A wearable's sampling rate is the least discussed number on its specification sheet and the one that determines what the device is capable of detecting.
Sampling is snapshots, not continuous recording
A sensor measures at intervals, and everything between two measurements is unobserved. The device reconstructs motion by connecting the points it captured.
If the movement being measured completes faster than the interval between samples, the device may miss it entirely or record a distorted version.
The limit is physical. No amount of processing recovers detail that was never captured, which is why sampling rate cannot be fixed in software.
Different questions need different rates
Position over a match changes relatively slowly, so a modest rate describes running paths and total distance perfectly well.
Acceleration during a change of direction happens far faster, so detecting it requires sampling many times more often than positional tracking needs.
Impacts are faster again, lasting a small fraction of a second, which is why devices intended to capture collisions run their accelerometers at very high rates.
Undersampling produces confident wrong answers
When a signal changes faster than the sampling rate, the reconstruction can show a slower pattern that was never present. The output looks plausible and is fabricated by the sampling itself.
In practice this appears as smooth acceleration traces from a device that could not have resolved the sharp changes actually occurring.
An analyst reading that trace sees a player accelerating gently and concludes something about their movement that the instrument was incapable of measuring.
Higher rates cost battery and storage
Sampling more often means more power drawn and more data written, and a device worn for a full session has a fixed energy budget.
Manufacturers therefore run different sensors at different rates within the same unit, sampling position modestly and acceleration aggressively.
Some devices raise the rate only when motion crosses a threshold, capturing detail where it matters while sleeping through quiet periods.
What buyers should actually ask
The useful question is which rate applies to which sensor, since a single headline figure usually describes the fastest one rather than the one producing a given metric.
It is equally important to ask what the device transmits, because many report summaries computed on board and discard the raw samples entirely.
Without the raw signal, a club cannot recompute a metric when its definition changes, which turns a measurement into a vendor's opinion about a measurement.