Selected server, connection delay, lost game data and visual stutter.
COCKPIT // PRIVATE DIAGNOSTIC SESSION
Trace the symptom.
Test one change.
Start with what the game feels like. Your answers suggest one safe first check. Retest the same conditions to see whether that lead explains your symptom.
● Your answers and calculated results stay in this tabStep 1 · Describe the symptom
Tell us what the game feels like.
Use numbers shown inside the game when available. If you do not know a number, leave it blank instead of guessing.
Step 2 · Optional browser-path check
See how browser requests behave when the connection is busy.
Use this only when the first diagnosis points to Wi-Fi or household traffic. We compare a quiet moment with controlled downloading and uploading.
Step 3 · Check whether a change helped
Compare game readings before and after.
Enter three readings before the change and three after it. Every run stays in the result. The middle reading shows the typical run, and the worst run and the lowest-to-highest spread show whether a spike is still there.
Step 5 · Keep the raw evidence
Use a game-specific six-run sheet.
Each template includes game version, server or realm, connection, before-and-after setup labels, symptoms and evidence filenames.
Continue with the current game
Published game guides keep the in-game context attached.
Telemetry HUD, subtick jitter, misdelivery and SDR relay checks.
Matchmaking region, Net Debug Stats and texture streaming isolation.
Performance Display, four red telemetry icons and Data Center selector.
Riot shard routing, Ctrl+F overlay checks and Movement Prediction.
LAT vs IND interpolation telemetry and Ctrl+Shift+N NetGraph.
Raid, map, party leader, device and lost-data clues.
Home and World response time, realm, instance and add-ons.
Account server, solo or co-op state, platform, device heat and slowdown.
Display Network Information HUD, Valve SDR relay and In/Out loss.
Four yellow/red connection icons, Data Center code and jitter checks.
F1 perf 6 console, Unity GC pauses, wipe-day host tick and ping.
Make the comparison fair
Change one thing, then check the same situation.
Read the middle, the worst, and the spread
The middle reading is the typical run. A spike still counts: the worst run and the spread from lowest to highest stay in the result. Three readings do not establish statistical significance.
Keep every run
Do not delete a result just because it is inconvenient.
Match the conditions
Keep the device, server and game mode the same.
Confirm inside the game
A better browser result is useful only when the original game problem improves too.
Turn a ping log into a readable support summary.
Paste a complete English Windows ping run, including every reply, timeout and the packet summary. Get median, p95, slowest reply and timeout counts. Add a second run only for the same destination after one change. This reads ICMP logs; it does not measure game packets.
How to collect a compatible log
In Command Prompt, run ping /n 30 1.1.1.1 and wait for the summary. This is Cloudflare's public resolver, not your game server. For the local link, find your active adapter's default gateway with ipconfig, then run ping /n 30 YOUR_GATEWAY. Use the actual gateway in place of YOUR_GATEWAY. Compare gateway runs separately from public-control runs. See the full walkthrough and interpretation limits.
Other languages, Linux/macOS output, merged runs and summary-only logs are not supported. Destination-unreachable or local-send errors need separate diagnosis and will not be treated as ordinary reply loss.
Ready to analyze. Logs stay in this tab; no log text is uploaded.
The exported summary omits raw logs and destination addresses. Median and p95 use received echo replies only; timeouts remain separate. Less-than values are reported as upper bounds. Read the calculation method. These two runs cannot prove causation or a durable improvement.