Symptoms
The phone and wearable show plausible counts separately but the combined app total doubles.
A status icon is not the same as a successful result. The phone and wearable show plausible counts separately but the combined app total doubles. Reproduce the outcome itself before treating a setting as the cause.
- The failing path: The phone and wearable show plausible counts separately but the combined app total doubles.
- Control test: Walk a short counted distance with one tracker at a time and then both.
- A successful outcome: The combined app count approximates one counted walk, not twice the walk.
Why It Happens
Sensor history can contain both legitimate gaps and app display issues. Compare the device reading and timestamp with the phone rather than interpreting health trends.
There are three genuinely different layers worth checking; use the observation in the table to decide which one is implicated. None of these is proven merely by the symptoms.
- Sources summed instead of deduplicated. This explanation predicts a different result in the comparison test or the more advanced inspection.
- Duplicate device registration. This explanation predicts a different result in the comparison test or the more advanced inspection.
- Imported records overlap. This explanation predicts a different result in the comparison test or the more advanced inspection.
Before Making Changes
Do not delete historical health data to test a counting rule.
Save the current configuration or capture screenshots of the exact option values; where a file or recording is involved, work on a disposable duplicate. If this is shared equipment, ask the owner or administrator before changing it. Prefer a safe test item over real customers, medical decisions, paid transactions or irreplaceable data.
Write down the device model, platform, observed time of failure, and which action still works. These details will distinguish an unsupported feature from a malfunction and will be useful if escalation is required.
Problem-Specific Diagnosis
Use the following route rather than guessing based on a product name. The expected differences are designed for this particular failure.
| Action or observation | Meaning of the result | Decision |
|---|---|---|
| Walk a short counted distance with one tracker at a time and then both. | Does the failure follow the affected file, profile, connection or location? | Keep this result as the baseline; do not alter two systems at once. |
| Choose a primary step-data source or deduplication preference in the health aggregator. | Does the low-impact correction restore the exact failure without altering the control case? | Re-test; reverse the change if the original symptom is unchanged. |
| Inspect data-source priority and whether simultaneous steps are merged or summed by the app. | Does this separate duplicate device registration from imported records overlap? | Use documented diagnostic views rather than blind replacement. |
| The combined app count approximates one counted walk, not twice the walk. | Is the same original symptom gone after a normal repeat? | Record the pass condition and preserve the working settings. |
If the control case fails as well, the fault may be broader than this guide. Stop at that point and check the shared power, account, access or media path before applying the narrow fix.
Beginner and Advanced Paths
Start with the reversible Beginner path. The Advanced path examines a different layer when the first route fails; it is not a repetition of the same restart or setting.
Step-by-Step Troubleshooting
Choose the simple reversible checks first, then investigate the deeper system layer only if necessary.
Record the exact failure
Confirm the behavior without guessing: The phone and wearable show plausible counts separately but the combined app total doubles. Note the affected account, file, connection or physical control and record the current settings. Do not delete historical health data to test a counting rule.
Run the passing comparison
Walk a short counted distance with one tracker at a time and then both. Keep the working and failing cases separate. If both fail, stop and investigate the common path instead of forcing a narrow repair.
Apply the least disruptive correction
Choose a primary step-data source or deduplication preference in the health aggregator. Change only this setting or route; do not reset unrelated profiles, storage or pairing records. Save the previous value so it can be restored.
Retest and keep the result
The combined app count approximates one counted walk, not twice the walk. Repeat after a normal reopening or reconnect. If the control case gets worse, undo the last change and write down the difference.
Separate the three underlying layers
Test the hypotheses independently: sources summed instead of deduplicated; duplicate device registration; imported records overlap. A green status indicator does not distinguish them. Use the recorded control case as evidence.
Inspect the deeper component
Inspect data-source priority and whether simultaneous steps are merged or summed by the app. Use only read-only diagnostics or documented configuration screens first. Capture logs and timestamps rather than removing the component or reinstalling software blindly.
Cross-check with an independent path
Where possible, repeat “Walk a short counted distance with one tracker at a time and then both.” using a second supported account, safe sample file, physical input, known-good cable or alternate compatible application. Change only one variable and note which outcome follows it.
Choose a reversible targeted escalation
If the deeper check isolates a fault, use a vendor-supported fix or authorized service for that layer; otherwise restore original values. Finish only when: The combined app count approximates one counted walk, not twice the walk. Do not delete historical health data to test a counting rule.
Solution Checklist
Solution Checklist
Common Mistakes to Avoid
- Treating sources summed instead of deduplicated as certain just because the behavior looks familiar.
- Changing several independent settings together, which makes the result impossible to interpret.
- Skipping the safety boundary: Do not delete historical health data to test a counting rule.
- Calling the issue fixed after one status indicator changes but before repeating the original operation.
- Discarding the working configuration or the only original file before a replacement has been verified.
How to Verify the Fix
The combined app count approximates one counted walk, not twice the walk.
Use the same baseline case: Walk a short counted distance with one tracker at a time and then both. Then repeat the previously failing operation after a normal close-and-open, reconnect or next scheduled cycle, whichever matches the scenario. Confirm the part of the system that was already working is still unaffected.
Where a result can be measured, retain the test timestamp, file name, screenshot or device state. A temporary success that cannot be repeated is not yet a confirmed repair.
If the Problem Continues
Escalate from the surface symptom to the narrower mechanism: Inspect data-source priority and whether simultaneous steps are merged or summed by the app. If this identifies a hardware, account-policy or service limitation, use the manufacturer’s authorized support route rather than bypassing controls.
When opening a support case, supply the model, software/firmware version actually installed, the failing operation, the successful comparison, the three possible explanations checked, and the final observed result. Redact account identifiers, access codes, serial-number photos and personal information from public posts.
Prevent Recurrence
Keep a short note of the confirmed working configuration, particularly the setting implicated by sources summed instead of deduplicated. Repeat the disposable comparison after an app update, device replacement, major account change or new connection only when that change plausibly affects this issue.
Retain a recovery path: spare test material, a saved profile, a restore point, a backup copy or official service information, as appropriate for the device. Do not schedule risky resets as routine maintenance.
FAQ
Is sources summed instead of deduplicated definitely the reason this happens?
No. The same symptom can come from duplicate device registration or imported records overlap. First compare the affected and unaffected paths: Walk a short counted distance with one tracker at a time and then both. Do not assume any cause without a different outcome.
How can I safely rule out duplicate device registration?
Try the reversible correction only after documenting the control: Choose a primary step-data source or deduplication preference in the health aggregator. If the behavior does not change, restore the old value and use the deeper observation: Inspect data-source priority and whether simultaneous steps are merged or summed by the app.
What if imported records overlap turns out to be responsible?
Do not force an unsupported setting or a factory reset. Keep the evidence and ask the manufacturer or responsible administrator to investigate the isolated layer. The measurable pass condition remains: The combined app count approximates one counted walk, not twice the walk.