Define the observation first
Packet loss means packets did not arrive as expected for a particular flow and measurement interval. A frozen call or stalled download is a symptom, not by itself a loss measurement. Note the affected application, direction, time and any reported error before assigning the problem to Wi-Fi, a cable or your provider.
Use an appropriate measurement
VelocityVerify estimates HTTP round-trip timing, not UDP or ICMP packet timing. It does not measure a separate upload-loaded latency series or end-to-end packet loss. New letter grades are withheld. See the measurement method and validity rules before interpreting a result. Use the affected application’s own network statistics where available, or a diagnostic tool with documented packet type, destination, interval and loss calculation. Do not compare percentages from different methods as if they described the same flow.
Read percentages with their denominator
Keep sent and received counts, observation duration and the pattern of missing replies. A short burst and a few isolated misses can affect applications differently. This page does not assign a universal good/bad threshold. Check whether the application recovered, whether the symptom repeated and whether the diagnostic endpoint may handle probes differently from application traffic.
Do not convict an intermediate hop
A traceroute hop that does not reply is not sufficient evidence that it is dropping forwarded application traffic. RFC 1812, section 4.3.2.8 allows rate limiting of ICMP error messages and notes resource-related limits on echo replies. Compare destination behavior and repeated observations; do not treat asterisks or an isolated intermediate-hop percentage as an exact fault location.
Make a controlled wired comparison
If practical, test the same device over Ethernet and record the full path, including adapters, docks and switches. Inspect loose connectors and reported link errors; try one known-working component at a time. Do not replace the whole network or force interface settings solely because a web page loads slowly.
Keep the wireless comparison specific
Repeat the original task near the access point and at the troublesome location, noting the band, signal and active clients where visible. Changing the device and location together makes a difference harder to interpret. Better behavior in one wireless condition narrows the investigation but does not establish a universal cable-versus-Wi-Fi ranking.
Send a reproducible support record
Provide timestamps, destination, connection method, application symptoms, diagnostic settings and destination-level results. Include any modem or router errors without publishing account details or raw logs containing private information. Ask support to investigate the documented pattern rather than naming a guilty hop from one traceroute.