NetDiag+
Why you lag when the speed test says everything is fine
A speed test gives you one number about an empty pipe. A game or a voice call cares about three numbers about a full one: how far the server is (median), how much that distance wobbles from packet to packet (jitter), and how many packets never come back (loss). It also matters which server — the region your game actually runs in, the CDN edge that serves your video, your own carrier — not a speed-test host chosen to look good. The Latency Dashboard in NetDiag+ pings the services people in your region really use, sorted worst-first, with all three numbers per row.
Three steps that tell you what is actually wrong
- Check your own router first. Ping the gateway before anything else. A healthy Wi-Fi link answers in a few milliseconds and barely wobbles. If the gateway itself jitters by 15 ms or more, every remote number below it is inflated by your Wi-Fi — move closer, leave Low Power Mode, re-run. Blaming a server before the router is checked is how people buy a faster plan that changes nothing.
- Read three numbers per server, not one. Median is geography — a far region stays far, and the fix is choosing a nearer one in the game. Jitter is stability, and it is the number that makes games rubber-band and voices robotic; 30 ms of jitter on a 40 ms path is worse than a steady 120 ms. Loss is packets that never returned — that one deserves a traceroute to find the hop where they vanish.
- Know what kind of number you are looking at. Cloud regions drop ICMP by policy, so they are measured by a TCP handshake instead — always higher than a ping, and not comparable to one. NetDiag+ labels those rows. And none of this is your in-game ping: the game adds its own server tick rate and matchmaking on top. What you are measuring is the network path to the place where those servers run — which is exactly the part your provider is responsible for.
Free · 27 tools · one-time purchase removes ads · no subscription
Questions people actually ask
Is this my in-game ping?
No. It is the network round-trip to the region where the game's servers run. The game adds server tick, matchmaking and its own protocol on top, so in-game ping is always somewhat higher. But if the network path is 40 ms with 3 ms of jitter and the game shows 140, the problem is in the game or the server — not your connection.
Why does a nearby server show Unstable while a far one is Stable?
Because jitter and distance are independent. A server 250 ms away with 5 ms of wobble is far but predictable. A server 40 ms away with 35 ms of wobble is close and unplayable. When several nearby rows are unstable at once, look at the gateway row: if it wobbles too, the cause is your Wi-Fi, not the internet.
My speed test shows 300 Mbps, but PUBG or Valorant rubber-bands. Why?
Bandwidth is not latency. Rubber-banding is jitter or loss, and a 300 Mbps line can have both. It also gets worse the moment someone in the house starts a download — that is bufferbloat, a separate test.
A row says No reply. Is the server down?
Not necessarily. Many hosts refuse ICMP but answer on their real service port, and NetDiag+ tries that second and labels the row "filtered" when it works. A row that fails both ways is genuinely unreachable from your network right now. Regions do get degraded — in 2026 the Bahrain cloud region that hosts several Middle-East game servers was one — and a red row next to a green neighbour is the diagnosis, not an error.
Which servers should I care about?
The ones your apps actually talk to: the cloud region hosting your game, the CDN edge of the apps you use, and your own carrier. That list is different in Riyadh, Istanbul, Moscow and Cairo, which is why the dashboard picks it from your App Store region rather than shipping one global list. Compare rows against each other, not against a number from a forum.
Stable-looking latency can still collapse the moment someone starts a download. Next: bufferbloat — latency under load →