Three kinds of test, and what each really measures
Players use the word "ping test" for three quite different measurements. Confusing them is the source of most bad conclusions about game connections.
The in-game ping display
Most online games can show your current latency to the game server you are actually playing on, either permanently in a corner of the screen or through a network statistics overlay in the settings. This is the most honest number available to you, because it measures the real journey: your device, your network, your provider's route, and the real server, under real match conditions. Its limitation is that it only exists while you are connected to that game, and some games include a little server processing time in the figure. For judging your game connection, it is still the gold standard.
Ping and route tools
Tools such as the ping command send small test packets to an address you choose and report how long each round trip takes and whether any packets were lost. They are flexible, you can aim them at different destinations and watch stability over time, but they measure only the destination you aim at, and they use a type of traffic (ICMP) that some servers answer slowly or not at all. A ping tool is a torch: brilliant at what you point it at, silent about everything else.
Speed tests
Speed-test websites and apps measure something else again: how much data your connection can push to a nearby test server, and often the latency to that same nearby server. That nearby server might be in your own city. Your game server might be in another country. A glowing speed-test result therefore tells you your local connection is healthy, and almost nothing about your game's journey, which is why a "fast" result alongside a laggy game is not a contradiction. Our guide to reducing high ping in online games explains the distance and routing reasons in full.
| Test | What it measures | Its blind spot |
|---|---|---|
| In-game ping display | Your real latency to the actual game server, during play | Only exists inside that game; may include some server processing time |
| Ping command / ping apps | Round-trip time and loss to an address you choose | Only as relevant as the address you chose; some servers deprioritise replies |
| Speed test | Bandwidth, plus latency to a nearby test server | The test server is rarely where your game server is |
| Console network test | Connection health and latency to the console platform's servers | Not your game's servers; a good result does not clear a bad game route |
Test towards the game, not just towards "the internet"
The destination matters more than the tool. Every game is a different journey from your house to a different building, so a single generic test can never speak for all of them. Wherever possible, prefer measurements in this order: the in-game display first; a ping test to an address in the same city or region as your game server second; a generic speed test last, as a rough health check of the local connection only.
This is also why region choice belongs in your test notes. If the game has connected you to a distant region, your readings describe that distant journey, not your normal one. Test on the region you actually intend to play on, and write the region down beside the numbers.
Testing on Windows
Start in the game itself: enable its ping or network display and play for a few minutes, watching how the figure behaves rather than glancing once. Then, for a steadier view, use the tools built into Windows.
- Open Command Prompt (type "cmd" into the Start menu search) and type ping followed by a space and an address, your router's address for a home-network check, or a public server's address for a wider one. Press Enter.
- Read the reply lines. Each line is one round trip in milliseconds. The summary at the end gives you minimum, maximum and average, plus the percentage of packets lost. A healthy result is consistent: replies close together, nothing lost. A troubled one shows replies swinging widely, or requests timing out, those timeouts are packet loss, made visible.
- For a longer watch, add -t after the word ping (for example, "ping -t" followed by the address) and the test keeps running until you stop it with Control and C together. Leave it running in a window while you play a match on the same machine, and you can see whether your spikes in game coincide with spikes in the test.
- To see the journey, type tracert followed by an address. This lists the hops your data passes through, with timings for each. Treat it as a sketch, not a verdict: many routers deliberately answer these test packets slowly, so one slow-looking hop whose delay does not carry through to the final figure is usually not a fault at all.
Testing on console
Both PlayStation and Xbox include a built-in network test in their settings menus, which checks the connection, reports download and upload speeds, and confirms details such as NAT type. Run it, it is a good health check of the console's link to its platform's servers, but understand its limit: it does not test your game's servers, so a clean result does not prove your game connection is clean, and a poor game connection alongside a clean console test points away from your home network and towards the route or that game's servers.
For the real figure, use the game's own ping display where it offers one, exactly as on PC. And the most valuable console test is not a menu at all: play wired, play wireless, compare. The console versions of the isolation tests in our guide to troubleshooting unstable game connections depend on honest readings like these.
Testing on a phone
Many mobile games show a small ping readout on the match screen, often colour-coded, and that remains your most truthful measure on a phone. Speed-test apps on phones measure a nearby test server, with the same blind spot as on any platform. If you use a booster app on your phone, note that GearUP's own pages list a ping test among the app's free features, useful for a quick check of how the boosted route behaves, and worth comparing against the in-game figure rather than trusting on its own. Whichever tool you use, test on the connection you play on: a reading taken on home Wi-Fi says nothing about your mobile data connection at a friend's house.
How many readings, and how to record them
Ping is not a fixed property like your screen size. It breathes, varying from second to second and hour to hour, so a single reading is closer to a rumour than a measurement. The minimum standard used throughout this site is:
- Three readings, a few minutes apart, under identical conditions.
- Use the middle value as your result, and glance at the highest: a wide gap between middle and highest is instability introducing itself.
- Record five things beside every result: date and time, game, server region, connection type (wired or Wi-Fi), and the reading itself. A notes app is fine; a simple table is better.
- Keep conditions identical between tests you intend to compare. Different time, different region, different connection, different test.
Recording feels fussy until the first time it saves you. It is what lets you say "this fix helped" instead of "this fix felt like it helped", and the two statements have very different relationships with the truth.
Test packet loss and jitter at the same time
A ping figure alone can certify a miserable connection as healthy. Two companions must be checked with it:
- Packet loss, packets that never arrive. In a continuous ping test it appears as timed-out replies; the summary states it as a percentage. Any consistent loss above zero is worth attention, because even 1–2% loss is plainly visible in a fast game as stutter and snapping movement, even when the ping number looks good.
- Jitter, the variation between your readings. If your replies read 28, 31, 29, 30 ms, your connection is steady. If they read 22, 88, 25, 140 ms, the average may look acceptable and the game will feel dreadful. Our explainer on what network jitter is and how it affects gaming covers why the spread matters as much as the average, and how to think about the numbers you record.
When you write down a result, write all three: middle ping, loss percentage, and the rough spread between lowest and highest readings. Three numbers, honestly taken, describe a game connection better than any single score.
Peak hours versus quiet hours
Always test at the time you actually play. Congestion, on your provider's network, on the route, or at the game server, gathers in the evening and at weekends, and a test taken on a quiet Tuesday morning will certify a connection that collapses every Friday night. If your problems keep a schedule, your tests must keep the same schedule to see them. A useful habit for a connection you are unsure about: one set of readings at your normal playing time, one set at a quiet hour, recorded separately. The difference between the two is often the diagnosis.
The fair before-and-after test
Every fix you will ever try, a cable, a router setting, a different region, a routing tool, should pass the same examination. This is the procedure used across this site:
- Baseline. On your current setup, take three readings in your game, on your region, at your usual time. Record them.
- Change one thing. Only one. If you change three things, the test cannot tell you which one mattered.
- Retest identically. Same game, region, time of day and connection type, three readings, middle value. Compare all three numbers, ping, loss, spread.
- Repeat on another day before concluding. Connections have good and bad days for reasons nothing to do with your change; one evening's improvement can be weather, not progress.
- Be willing to find nothing. If the numbers do not improve consistently, the change did not help your connection, whatever it cost or promised. Revert it and test the next suspect.
This applies to boosters exactly as it applies to cables. If you test a routing tool such as GearUP Booster, run your baseline on your ordinary route first, then the identical test with the boost active, GearUP offers a free trial, so a careful comparison costs nothing but two evenings of method; trial terms change, so check the current offer rather than assuming its length. Judge by your recorded numbers across several sessions, never by one good match.
Frequently asked questions
Why does my speed test show 12 ms but my game shows 90 ms?+
Because they measured different journeys. The speed test connected to a server near you; your game server is further away, reached by a different route. Both numbers can be accurate at once, the in-game figure is the one that describes your gaming.
Is pinging a search engine's address a good gaming test?+
It is a fair basic health check of your connection to the wider internet, and useful for spotting loss or wild instability. It cannot tell you your latency to any particular game server, which may be in a different city or country entirely.
How many times should I test before trusting a result?+
Three readings a few minutes apart, using the middle one, is the minimum for a stable connection. If you are investigating instability, test repeatedly during the bad period itself, that is where the evidence lives.
Can I test my ping properly on a console without a PC?+
Partly. The built-in network test checks your connection to the console platform's servers, and the game's own ping display shows the real figure during play. What you cannot easily do on console is aim a ping test at a custom address, which is one more reason the in-game display matters.
What loss and jitter figures should I aim for?+
As a rough guide: packet loss at 0% is the goal, and consistent loss is a fault worth chasing at almost any level. For jitter, a spread of only a few milliseconds between readings feels rock-steady; once readings routinely swing by 30 ms or more, most players feel it clearly.
Testing is the least glamorous part of fixing a game connection, and the part everything else stands on. Learn what each test really measures, test towards the game itself, record three honest numbers at the time you actually play, and from then on, every fix, setting and booster you meet can be judged by evidence instead of by hope.
