Preserve the denominator
A result of zero lost packets out of 100 probes is different evidence from zero out of 10,000. Both can display 0%, but the number of observations and their timing differ. Do not quote only the percentage when support needs to understand the sample.
An illustrative one lost probe out of 100 is 1% for that sample. It is not automatically the loss rate of every application or every later minute. Bursty loss can also matter differently from evenly spaced loss, so preserve timing where the tool exposes it.
Check what traffic was measured
Diagnostic probes may use a different protocol, packet size, destination or handling policy from media requests. Some devices deprioritize or ignore probes while forwarding ordinary traffic. An unanswered probe is not always proof that an equivalent media packet was lost.
Conversely, a clean probe sample to one server cannot certify another destination's path. Record which endpoint was tested and whether the television or a different device generated the traffic. Avoid calling a phone's result a television measurement.
Align the sample with the symptom
If the fault is intermittent, a short clean test hours later provides limited context. Record the original playback time and run an appropriate supported observation near a recurrence where practical. Do not generate an uncontrolled flood of probes or traffic to create a more impressive denominator.
Compare a known local path only when the equipment supports it and keep the ordinary viewing outcome visible. A loss counter without a picture/sound timeline can send the investigation toward a network event that the player actually recovered from.
State the limited conclusion clearly
A useful report says “0/100 probes lost to this endpoint at this time; the original programme stalled later at this timestamp.” That keeps both observations intact. “No packet loss anywhere” overstates the evidence and makes later troubleshooting harder.
If loss repeats under a controlled local comparison, provide the result and conditions to the network administrator. If only one resource fails at a fixed point while the network sample is clean, retain the resource-specific clue. Neither measurement should erase the other.
Related questions
Does one lost probe prove my internet connection is unusable?
No. Interpret the method, sample, timing and actual application outcome before drawing that conclusion.
Sources and scope
Scope notes distinguish published documentation, supplied information and editorial methods. They do not report a hands-on service test.
- RFC 2680 — Relevant protocol definition only; household workflows and numerical examples are editorial.
