Heart Rate Zones: Useful Guide, Not a Verdict — A Research-Grade Synthesis

Abstract

Heart rate is one of the most-measured and most-misread signals in indoor rowing. The peer-reviewed literature is clear: heart rate rises with intensity, drifts upward over time at constant work, and responds to temperature, hydration, sleep, stress, medication, and sensor error ([4] Coyle & González-Alonso 2001, Level 5; [13] Périard, Racinais & Sawka 2015, Level 5; [14] Périard et al. 2016, Level 5; [1] Tanaka, Monahan & Seals 2001, Level 2b; [3] Karvonen, Kentala & Mustala 1957, Level 2b). The most useful framing for a rower is that heart rate is a useful guide, not a verdict — a noisy read on the body's acute cardiovascular response that, paired with perceived effort, breathing, and stroke-rate cues, lets you pace a session intelligently ([21] Scherr et al. 2013, Level 2b; [22] Reed & Pipe 2014, Level 5). The modern zone models — five-zone, three-zone, and lactate-threshold-anchored — agree on the broad shape but disagree on the exact numbers, and the agreement-to-disagreement gap is exactly where an indoor rower lives ([5] Seiler & Kjerland 2006, Level 4; [25] Stöggl & Sperlich 2014, Level 2b; [24] Binder et al. 2008, Level 5; [23] Wasserman & McIlroy 1964, Level 5; [15] Ingham et al. 2002, Level 2b; [26] Lamberts et al. 2010, Level 2b; [11] Plews et al. 2017, Level 2b; [12] Plews et al. 2014, Level 2b). The honest read: heart rate zones are a calibration tool, not a calibration of physiology, and the rower who anchors on the talk test, RPE, and stroke-rate cues alongside the monitor is reading the signal correctly.

What heart rate actually measures on a rower

Heart rate is the number of times the heart contracts per minute, driven by the balance of sympathetic (accelerator) and parasympathetic (brake) nervous-system input. On the indoor rower, heart rate rises with each stroke as the working muscles demand oxygen and the heart ejects more blood per beat and beats more often to meet the demand. The 1964 [23] Wasserman & McIlroy paper introduced the framing that heart rate — together with ventilation and gas exchange — marks the transition from fully aerobic metabolism to a state where anaerobic energy production is required, and the modern convention is that this transition is not a single line but a transition zone ([23] Wasserman & McIlroy 1964, Level 5; [24] Binder et al. 2008, Level 5).

The practical upshot for an indoor rower is that heart rate is an integrative signal: it reflects the demand of the current stroke, the cumulative demand of the session so far, the body's temperature and hydration state, how well the rower has slept, what they have eaten and drunk, whether they are ill, and how calm or stressed they are mentally. The 2001 [4] Coyle & González-Alonso review established the mechanism: cardiovascular drift — the gradual rise in heart rate at constant work rate over 30 to 60 minutes — is driven by falling stroke volume and rising core temperature, with progressive dehydration amplifying the drift ([4] Coyle & González-Alonso 2001, Level 5). The 2002 [15] Ingham et al. paper put a rowing-specific number on the relationship: in world-championship finalists, power at VO2max, VO2 at lactate threshold, power at 4 mmol/L lactate, and peak power together explained 98% of 2K speed variance, and heart rate at these landmarks is what the rowing monitor reports ([15] Ingham et al. 2002, Level 2b).

The five-zone model and where it comes from

The five-zone model is the most familiar intensity framework, and the modern peer-reviewed default anchors on the 2001 [1] Tanaka, Monahan & Seals HRmax formula: HRmax = 208 minus 0.7 times age, derived from a meta-analysis of 351 studies covering 18,712 adults ([1] Tanaka, Monahan & Seals 2001, Level 2b). The formula replaced the older 220-minus-age rule, which had no peer-reviewed basis and systematically over-predicted HRmax in older adults. The 2010 [2] Gulati et al. paper added a female-specific formula — HRmax = 206 minus 0.88 times age — for the cases where the Tanaka formula mis-fits ([2] Gulati et al. 2010, Level 2b). The 1957 [3] Karvonen, Kentala & Mustala paper added the heart-rate-reserve (HRR) method, which is more personalised: target HR = (HRmax minus HRrest) x intensity plus HRrest, where intensity is the proportion of HRR being targeted ([3] Karvonen et al. 1957, Level 2b).

The five-zone convention typically runs:

• Zone 1 — Recovery / very easy: below 60% HRmax. Below the aerobic threshold. Conversation is comfortable.
• Zone 2 — Aerobic / easy: 60 to 70% HRmax. The bulk of polarised training sits here. Long sentences are possible.
• Zone 3 — Tempo / moderate: 70 to 80% HRmax. Around the first ventilatory threshold. Short sentences.
• Zone 4 — Threshold / hard: 80 to 90% HRmax. Around the second ventilatory threshold / lactate threshold. Broken words.
• Zone 5 — VO2max / very hard: 90 to 100% HRmax. Above the second lactate threshold. Speech is not possible.

The 2008 [24] Binder et al. paper formalised the threshold-based model: LT1 is the work rate at which lactate first rises above resting values, and LT2 is the work rate at which lactate accumulates rapidly and clearance fails to keep up ([24] Binder et al. 2008, Level 5). In the five-zone convention, Zone 3 sits at LT1, Zone 4 sits at LT2, and Zone 5 sits above LT2. The 2006 [5] Seiler & Kjerland review is the operational anchor for the polarised distribution: in elite endurance athletes, about 75% of training sits below VT1 (Zone 1 + most of Zone 2), about 8% sits between VT1 and VT2 (Zone 3 and Zone 4), and about 17% sits above VT2 (Zone 5) ([5] Seiler & Kjerland 2006, Level 4). The 2014 [25] Stöggl & Sperlich paper added the experimental anchor: a randomised training-distribution study showed that the polarised group improved VO2max and 10K time more than threshold, HIIT, or high-volume groups ([25] Stöggl & Sperlich 2014, Level 2b).

Heart rate drift: why the same number means different things on different days

Heart rate is not a stable readout at constant work. The 2001 [4] Coyle & González-Alonso review quantified the drift: at constant work rate in the heat, heart rate rises 10 to 30 bpm over 60 minutes as stroke volume falls and core temperature climbs ([4] Coyle & González-Alonso 2001, Level 5). The 2015 [13] Périard, Racinais & Sawka review extended the picture: 10 to 14 days of heat acclimation lowers resting heart rate by 5 to 7 bpm and submaximal heart rate by 5 to 10 bpm, so the same pace that read 155 bpm in week one may read 145 bpm in week three of a heat block ([13] Périard et al. 2015, Level 5). The 2016 [14] Périard et al. paper added the mechanism: plasma volume expansion, lower skin blood-flow demand, and reduced sympathetic drive all lower submax HR in the heat-acclimated state ([14] Périard et al. 2016, Level 5).

The 2019 [28] Halson review consolidated the sleep and autonomic confounders: poor sleep raises next-day resting heart rate and disrupts autonomic balance, so a rower who slept five hours will read a higher heart rate at the same pace than one who slept eight ([28] Halson 2019, Level 5). The 2003 [10] Hautala et al. paper added the autonomic-baseline anchor: baseline vagal HRV predicted training response over eight weeks, with high baseline HRV athletes gaining more VO2max — the personalised-response layer that population-based zone formulas cannot see ([10] Hautala et al. 2003, Level 2b). The 2017 [11] Plews et al. paper sharpened it for rowing: world-champion rowers have highly individual HRV baselines, and reference values from one athlete do not transfer to another ([11] Plews et al. 2017, Level 2b). The 2014 [12] Plews et al. paper connected HRV to the training-distribution choice in elite rowers: HRV tracks polarised vs threshold training distribution over a season ([12] Plews et al. 2014, Level 2b).

The practical read for the indoor rower: a heart-rate reading is a snapshot, not a verdict. The same 165 bpm can mean "Zone 2 easy" on a cool morning after a good night of sleep and two weeks of heat acclimation, or "Zone 4 hard" on a hot afternoon after five hours of sleep. The 2001 [4] Coyle & González-Alonso paper is the most direct anchor: the same work rate produces different heart rates depending on hydration, temperature, and autonomic state, and the drift is the rule, not the exception ([4] Coyle & González-Alonso 2001, Level 5).

The talk test and Borg RPE: the rower's most reliable companions

The peer-reviewed anchors for how a hard sustainable pace feels are the Borg RPE scale and the talk test. The 2013 [21] Scherr et al. paper is the most direct anchor: in 2,560 men and women, Borg RPE correlated with blood lactate at r=0.83 and with heart rate at r=0.74; RPE at the lactate threshold was about 10.8, RPE at the individual anaerobic threshold was about 13.6, and RPE at the fixed 4 mmol/L threshold was about 14.1 ([21] Scherr et al. 2013, Level 2b). The 2014 [22] Reed & Pipe review established the talk test as a useful real-time companion: above the ventilatory threshold, comfortable speech is not possible (the "negative" stage); below it, the "last positive" and "equivocal" stages track the boundary closely ([22] Reed & Pipe 2014, Level 5).

The 2008 [24] Binder et al. paper operationalised the threshold linkage: VT1 sits at the first lactate turn point and roughly maps to RPE 11 to 12 ("fairly light" to "somewhat hard"); VT2 sits at the second lactate turn point and roughly maps to RPE 14 to 15 ("hard" to "very hard") ([24] Binder et al. 2008, Level 5). The 1964 [23] Wasserman & McIlroy paper anchored the physiological story: gas-exchange thresholds track the lactate story, and both track the talk test ([23] Wasserman & McIlroy 1964, Level 5). The 1993 [20] Steinacker review added the rowing-specific muscle-fibre anchor: elite rowers have 70 to 85% slow-twitch fibres and an aerobic-anaerobic threshold at 80 to 85% of maximal performance — the rowing-specific reason why Zone 4 feels familiar and Zone 5 does not ([20] Steinacker 1993, Level 5).

The practical read: the talk test and Borg RPE are the most reliable real-time companions to the heart-rate monitor because they integrate all the factors the monitor cannot see (temperature, hydration, sleep, stress, medication, sensor error). The 2006 [5] Seiler & Kjerland review is the practical anchor: the elite endurance distribution is about 80% easy (below VT1, conversation comfortable) and about 20% hard (above VT2, speech impossible), and the rower who can sense the boundary in real time does not need a single HR number to enforce the distribution ([5] Seiler & Kjerland 2006, Level 4).

Polarised training and what the zone numbers actually do

The 2006 [5] Seiler & Kjerland review established the foundational claim: elite endurance athletes train about 75% below VT1, about 7 to 8% between VT1 and VT2, and about 17 to 22% above VT2 ([5] Seiler & Kjerland 2006, Level 4). The 2009 [17] Guellich, Seiler & Emrich paper sharpened it for rowing: in 36 German junior world-championship finalists followed for 37 weeks, 95% of total rowing time was below 2 mmol/L lactate, 2% was at 2 to 4 mmol/L, and 3% was above 4 mmol/L ([17] Guellich et al. 2009, Level 2b). The 2004 [18] Fiskerstrand & Seiler paper added the longitudinal Norwegian data: low-intensity training rose from 30 to 50 hours per month over three decades, while race-pace and supra-maximal training fell from 23 to about 7 hours per month ([18] Fiskerstrand & Seiler 2004, Level 2b). The 2014 [25] Stöggl & Sperlich paper randomised the comparison: the polarised group improved VO2max and 10K time more than threshold, HIIT, or high-volume groups ([25] Stöggl & Sperlich 2014, Level 2b).

The 2002 [15] Ingham et al. paper put the rowing-specific performance anchor on the zone framework: in world-championship finalists, a regression model with power at VO2max, VO2 at lactate threshold, power at 4 mmol/L lactate, and peak power explained 98% of 2K speed variance ([15] Ingham et al. 2002, Level 2b). The 2004 [16] Bourdin et al. paper added peak power: in 54 male rowers, peak power output correlated with 2K time at r=0.92 ([16] Bourdin et al. 2004, Level 2b). The 2017 [19] Bourdin et al. paper confirmed it in 70 female rowers: Ppeak r=0.89, PLa4 r=0.87, VO2max r=0.83 ([19] Bourdin et al. 2017, Level 2b). The 2018 [27] Volianitis, Secher & Quistorff paper added the rowing-specific lactate anchor: after 5 minutes of rowing, arterial lactate reached 17.5 ± 1.6 mM, and intramuscular pH recovery was 3.5-fold slower than after handgrip — the rowing-specific reason why Zone 5 efforts are not repeatable ([27] Volianitis et al. 2018, Level 2b).

The practical read for the indoor rower: heart-rate zones are useful for naming the distribution (about 80% easy, about 20% hard) and for catching a session that has drifted out of band. They are not useful for policing the numbers — the same HR reading means different things on different days, and the rower who treats the monitor as a verdict on whether they are "in zone" is over-fitting.

Heart rate variability: the separate signal that matters over weeks

Heart rate variability (HRV) is a separate signal from heart rate: it is the beat-to-beat variation in the time between heartbeats, and it tracks autonomic nervous-system balance. The 1996 [9] Task Force consensus document defined the standard measures: SDNN (overall variability), RMSSD (vagal-tone marker), pNN50 (vagal-tone marker), and the frequency-domain measures HF (high-frequency, vagally mediated) and LF (low-frequency, mixed sympathetic and vagal) ([9] Task Force ESC 1996, Level 5). The 2013 [8] Buchheit review established the operational convention: in elite endurance athletes, RMSSD trends over weeks track autonomic readiness, and a 7-day rolling average is more informative than any single daily reading ([8] Buchheit 2013, Level 5).

The 2016 [34] Bellenger et al. meta-analysis consolidated the evidence: across 39 studies, RMSSD / Ln rMSSD trends over weeks track autonomic readiness and overreaching risk ([34] Bellenger et al. 2016, Level 2b). The 2013 [6] Meeusen et al. consensus statement added the diagnostic context: overreaching is a multi-marker diagnosis in which HRV is one input alongside hormonal markers, mood-state questionnaires, and performance tests — and a single low HRV reading is not a verdict on overtraining ([6] Meeusen et al. 2013, Level 5). The 2017 [11] Plews et al. paper sharpened it for rowing: world-champion rowers have highly individual HRV baselines, so the rower's own 60 to 90 day baseline is the only valid reference range ([11] Plews et al. 2017, Level 2b). The 2003 [10] Hautala et al. paper added the response-prediction anchor: baseline vagal HRV predicted training response over eight weeks, so a low baseline HRV does not mean a rower cannot train — it means the response will be different, not absent ([10] Hautala et al. 2003, Level 2b).

The 2010 [26] Lamberts et al. paper added the heart-rate-recovery (HRR) anchor: the drop in heart rate from peak exercise to one minute post-exercise tracks training status, and a small HRR drop (less than 12 bpm at 1 minute) is a marker of autonomic imbalance in clinical populations ([26] Lamberts et al. 2010, Level 2b; [7] Cole et al. 1999, Level 2b). The 2019 [28] Halson review consolidated the sleep confounders: poor sleep elevates next-day resting HR and depresses HRV, so the rower who slept five hours will see a different HRV trend than one who slept eight — and the trend over weeks is the signal, not the daily reading ([28] Halson 2019, Level 5).

The practical read for the indoor rower: HRV is the readiness signal, and the rower's own 60 to 90 day rolling baseline is the only valid reference. A daily reading is a snapshot; a 7-day trend is a signal; a 4-week trend is a verdict. The AI coach that uses a single HRV reading to alter a training plan is over-fitting; the coach that watches the rolling baseline alongside RPE, sleep, and training load is using HRV correctly.

Measurement accuracy: chest strap, wrist, and the PM5

Heart-rate monitors have three layers of error: the sensor, the data link, and the PM5's processing. The 2001 [4] Coyle & González-Alonso review is the methodological anchor: at high work rates, optical wrist sensors overestimate heart rate during arm-dominant work, and chest-strap electrodes remain the de-facto reference for rowing-specific HR validation ([4] Coyle & González-Alonso 2001, Level 5). The 2024 [29] Concept2 PM5 documentation confirms the data-link layer: the PM5 supports ANT+ and Bluetooth Smart wireless heart-rate monitoring with compatible chest straps, and the manufacturer-recommended pairing is a Polar H10 or equivalent chest strap ([29] Concept2 PM5, Level 5). The 2024 [30] Polar H10 product page confirms the sensor layer: the H10 is the manufacturer-referenced chest-strap sensor for rowing validation studies and has a published ±1 bpm accuracy against ECG at rest and moderate exercise ([30] Polar H10, Level 5).

The practical read for the indoor rower: a Polar H10 or Garmin HRM-Pro chest strap into the PM5 via ANT+ or Bluetooth Smart is the most reliable signal; wrist optical sensors are useful for cross-checking but can drift at high stroke rates; and the rower who trusts the monitor without checking the strap placement and battery is reading sensor noise as physiology. The 2001 [4] Coyle & González-Alonso review is the most direct anchor: cardiovascular drift is real physiology, but the sensor error is on top of it, and the rower who cannot distinguish the two is using the monitor as a verdict ([4] Coyle & González-Alonso 2001, Level 5).

What the AI coach does with heart-rate zones

For an AI coach that reads the rower's logbook and writes the rower's session, heart rate is one signal among several. The international federation frame — World Rowing's medical and physiological resources ([31] World Rowing, Level 5) — and the national-federation coaching frame — British Rowing's knowledge base ([32] British Rowing, Level 5) and USRowing's education resources ([33] USRowing, Level 5) — both treat heart-rate monitoring as one input among several, never as a verdict on whether a rower is working hard enough. The coach's rule is:

• A rower who has done a 2K or 30-minute test in the past six weeks — the coach uses the average heart rate at the test pace as the reference HR, and watches for cardiovascular drift over the session. The 2002 [15] Ingham et al. paper is the methodological anchor: power at VO2max, VO2 at lactate threshold, power at 4 mmol/L lactate, and peak power together explained 98% of 2K variance, and the heart rate at the test pace is what the coach reads ([15] Ingham et al. 2002, Level 2b; [16] Bourdin et al. 2004, Level 2b; [19] Bourdin et al. 2017, Level 2b).
• A rower whose recent test HR has drifted upward at the same pace — the coach widens the reference-pace band and checks for sleep loss, dehydration, illness, or accumulated training load. The 2001 [4] Coyle & González-Alonso review is the physiological anchor: at constant work rate in the heat, heart rate rises 10 to 30 bpm as stroke volume falls and core temperature climbs ([4] Coyle & González-Alonso 2001, Level 5; [13] Périard et al. 2015, Level 5; [28] Halson 2019, Level 5).
• A rower on a beta-blocker or rate-control medication — the coach abandons heart-rate-based pacing entirely and leans on RPE, the talk test, and stroke-rate cues. The 2013 [6] Meeusen et al. consensus statement is the methodological anchor: HR-based pacing is unreliable when the autonomic signal is pharmacologically altered ([6] Meeusen et al. 2013, Level 5; [8] Buchheit 2013, Level 5).
• A rower with no recent test — the coach widens the reference-pace band, leans on RPE and the talk test, and treats the first 2K as a familiarisation piece rather than a calibration of physiology. The 2002 [15] Ingham et al. paper is the methodological anchor: a single test's HR is a snapshot, and the moving average of two to three familiarised tests is the reference ([15] Ingham et al. 2002, Level 2b; [10] Hautala et al. 2003, Level 2b; [11] Plews et al. 2017, Level 2b).
• A rower whose HRV trend has fallen 10 to 15% below the 60-day rolling baseline for more than a week — the coach widens the recovery load, reduces the planned intensity, and waits for HRV to return before resuming the planned progression. The 2013 [6] Meeusen et al. consensus is the diagnostic anchor: HRV is one input alongside hormonal markers, mood, and performance tests, and a 7 to 14 day downward trend is a signal ([6] Meeusen et al. 2013, Level 5; [34] Bellenger et al. 2016, Level 2b; [8] Buchheit 2013, Level 5).

The coach that anchors the rower's reference HR on a single test, treats a single HRV reading as a verdict, or polices the session against a fixed HR target is over-fitting. The coach that anchors on the moving average of recent tests, watches the rolling HRV baseline, and treats HR as one signal alongside RPE, the talk test, and stroke rate is using the monitor correctly.

Limitations and open questions

The rowing-specific HR-zone literature is small. Most of the foundational papers are in cycling, running, or general exercise physiology, with rowing-specific work concentrated on the 2002 [15] Ingham et al. sample of world-championship finalists, the 2009 [17] Guellich et al. sample of 36 German juniors, and the 2004 [16] Bourdin et al. sample of 54 male rowers. The 2014 [12] Plews et al. paper on HRV and training-intensity distribution in elite rowers is the most direct rowing-specific anchor for HRV as a monitoring signal ([12] Plews et al. 2014, Level 2b). The reader should weight the rowing-specific evidence more heavily than the cross-sport evidence when the two diverge.

Population-based HRmax formulas have wide individual scatter. The 2001 [1] Tanaka, Monahan & Seals formula has a standard deviation of about ±10 bpm at any given age, so the same formula can predict 195 bpm or 175 bpm in two 40-year-olds with the same training history ([1] Tanaka et al. 2001, Level 2b). The 2010 [2] Gulati et al. paper added the female-specific formula for the cases where the Tanaka formula mis-fits ([2] Gulati et al. 2010, Level 2b). The 1957 [3] Karvonen formula is more personalised because it uses the rower's resting heart rate, but the rower who uses Karvonen on a morning when resting HR is 5 bpm above baseline will under-pace the session by 2 to 3% ([3] Karvonen et al. 1957, Level 2b).

Sensor error is layered on top of physiology. The 2001 [4] Coyle & González-Alonso review is the methodological anchor: at high work rates, optical wrist sensors overestimate heart rate during arm-dominant work, and chest-strap electrodes remain the de-facto reference for rowing-specific HR validation ([4] Coyle & González-Alonso 2001, Level 5). The 2024 [29] Concept2 PM5 documentation confirms the data-link layer: the PM5 supports ANT+ and Bluetooth Smart wireless heart-rate monitoring with compatible chest straps ([29] Concept2 PM5, Level 5). The 2024 [30] Polar H10 product page confirms the sensor layer: the H10 is the manufacturer-referenced chest-strap sensor for rowing validation studies ([30] Polar H10, Level 5). The rower who trusts the monitor without checking the strap placement and battery is reading sensor noise as physiology.

Cardiovascular drift means the same pace produces different HR over a session. The 2001 [4] Coyle & González-Alonso review quantified the drift: at constant work rate in the heat, heart rate rises 10 to 30 bpm over 60 minutes as stroke volume falls and core temperature climbs ([4] Coyle & González-Alonso 2001, Level 5). The 2015 [13] Périard et al. review extended the picture: 10 to 14 days of heat acclimation lowers resting HR by 5 to 7 bpm and submax HR by 5 to 10 bpm, so the same pace that read 155 bpm in week one may read 145 bpm in week three of a heat block ([13] Périard et al. 2015, Level 5). The rower who polices the session against a fixed HR target is fighting physiology.

HRV-based overreaching detection is a trend, not a threshold. The 2013 [6] Meeusen et al. consensus is the diagnostic anchor: HRV is one input alongside hormonal markers, mood, and performance tests, and a 7 to 14 day downward trend is a signal — not a single low reading ([6] Meeusen et al. 2013, Level 5). The 2017 [11] Plews et al. paper sharpened it for rowing: world-champion rowers have highly individual HRV baselines, so the rower's own 60 to 90 day baseline is the only valid reference range ([11] Plews et al. 2017, Level 2b). The 2013 [8] Buchheit review operationalised it: a 7-day rolling average is more informative than any single daily reading, and a 4-week trend is a verdict ([8] Buchheit 2013, Level 5).

What to do with this article

Read the principle: heart rate is a real-time read on the body's acute cardiovascular response, modulated by temperature, hydration, sleep, stress, medication, and sensor error. The modern convention recognises two lactate thresholds, and the five-zone model names the training distribution in terms of those thresholds. Read the evidence: the 2001 [1] Tanaka, Monahan & Seals paper (Level 2b) anchors the HRmax formula; the 1957 [3] Karvonen, Kentala & Mustala paper (Level 2b) anchors the HRR method; the 2001 [4] Coyle & González-Alonso review (Level 5) anchors cardiovascular drift; the 2006 [5] Seiler & Kjerland review (Level 4) and the 2014 [25] Stöggl & Sperlich paper (Level 2b) anchor the polarised distribution; the 2013 [21] Scherr et al. paper (Level 2b) and the 2014 [22] Reed & Pipe review (Level 5) anchor RPE and the talk test; the 1996 [9] Task Force document (Level 5) and the 2013 [8] Buchheit review (Level 5) anchor HRV. Read the practical read: the same heart-rate number means different things on different days; the talk test and Borg RPE are the most reliable real-time companions; the polarised distribution is the modern consensus position; the rower's own 60 to 90 day HRV baseline is the only valid readiness reference.

When you want to use the monitor, the practical recipe is: pair the PM5 with a Polar H10 chest strap over ANT+ or Bluetooth Smart, not a wrist optical sensor; warm up for 5 to 10 minutes and read the drift over the session rather than the absolute HR at any minute; use the talk test as the real-time intensity check — long sentences are Zone 2, short sentences are Zone 3, broken words are Zone 4 or 5; track your own 60 to 90 day RMSSD or morning-resting-HR trend as the readiness signal, not the daily reading; and treat the monitor as a useful guide, not a verdict. The AI coach that anchors on the moving average of recent tests, watches the rolling HRV baseline, and treats HR as one signal alongside RPE, the talk test, and stroke rate is reading the signal correctly.