Your Smart Watch Says You’re Tired — Should You Actually Change Your Workout?

You wake up feeling fairly normal. You slept reasonably well, you’ve got a training session planned and physically you don’t feel too bad. Then you look at your watch. Your sleep score is poor, your HRV is down, your Body Battery hasn’t recharged and your training readiness is telling you that today probably isn’t the day to push yourself. So what do you do?

Wearable technology has become a huge part of modern fitness. Garmin, Apple Watch, WHOOP, Oura and similar devices can now give us information on heart rate, sleep, heart-rate variability, training load, stress, recovery and even how “ready” we are to train. The amount of information available is impressive, but the more important question is whether we are actually using it properly. A low score shouldn’t automatically mean cancelling your workout. Equally, repeatedly ignoring signs that your body is struggling to recover isn’t particularly sensible either. The useful approach sits somewhere in the middle.

What Is Your Watch Actually Measuring?

Most wearable devices are not directly measuring “recovery”. Instead, they collect several pieces of information and use algorithms to estimate your current state. Garmin’s Training Readiness feature, for example, considers recent sleep, recovery time from previous exercise, heart-rate variability, acute training load, sleep history and stress history. Its Training Status feature also considers changes in VO₂ max, recent training load and HRV over time. Garmin itself describes these measurements as estimates designed to help users understand their training rather than definitive measurements of what is happening inside the body. (Garmin, 2026).

That distinction matters. Your watch isn’t looking inside your muscles and determining that your legs are 73% recovered. It is taking information that it can measure and using that information to make an educated prediction. That prediction can be extremely useful, but it still needs context.

HRV: Focus on the Trend, Not One Bad Reading

One of the most talked-about recovery measurements is heart-rate variability, or HRV. Despite the name, a healthy heart doesn’t beat perfectly regularly. There are tiny variations in the amount of time between individual heartbeats. These variations are influenced by the autonomic nervous system and can provide useful information about how the body is responding to training and other forms of stress. Hard training, poor sleep, illness, alcohol, psychological stress, travel and inadequate nutrition can all influence it.

This is why comparing your HRV with somebody else’s isn’t particularly useful. A “good” HRV number for one person could be completely normal or abnormal for somebody else. The trend matters much more. Garmin, for example, assesses HRV status by comparing a rolling seven-day average with an individual’s established baseline rather than treating one night in isolation. Research into the use of HRV for athlete monitoring has similarly highlighted the importance of averaging measurements and looking at patterns rather than over-interpreting individual daily readings. (Plews et al., 2013).

So if your HRV suddenly drops one morning but you feel good and everything else looks normal, it doesn’t necessarily mean you need to abandon your training session. If HRV has been suppressed for several days, resting heart rate is elevated, sleep has deteriorated and you’re also feeling unusually tired, that becomes much more meaningful. This is where wearable data starts becoming useful: not because one metric told you what to do, but because several pieces of information are pointing in the same direction.

Resting Heart Rate Can Be a Useful Warning Sign

Resting heart rate is another simple but useful measure. It tends to be relatively stable when measured under similar conditions, so if yours suddenly sits noticeably higher for several mornings, something may have changed. That doesn’t automatically mean you’re overtraining. It could be increased training load, poor sleep, dehydration, alcohol, psychological stress or the early stages of illness. Again, context matters.

Wearable heart-rate measurements are also one of the better-performing measurements these devices provide. A large systematic review and meta-analysis found that wrist-worn devices showed relatively small average errors for heart rate across many everyday and aerobic activities, although accuracy became poorer during activities such as resistance training and cycling. (Zhang et al., 2020). This means wrist-based heart rate can be very useful for tracking general trends, but if accurate heart-rate data is important during training, a chest strap will generally provide a more reliable measurement.

How Much Should You Trust Your Sleep Score?

Sleep is another area where wearables can be useful, but not perfect. Your watch doesn’t actually know that you slept for exactly seven hours and 14 minutes, with precisely 58 minutes of deep sleep and 91 minutes of REM sleep. It is estimating those numbers from signals such as movement and heart rate.

A 2026 systematic review and meta-analysis comparing wearable sleep trackers with polysomnography — the laboratory reference standard for assessing sleep — found that wearables could provide useful estimates of overall sleep but tended to overestimate total sleep time and sleep efficiency while underestimating periods of wakefulness during the night. Accuracy also varied considerably between devices. (Agostinho et al., 2026). Another meta-analysis involving devices including Garmin, Apple Watch, Fitbit and WHOOP also found meaningful differences between consumer wearables and polysomnography for total sleep time, sleep efficiency, sleep latency and waking after sleep onset. (Lee et al., 2024).

That doesn’t make your sleep score useless. Far from it. If your watch consistently tells you that you’re getting six hours of sleep when you need closer to eight, you probably don’t need a sleep laboratory to tell you that more sleep would be beneficial. The danger comes when we become obsessed with the precision of individual sleep stages. Going to bed feeling like you slept well and then deciding you must be exhausted because your watch gave you a sleep score of 62 is probably using the technology backwards.

Don’t Let the Watch Tell You How You Feel

There is an interesting problem that comes with having access to constant health information. Sometimes we feel fine until our watch tells us that we aren’t. You wake up feeling ready to train, see a poor readiness score and suddenly start wondering whether you’re tired. For some people, sleep tracking can even increase anxiety around sleep. The term “orthosomnia” has been used to describe people becoming excessively focused on achieving perfect sleep data from wearable technology.

The purpose of measuring something should be to make better decisions. If the measurement itself starts creating additional stress, its usefulness becomes questionable. Before looking at your recovery score in the morning, it can therefore be useful to ask yourself some very basic questions. How do I actually feel? Do my legs feel unusually heavy? Am I motivated to train? Do I feel ill? Am I unusually sore? Did I sleep badly? Once you’ve formed that subjective picture, your wearable can provide another layer of information.

Should You Actually Change Today’s Workout?

This becomes particularly important when deciding whether to change a workout. Imagine you have a hard strength and conditioning session planned. Your Garmin says your readiness is low, but you feel good, your warm-up feels normal and nothing unusual has happened during the previous few days. There may be absolutely no reason to change the session.

Now consider the opposite scenario. Your HRV has been below your usual range for three days. Resting heart rate is up. You’ve slept poorly. Your previous few sessions have felt harder than normal and during the warm-up your legs feel flat. That is probably a good day to adjust something.

But adjusting training doesn’t necessarily mean doing nothing. You might reduce the weight slightly, remove a few sets, avoid maximal conditioning, complete some lower-intensity aerobic work or simply finish the session earlier than planned. Good programming isn’t about choosing between training at 100% and staying on the sofa. There is a huge amount of space between those two options.

Your Warm-Up Is a Readiness Test Too

One of the best ways of assessing whether you are ready to train is still remarkably simple: warm up. If your normal movements feel good, coordination is normal and your usual training weights move as expected, that’s valuable information. If everything feels unusually heavy, movement quality is poor and your heart rate seems disproportionately high for the workload, that is valuable information too.

Experienced coaches have effectively been using this approach for decades. Before watches provided readiness scores, coaches watched how athletes moved, spoke to them, looked at recent training loads and adjusted sessions accordingly. Technology can improve that process. It shouldn’t replace it.

Look for Patterns, Not Perfect Scores

The biggest mistake people make with recovery data is reacting too aggressively to individual measurements. Human physiology is noisy. HRV changes. Resting heart rate changes. Sleep varies. Training performance fluctuates. Nobody performs exactly the same every day.

If you reduce your training every time one metric isn’t perfect, you may actually remove many of the challenging sessions required to improve. Training itself is a stress. We don’t want to eliminate stress completely. We want to apply enough stress to stimulate adaptation and then provide enough recovery for the body to adapt.

Garmin makes a similar point with its Body Battery feature. The aim isn’t necessarily to finish every day with a completely full “battery”. Exercise and demanding activity are expected to drain energy. Repeatedly avoiding challenging activity simply because your energy score isn’t perfect would defeat part of the purpose of training in the first place. (Garmin, 2026).

This is why patterns are more useful than snapshots. One poor night isn’t particularly concerning. A week of declining sleep, suppressed HRV, elevated resting heart rate, falling performance and increasing fatigue deserves attention.

Be More Cautious With Calories Burned

Another number worth treating with more caution is calorie expenditure. People regularly finish a workout and see that their watch claims they burned 700 calories. That number can look incredibly precise. It isn’t.

Wearables estimate energy expenditure using information such as heart rate, movement, age, body mass and proprietary algorithms. The problem is that estimating how much energy somebody has actually expended is considerably more difficult than simply measuring heart rate. A systematic review of wrist-worn activity trackers found that none of the included consumer brands demonstrated consistently accurate estimates of energy expenditure. (Fuller et al., 2022).

A much larger 2024 umbrella review examining 24 systematic reviews and 249 validation studies came to a similar conclusion: wearables can provide useful health and activity information, but accuracy varies substantially depending on what is being measured. Heart rate tends to perform reasonably well, while estimates of variables including energy expenditure, physical activity intensity and VO₂ max can contain considerably larger errors. (Doherty et al., 2024).

So if your watch says you burned 820 calories during training, don’t automatically assume that means you’ve earned an additional 820 calories of food. Think of calorie expenditure as an estimate rather than an accounting system.

What Data Actually Matters Most?

Perhaps the most useful thing about wearable technology isn’t any individual number. It’s the ability to identify changes in your normal behaviour and physiology over time. If we were looking at a member’s wearable data at SPC, we’d be much more interested in seeing that their resting heart rate had gradually increased over several days than whether it was 51 or 52 beats per minute this morning.

We’d be more interested in HRV repeatedly moving outside their personal baseline than comparing their HRV with somebody else’s. We’d be interested in whether sleep duration had gradually fallen because work had become busier, whether training load had suddenly increased and whether sessions were starting to feel harder than normal.

Most importantly, we’d combine those numbers with what the person was actually telling us. “I feel great” is information. “My legs have felt awful all week” is information. “This weight normally feels comfortable but today it feels incredibly heavy” is information. The watch simply adds another piece to the puzzle.

Your Watch Doesn’t Know Your Entire Life

That is ultimately the limitation of even the best wearable. It doesn’t know everything. It might recognise that your heart rate and HRV look different, but it doesn’t necessarily know that you spent yesterday travelling, barely ate lunch, had a stressful day at work, played sport in the evening and then went to bed two hours later than usual.

Human performance is influenced by training, nutrition, hydration, sleep, psychological stress, illness, work, relationships and dozens of other factors. An algorithm can detect some of the physiological consequences, but it can’t fully understand the context.

The best approach isn’t to ignore wearable technology and it isn’t to blindly obey it either. Use it, but interpret it.

The SPC Takeaway

Fitness watches are becoming increasingly sophisticated and the information they provide can genuinely help us understand our training and recovery. HRV trends, resting heart rate, sleep duration and training load can all provide useful information, but there is a big difference between using data and being controlled by it.

A poor recovery score isn’t automatically a reason to skip training. A perfect recovery score doesn’t guarantee that you’re going to have a great session either. Look at trends, consider how you actually feel, pay attention to your warm-up and recent performance and use your wearable data as one part of the decision-making process.

Your watch can provide information. It shouldn’t be your coach.

References

Garmin Ltd. (2026). Garmin Training Status and How to Use It and Body Battery Energy Monitoring. Garmin Human Performance Lab.

Agostinho, M., Borges, M., Pereira, T., Borges, D.F. & Soares, J.I. (2026). Are Wearable Sleep-Tracking Devices Reliable Alternatives to Polysomnography? A Systematic Review and Meta-Analysis. Behavioral Sleep Medicine.

Doherty, C. et al. (2024). Keeping Pace with Wearables: A Living Umbrella Review of Systematic Reviews Evaluating the Accuracy of Consumer Wearable Technologies in Health Measurement. Sports Medicine.

Fuller, D. et al. (2022). Accuracy and Acceptability of Wrist-Wearable Activity-Tracking Devices: Systematic Review of the Literature. Journal of Medical Internet Research.

Lee, Y.J. et al. (2024). Performance of Consumer Wrist-Worn Sleep Tracking Devices Compared to Polysomnography: A Meta-Analysis.

Plews, D.J., Laursen, P.B., Stanley, J., Kilding, A.E. & Buchheit, M. (2013). Training Adaptation and Heart Rate Variability in Elite Endurance Athletes: Opening the Door to Effective Monitoring. Sports Medicine.

Zhang, Y. et al. (2020). Validity of Wrist-Worn Photoplethysmography Devices to Measure Heart Rate: A Systematic Review and Meta-Analysis. Journal of Sports Sciences.