Training

The data density of modern GPS watches has created a paradox for endurance athletes: more information than any previous generation of athletes has had access to, and less clarity about which numbers to act on. Garmin, Polar, and COROS devices now generate metrics including training load, training readiness, body battery, VO2max estimate, lactate threshold estimate, race time predictions, recovery time, acute/chronic workload ratios, and real-time running dynamics — all from wrist-based sensors with varying evidence bases and error margins. Understanding which metrics have scientific validation, which are proprietary black boxes, and which are outright noise allows athletes to use technology as a genuine performance tool rather than an anxiety-generating data feed.

This guide focuses on the metrics with the strongest evidence base and clearest actionability — the ones that change what you do in training based on what the number tells you.

Heart Rate: The Foundational Metric

Heart rate remains the most validated wearable metric after decades of research. Optical wrist-based heart rate is accurate enough for steady-state aerobic training (error margin ±3–5 bpm at below 150 bpm), but degrades significantly during high-intensity interval work where wrist movement creates optical artefact (error ±15–30 bpm at above 170 bpm). Chest straps remain the gold standard for interval sessions and race effort. Key heart rate metrics:

• Resting HR trend: Track resting heart rate daily (most watches measure during sleep). Resting HR elevated by 5+ beats above your personal baseline for 2+ consecutive mornings signals accumulated fatigue — reduce training load. Chronically low resting HR (cardiac drift downward over months) signals improving aerobic fitness. This is the single most actionable daily metric available from wrist-based devices.
• Training zones: Heart rate zones based on lactate threshold or maximum heart rate provide the intensity structure for most endurance training systems. Verify zone boundaries with a ramp test or field test rather than accepting age-based formulas (HRmax = 220 − age has a standard deviation of ±12 bpm — individual variation makes this formula unreliable for training prescriptions).
• Heart rate decoupling (aerobic decoupling): The degree to which heart rate "drifts" upward while pace stays constant on a long easy run. Less than 5% drift over 60+ minutes indicates good aerobic efficiency. More than 10% drift suggests you were running too hard, the conditions were hot, or aerobic base fitness is limiting. A useful metric for calibrating Zone 2 effort that most watches do not display directly but can be calculated from workout data.

HRV (Heart Rate Variability): Powerful but Context-Dependent

HRV — the variation in time between consecutive heartbeats — is the strongest validated biomarker of autonomic nervous system recovery status available from wearable devices. High HRV indicates parasympathetic dominance (recovered); low HRV indicates sympathetic dominance (stressed or fatigued). Interpreting HRV requires personal baseline context — absolute values vary enormously between individuals (healthy resting HRV ranges from 20 to 100+ ms depending on age, fitness, and measurement method). Only changes relative to your own baseline are meaningful:

• Consistent suppression (3+ days below your baseline band): Meaningful signal to reduce load. Not acted on by most athletes, which is why HRV data is collected but not applied in most recreational athlete contexts.
• Measurement standardisation: HRV must be measured at the same time each day (morning, immediately on waking, before standing) in the same position (supine) for day-to-day comparison to be valid. Inconsistent measurement protocols produce noise that swamps the signal.
• Confounders: HRV is suppressed by alcohol (even 1–2 drinks), poor sleep, illness, significant psychological stress, and altitude. Knowing the confounder prevents false alarm responses to training load.

Estimated VO2max and Race Predictions

Watch-based VO2max estimates (primarily from Firstbeat Analytics algorithms used by Garmin and others) are validated against laboratory measurements with reasonable accuracy (±3.5 mL/kg/min in research conditions) — sufficient for tracking trends over weeks and months, not sufficient for absolute performance planning. The race time predictions derived from watch VO2max estimates assume optimal pacing and current fitness is reflected in recent training data — they are reasonable estimates, not precise predictions. Use them as trend indicators rather than definitive targets. When your watch predicts marathon time improves from 3:48 to 3:41 over a 12-week block, that trend is meaningful even if the absolute number is imprecise.

Running Dynamics: Ground Contact Time, Vertical Oscillation, and Stride Length

Running dynamics metrics require chest-strap or footpod accessories to be accurate — wrist accelerometers produce high error in these measurements. With accurate measurement hardware, the actionable metrics are:

• Ground contact time balance: The split between left and right foot contact time. Values above 51.5%/48.5% asymmetry indicate a compensatory gait pattern that elevates injury risk on the overloaded side. Most useful for runners managing asymmetric injuries.
• Vertical oscillation: Below 8cm is excellent; 8–10cm is good; above 12cm warrants review of cadence and forward lean. High vertical oscillation wastes energy on non-horizontal movement.
• Stride length trends: Declining stride length at the same pace (requiring more cadence to maintain speed) can indicate fatigue accumulation over a training block — a useful early indicator before subjective fatigue becomes obvious.

Use the NorthLine Training Load Calculator alongside your GPS data to track your acute-to-chronic workload ratio — the validated injury risk metric that combines GPS-derived training stress with time-based recovery windows. Monitoring both GPS metrics and load metrics together provides a more complete picture than either alone.