Ping, Latency and Server Location for Game Servers Explained
Ping is the round-trip time, in milliseconds, between your PC and the game server. For most survival, co-op and building games, a steady ping under about 60 ms feels fine, and anything under 100 ms is workable. Fast PvP gunfights want lower. The part you can't change with settings is physical distance, which sets a hard minimum.
This guide explains the terms people mix up (ping, latency, jitter, packet loss), what numbers are playable, how server location and routing affect them, how to measure your own connection, and which problems a hosting provider can and can't fix.
- Ping is round-trip time. Jitter is how much it varies, and packet loss is data that never arrives. Jitter and loss often hurt more than a slightly higher average ping.
- Rough guide: under 30 ms is excellent, 30-60 ms is good, 60-100 ms is playable, and over 100 ms is frustrating for fast real-time games.
- Light in fibre adds roughly 10 ms of round trip per 1,000 km of cable, so distance alone puts a floor under your ping. Real routes are longer than a straight line.
- Test with
ping, thentracertorpathping(Windows) ortraceroute/mtr(macOS/Linux). Loss that starts at one hop and continues to the last hop is real. Loss at a single middle hop usually isn't. - A host can fix location, server-side lag and attack traffic. It can't fix your Wi-Fi, your ISP's routing or the speed of light.
On this page
What ping, latency, jitter and packet loss mean
Latency is the time data takes to get from one point to another. Round-trip time (RTT) is the time for a request to go out and the answer to come back, which is double the one-way latency (Cloudflare's explainer uses the same definition). The ping you see on a scoreboard or in the server browser is a round-trip figure.
| Term | What it measures | What it feels like when it is bad |
|---|---|---|
| Ping / RTT | Time for a packet to reach the server and return, in ms | Delayed reactions: you shoot, then the hit registers late; doors and chests open after a pause |
| Jitter | How much the ping changes from one moment to the next | Rubber-banding and inconsistent hit registration, even when the average ping looks fine |
| Packet loss | Percentage of packets that never arrive | Stutters, items snapping back, players teleporting, and eventually disconnects |
A connection that sits at a steady 70 ms is usually easier to play on than one that averages 35 ms but spikes to 200 ms every few seconds. When someone says "my ping is fine but the game lags", jitter or loss is the first thing to check.
What ping is playable?
There is no universal cut-off, but the commonly used bands (the same ones T-Mobile's gaming latency guide uses) look like this:
| Ping | Rating | What to expect |
|---|---|---|
| Under 30 ms | Excellent | Competitive shooters and fighting games feel instant |
| 30-60 ms | Good | Fine for nearly everything, including most PvP |
| 60-100 ms | Playable | A slight delay is noticeable in aim-heavy fights; fine for building, farming and PvE |
| Over 100 ms | Frustrating | Fast real-time games suffer; around 200 ms and up many become unplayable |
Genre matters. Co-op survival, crafting and base-building games forgive a lot more than a game where you win or lose on the first shot. Hosts of PvP servers hear about ping first because a 40 ms difference decides gunfights, while on a friends-only building server nobody notices 80 ms.
How server location affects ping
Light travels through optical fibre at about two-thirds of its vacuum speed, roughly 200,000 km/s. That works out to about 5 microseconds per km one way, or about 10 microseconds per km for a round trip, so 1,000 km of cable costs you around 10 ms of ping (see the GEANT propagation delay note). No hardware upgrade removes that.
Cables don't run in straight lines and every router adds a little processing time, so real pings land above the physical floor. As a reference, Cloudflare gives roughly 5-10 ms for users about 100 miles from a data centre and roughly 40-50 ms for users about 2,200 miles away. The table below is the straight-line floor from Chicago, rounded; treat real-world numbers as that floor plus 10-40 percent or more, depending on the route.
| From Chicago to | Straight-line distance | Physical minimum round trip |
|---|---|---|
| Atlanta | about 590 mi / 950 km | about 9 ms |
| New York City | about 710 mi / 1,145 km | about 11 ms |
| Dallas | about 805 mi / 1,295 km | about 13 ms |
| Denver | about 920 mi / 1,480 km | about 15 ms |
| Miami | about 1,190 mi / 1,920 km | about 19 ms |
| Seattle | about 1,730 mi / 2,790 km | about 28 ms |
| Los Angeles | about 1,740 mi / 2,800 km | about 28 ms |
That is the case for central US hosting. Chicago sits roughly in the middle of the country, so the East Coast, the South and the Mountain states are all within about 1,200 miles, and the West Coast is under 1,800. A group spread across both coasts gets a tolerable ping for everyone instead of one side getting excellent and the other poor. ChicagoServers hardware is in two Chicago facilities (Digital Realty at 680 S. Federal St and Equinix CH1 at 350 E. Cermak Rd), so this is the geography your players get.
How to test ping and trace the route to your server
Use the server's IP address from your control panel. Run each test for a minute or more, and again at the time of day when you normally play.
- Measure round-trip time and loss
Open a terminal or Command Prompt. On Windows,
-nsets the packet count; on macOS and Linux it is-c. Read the min/avg/max and the percentage lost.ping -n 50 SERVER_IP (Windows) ping -c 50 SERVER_IP (macOS / Linux) - Trace the path
Windows
tracertlists every router between you and the server;/dskips name lookups so it finishes faster (Microsoft's tracert reference). On macOS and Linux usetraceroute.tracert /d SERVER_IP traceroute -n SERVER_IP - Measure loss per hop over time
pathpingon Windows sends repeated probes to every hop and reports loss for each router and link, then takes a minute or two to finish (Microsoft's pathping reference). On macOS and Linux,mtrdoes the same job if it is installed.pathping /n SERVER_IP mtr -rwc 100 SERVER_IP
How to read the output:
- Look for where latency jumps. A big step between two hops in different cities is normal distance. A big step between hops in the same area, with high numbers carried all the way to the end, points at a congested link.
- Loss that persists to the last hop is real. If hop 6 shows 30 percent loss but hops 7 through the server show none, that router is just deprioritising probes addressed to itself. Microsoft notes that routers that drop such packets may still forward other traffic normally.
- Rows of
*aren't automatically a problem. They mean a router didn't answer the probe, not that traffic isn't passing through it. - Some servers don't answer ping at all. If ping times out but the game connects, trust the in-game ping instead.
What a host can and can't fix
Separating network lag from server lag saves a lot of guessing. If everyone's ping is low but the game stutters, entity counts are high or the world is heavy, the problem is on the server. If ping or loss is bad for one player only, it is usually on their side of the connection.
| A host can | A host can't |
|---|---|
| Place the server near your players (location) | Make light travel faster, or shorten the distance between you and the data centre |
| Keep the machine from being overloaded: ask whether the host oversells its nodes or caps CPU and RAM. | Fix your Wi-Fi, a congested home router or a saturated upload |
| Absorb attack traffic that would otherwise cause spikes. 40Gbps DDoS protection is included | Control how your ISP hands traffic off to other networks |
| Show you server health: the server's CPU, memory and player count help you tell server lag from network lag | Make a badly written plugin or an enormous world cheaper to simulate |
On your side, the usual fixes are a wired connection instead of Wi-Fi, pausing cloud backups and large uploads while you play, and trying a different time of day if the trace shows congestion. A VPN occasionally finds a better route around a bad ISP handoff, but more often it adds distance, so test with and without it.
Frequently asked questions
What is a good ping for a game server?
Under 60 ms is good for almost every game, and under 30 ms is excellent for competitive shooters. For building, farming and PvE, 60-100 ms is playable. Steady beats low: spikes and packet loss hurt more than a moderately higher, stable ping.
Why is my ping high even though the server is close?
Distance isn't the only factor. Wi-Fi, a busy home network, your ISP's routing or a VPN can all add latency. Run a traceroute or pathping to see where the delay builds up, and check the server's CPU and player graphs to rule out server-side lag.
Is ping the same as latency?
In practice, yes. Strictly, latency can mean one-way delay, while ping (round-trip time) is the out-and-back measurement that games and tools display, so it is about double the one-way figure.
Does a VPN lower ping?
Rarely. A VPN adds a hop and often extra distance. It can help only when your ISP's normal route to the server is unusually poor, so compare the ping with and without it before settling on one.