The cost of longer battery life depends on what a mode saves
Before a marathon, the watch promises longer battery life through one switch. The choice seems convenient, but the eventual file may be less detailed than usual training recordings. The question is not whether every power-saving mode is bad. Know which functions and data the specific setting modifies. You can then choose a sensible compromise for your event and interpret the recording correctly later.
Apple’s guide, for example, treats Low Power Mode and Fewer GPS and Heart Rate Readings separately. On supported models and systems, the latter saves power through less frequent GPS and heart-rate readings during outdoor walks, runs or hikes, and requires Low Power Mode. The names are therefore not one identical switch. Another manufacturer may organise this differently. Source: Apple Watch Low Power Mode.
Begin with a list of needed features. Do you want uninterrupted recording, a detailed route, frequent heart-rate data, current pace, alerts or navigation? They do not always matter equally. A complete recording reaching the finish may be more valuable than a detailed system running flat, but missing running feedback can also be a significant difference. Tie the decision to your event rather than the largest displayed hour count.
Radio measurement, sensor readings and file writing are separate layers
Sparser file records do not prove equally sparse internal measurement on every device. Sensing and storage rates may differ. Likewise, selecting multiple satellite systems or frequency bands is not simply identical to how many points enter the file. The manual defines the setting. You can inspect actually stored data from a recording, but cannot always reconstruct every internal process.
Garmin’s FIT guide distinguishes change-responsive, irregular Record-message intervals under Smart Recording from every-second storage. Here the aim is reducing file size; do not automatically equate this with every power-saving sensor mode. Similarly sparse-looking point series can arise from different settings. Source: Garmin FIT, Smart Recording.
Give selected power mode and observed file density separate columns on your checklist. Add documented changes to measurement frequency if available. If you know only saved records, state the conclusion in those terms. Precise wording avoids inferring sensor failure or an undocumented internal measurement cycle from a storage method.
Enlarge imageA short change can fall between samples
In an invented timing example, heart rate is 150 at both 0 and 60 seconds, with a temporary change during ten seconds between them. With only the endpoints available, the intervening signal shape cannot reliably be recovered. A straight line connecting them stays smooth, but is the analysis app’s drawing rather than measurement of the intervening state. This is not a claim about a specific watch’s frequency or physiological event.
The same applies to routes. You leave a junction, detour briefly to a refreshment point and return. If retained points before and after the detour are at the same junction, coordinates alone do not show the intermediate path. Other data may let the system estimate distance, but surviving coordinates do not contain every turn. A smooth map is therefore not the complete route history.
Drawing more points between endpoints cannot recover lost detail later. Interpolation may fill a gap under an assumption, but new points do not become measurements. Recording detail is fundamental for investigating brief accelerations, heart-rate responses or tight turns. A more impressive graph cannot reliably reconstruct information forgone during recording.
Alerts and split pace may be separate losses in use
For Apple’s cited fewer-GPS-and-heart-rate-readings setting, the manufacturer lists further consequences: workout alerts switch off, splits and segments are not recorded and, for example, current pace is unavailable. Do not extend these to every other product. For that feature, however, the change is more than sparse measurement: the runner’s immediate workout experience differs too. Source: limitations of fewer readings.
If the marathon plan relies on important feedback, test it beforehand in the same mode. The timer starting is insufficient. Check desired warnings, lap indications, pace screen and navigation separately where used. Discover a missing function on a calm test loop rather than after the start. A working recording and familiar feedback are separate acceptance conditions.
Describe the result specifically. “The workout saved, but the usual pace alert did not work in this mode” is interpretable. “My watch is good for marathons” leaves too much open. The same compromise may suit you but not someone else. Weight features actually used, and avoid rediscovering a documented limitation at every race.
File timestamps describe retained density
The FIT encoding guide associates instantaneous position and sensor data with Record messages, stored at regular or irregular intervals. Differences between actual timestamps can therefore show the exported series’ density. They do not establish every internal measurement frequency, but reveal conditions for later analysis. Source: writing Garmin FIT Activity files.
In an example, one series has timestamps of 0, 1, 2 and 3 seconds; another 0, 15, 30 and 45. Each has four records but covers very different duration and detail. File row count alone is insufficient for comparison. With recordings of differing lengths, inspect intervals and covered time too. “Less data” becomes precise only with these.
Distinguish a pause from a long sample interval. FIT uses Timer events for recording-timer state, allowing stopped workouts and sparse points to be interpreted differently. A twenty-second gap does not automatically mean twenty seconds stationary. Treating every neighbouring row as equal time can yield a wrong mean under irregular recording. Use actual durations rather than row count alone. Source: Timer events and recording intervals.
Inspect the method used to summarise a sparse series too
Consider a simple value–time example: 100 at 0 seconds, 200 at 10 and 100 again at 11. The simple mean is 133.3, ignoring how much time each represents. Assuming the previous value holds gives approximately 109.1 over 0–11 seconds. Another interpolation may yield another result. These example values are not for heart-rate interpretation; they illustrate averaging irregular samples.
The lesson is not to apply one simple filling rule to every export. Know the processing purpose and data meaning. A device’s own summary may also have internal information absent from sparse exported records. If your recalculated mean differs, clarify time basis, valid samples and method first. The discrepancy is not immediately a device fault.
Take even greater care comparing short intervals retrospectively. Without a point near a repetition boundary, software may split the signal under different assumptions. Total long-run time may remain useful while analysis of individual accelerations becomes uncertain. Do not decide a recording’s value with one yes/no label. Name which metric has sufficient retained information and which is too sparsely supported.
Enlarge imageTest the entire setup before the start
Choose a familiar, safe short route with several turns and the race’s main features. Watch, sensors, app and mode should match what you plan to use. Note setting and charge before starting. Observe familiar feedback during running and inspect the complete saved activity afterwards. This need not be a separate hard workout; the technical aim is checking a known use situation.
The Garmin Forerunner 955 manual, for example, describes UltraTrac recording track points and sensor data less often, with reduced detail in return for battery life. Do not automatically equate this specific mode with differently named features. The shared goal of saving power can follow different technical paths. Your manual and actual trial together establish the compromise chosen. Source: Forerunner 955 UltraTrac.
Retain consumption as a separate outcome. Note starting and ending charge, recording duration and important active features. Do not estimate marathon battery life to a tenth of an hour from a short trial: display resolution, use and battery condition matter. Planning needs reserve rather than a tight limit based on one favourable occasion. For continuous recording, choose a previously tested solution.
Name the compromise in later analysis too
Mark power-saving recording in the log, particularly when comparing usual workouts. A sparser heart-rate or route series from another mode may not support the same short-window analysis. A missing acceleration peak does not automatically mean no acceleration; a smooth line does not establish more even running technique. Check which data the mode retained first.
Reconsider priorities before the next event too. Detailed mode may provide ample marathon battery life, while a much longer day requires another setting. You need not choose one profile forever. Change modes deliberately and document it, with important features tested beforehand. Power saving then appears as a known decision rather than accidental information loss.
Longer battery life and retained information detail together determine a useful setting. Precise manufacturer descriptions, a short personal trial and checking the original file are three separate references. If aligned, you can rely on chosen functions during the marathon and later know what the recording can explain. Sparse data do not undo the run, but limit which brief changes can reliably be analysed afterwards.
Sources
- Apple Watch: Low Power Mode
- Garmin FIT: decoding Activity files
- Garmin FIT: writing Activity files
- Forerunner 955: satellite settings
How Futószív articles are prepared: our editorial guidelines.


