When updating the test lifecycle tags, we only go through the tests present in the report for the reliable period to find tests combination to mark as reliable.
This means that a test that has a failure rate above the threshold for the unreliable period can be marked as unreliable even though it has not run at all during the reliable period.