Readiness

What is driving the number

Each input as a z-score against your own baseline. The grey band is plus or minus half a standard deviation, the smallest change worth reacting to. Impact is that same z-score after its weight is applied, in points on the 0-100 score, renormalised for any input that has no data yet, so it says exactly how much of today's number that row is responsible for. Read the components before the score: when they disagree, the composite is hiding something.

Today's readiness of 48 (normal) is mainly a function of load balance (acute vs chronic), currently costing roughly 3 points off the score. To improve it in the coming days: if this came from a hard spike, a couple of easier days brings it back; if training has instead dropped off, resuming normal volume closes it. Resting HR (7-day mean) is the biggest thing working in your favour, adding roughly 2 points. HRV stability (CV) is also below baseline and worth the same attention.

Where you sit against your own normal

Your HRV coefficient of variation over the last 28 days is 21.6%, which reads as widening. A widening spread over a training block has been linked to maladaptation independently of where the average sits, so this is worth watching even when the mean looks fine.

Vitals

Body Battery

Garmin's 0-100 energy-reserve gauge, combining stress, HRV and activity: charges mostly overnight, drains through the day, rising at +2.5/30d. Useful as your own short-term trend, not a validated clinical measure.

Stress

Garmin's continuous 0-100 stress score, banded on its own published scale (0-25 rest, 26-50 low, 51-75 medium, 76-100 high), falling at -9.6/30d over the last 14 days.

Training load focus, this month

Monthly training load split by intensity type against Garmin's target ranges for your profile: a training-type balance check, distinct from the acute:chronic training-amount comparison under Readiness today. Proprietary model, not independently validated, so read the position as a nudge rather than a verdict.

Garmin's own label for this: "Aerobic High Shortage."

Training load

Load is Banister TRIMP, which weights minutes by how much of your heart rate reserve they used, so hard minutes count for far more than easy ones. The last few days are a sharp step up from your recent normal. Treat the ratio sceptically: the acute figure is contained inside the chronic one, which inflates the correlation, and the widely repeated 0.8 to 1.3 safe range has not survived reanalysis. The uncoupled difference avoids that specific flaw.

Are you getting fitter

"Improving" or "regressing" only applies when the 95% confidence interval on the slope excludes zero. Otherwise the data cannot rule out a flat line, which over a window this short is the common case, not the exception, so it is labelled "not significant" rather than guessing a direction from a noisy point estimate.

From Garmin's nightly sleep tracking, 16 nights in this window. Average 7.1 h, sleep score 76, stage mix 25% deep / 17% REM / 58% light of total sleep time.

Heart rate recovery

No recovery curves captured yet. Heart rate recovery needs the watch to keep recording after you stop working. Add a three-minute cooldown lap and leave the timer running, or press lap rather than stop. Once a few sessions have a tail, this section fills in.

Beat-to-beat HRV

No beat-to-beat data found. Your watch's optical sensor records averaged heart rate, not individual beat intervals, so the real HRV indices and the DFA threshold estimate cannot be computed from it. Pair a chest strap that transmits R-R (HRM-Pro, HRM-Dual, Polar H10) for one ride and this section appears, with an aerobic threshold estimate from the beat series itself.

Lifting sessions

Sets, reps and volume from Hevy (automatic, internal API); heart rate from the watch where a Garmin activity matched the same start time within 45 minutes.

Weekly tonnage across 5 weeks, oldest first.

Sessions showing no watch data were logged without the watch recording, and still count toward weekly volume.

Weekly set volume by region

Weekly set count per muscle group is the volume landmark most closely tied to growth, with a graded dose-response up to roughly ten or more sets per muscle per week (Schoenfeld, Ogborn & Krieger, 2017). The upper and lower split here is keyword matching on exercise names, so treat it as approximate. For upper-body mass specifically, this table and the tonnage chart predict progress far better than anything on the heart rate panels above: HRV tells you whether your whole system recovered, not whether a given muscle is ready to be loaded again.

Other lifting sessions (not logged in Hevy)

These are strength sessions the watch detected that have no matching Hevy log, so sets and reps here are Garmin's wrist-motion guess rather than what was actually on the bar. Be honest about what heart rate can and cannot tell you too: the readiness metrics above were developed on endurance athletes, and HRV reflects whether your whole system has recovered, not whether a specific muscle group is ready to be loaded again.

Each row splits that measure's own logged history into an earlier and a more recent half and runs a proper significance test (Welch's t-test, which does not assume equal variance between the two halves) on the difference, rather than comparing two averages and guessing whether the gap is real. "Not significant" does not mean nothing changed; it means the data so far cannot rule out that it didn't. Weight, body fat and lean mass are shown without a colour judgement either way, since which direction counts as progress depends on your goal, not on the number itself.

Lifting volume per session (kg), earlier vs recent half of history

Volume by time of day, whole history

Average volume per session in each time window, with the percentage above or below that type's overall average and how many sessions fall in the window. "Best window" only names one when it beats every other window combined by a real margin (Welch's t-test), not just whichever bin happens to have the highest of several noisy averages; a session type shown here without a named best window has not shown a real time-of-day effect yet.

Placement against reference populations for age 19, 190 cm, 85 kg. The confidence column is doing real work: only some of these measures have normative data collected the same way yours was. Two have none that applies, and they are listed separately rather than given a number that would look just as authoritative as the rest.

Cardiorespiratory

Body

Where a percentile would mislead

• Overnight HRV: Overnight averaging is not comparable to short supine recordings, and between-person HRV differences are mostly constitutional. Use your own baseline instead.