hansons cumulative fatigue, 80/20 polarized training, and Banister TRIMP fatigue quantification.
- Banister TRIMP: converts each session's duration and heart-rate intensity into a "training impulse" to quantify fatigue and fitness, preventing overtraining.
- Higdon's 10% rule: weekly mileage generally increases by no more than ~10% as a conservative planning guideline — it is a heuristic, not a guarantee against injury.
- ACSM nutrition & hydration: combines sweat rate and session duration into post-run hydration and fueling guidance.
5. The three-layer decision model: composing theory into daily training
PACER wraps the science above in a three-layer decision pipeline that uses synced data for conversational adjustments and weekly reviews:
- ① Multi-dimensional state sensing — reads body and environment signals: running watch data sync (COROS, Garmin via intervals.icu, Polar), temperature and environment adjustments, sleep and HRV recovery monitoring.
- ② Core computation engines — makes decisions with scientific models: Daniels VDOT for current fitness, Banister TRIMP for fatigue, runner injury medicine for risk alerts.
- ③ Dynamic plan output — produces executable daily training: weekly adaptive plans, pre-race taper, precise hydration and nutrition.
Because decisions are built on your real data, PACER keeps one rule: when information is missing it says so and asks you to sync — AI advice is based on your authorized data and may still be wrong, so use your own judgment for important decisions.