What happens if I wear two devices?

One wins per metric. Base never adds two sources together, because bridged providers write into Apple Health and their samples are already inside Apple's totals — summing them would double-count. The Health Sources screen names the winning provider for each of today's numbers.

Why does Base prefer Apple Health for steps but my ring for HRV?

Because the priority order differs per metric on purpose. For totals like steps and energy, Apple Health is the union of every source including bridged ones, so it is the complete picture. For specialist metrics like heart-rate variability and sleep staging, the device that actually measured it is preferred, because Apple is only relaying it.

Can I choose which device wins?

The priority order is fixed per metric rather than user-configurable. What you can do is control what reaches the health store — turning off a source's write permission in the Health app removes it from consideration entirely, which is the effective way to change the winner.

Why does the screen list apps I have not set up?

It lists anything writing into your health store that Base has no explicit row for, so a Polar or Withings user learns their data is already arriving. It is informational rather than a setup step — those sources are already contributing through the store.

A metric shows no value at all. What does that mean?

No provider held a value for it that day. Base leaves it empty rather than falling back to zero, because zero is a claim and absent is not. If you expected data, the per-type diagnostics on the same screen distinguish an empty day from a permission that was never granted.

Which Device Base Believes, and Why

The rule that matters: never sum

The single most dangerous thing a multi-device app can do is add.

On iOS, Garmin Connect, Fitbit, Whoop and Oura all write into Apple Health. A bridged provider's samples are therefore already inside Apple's own totals.

Apple Health steps   11,000
Garmin steps         11,000   ← the same 11,000, relayed

summed:              22,000   ← wrong by a full workout's worth of energy

That error does not stay contained. Steps feed a walking-energy estimate, which feeds expenditure, which feeds your calorie target and your measured maintenance. A doubled step count quietly inflates everything downstream.

So every metric picks one winner: the first provider in that metric's priority order holding a non-null value for the day.

Why the order differs per metric

Metric type	Prefers	Reasoning
Totals — steps, active energy, basal energy	Apple Health / Health Connect	It is the union of every source including bridged ones, so it is the complete picture
Specialist — HRV, sleep staging, resting HR	The measuring device	The chest strap or ring measured it; the store is only relaying it

That split is the whole design. For a total, completeness wins — you want every source's contribution, and the store is the only place they are already combined. For a specialist reading, provenance wins — a ring measuring HRV overnight has better data than a phone reporting what the ring told it.

Seeing who won

Health Sources. The today's-numbers table names the provider behind each value, which is the answer to "where is this number coming from" when two devices disagree.

The today's numbers table lists each metric with its value and the provider that supplied it. That is the answer to "where is this coming from", and it is the first place to look when a number surprises you.

The screen also lists apps writing into your health store that Base has no explicit row for. That is informational — a Polar or Withings user learns their data is already arriving without any setup — rather than something to act on.

Changing the winner

The priority order is fixed rather than configurable, which is a deliberate simplification: a per-metric source picker is a settings screen nobody maintains correctly, and getting it wrong silently corrupts the data.

What you can control is what reaches the store at all. Turning off a source's write permission in your platform's Health app removes it from consideration entirely, which is the effective way to change which device wins.

If you have two devices and one is clearly better, that is the lever.

Empty is not zero

A metric with no value anywhere stays empty. Base does not fall back to zero.

Zero is a claim — it says you took no steps, or your HRV was zero. Absent says nobody measured. Those are entirely different, and conflating them corrupts every average that reads the series afterwards.

If a metric is empty and you expected data, the same screen's per-type diagnostics distinguish an empty day from a permission that never landed. See fixing missing health data.

Common mistakes

Expecting two devices to combine. They never do, deliberately.

Wearing two devices for better data. It does not work that way — one wins, the other is ignored for that metric. The second device adds value only where the first has no data.

Assuming the newest device wins. The order is per metric and fixed.

Reading an empty metric as zero. It is not, and the distinction is load-bearing.

Not checking the winner when a number looks wrong. It is one screen and it names the source directly.

Leaving a retired device writing to the store. An old tracker still syncing can win a metric it is no longer well placed to measure.

What to do next

Open Health Sources and read the today's-numbers table. If a metric is coming from a device you no longer wear, revoke its write permission in your platform's Health app — that is the only way to stop it winning.