Jitter, Latency, Ping, and Packet Loss for Gamers
TL;DR. Latency is how long a packet takes to get from your machine to the game server. Ping is the same number reported by the pingtool — round-trip in milliseconds. Jitter is how much that number varies between samples. Packet loss is the percentage of packets that never come back. Average ping is the one everyone quotes, but jitter and loss decide whether your shots land.
The four definitions, properly
Latency
Latency is the time it takes a packet to travel from your device to a destination, measured one way. For a packet from your laptop to a game server in Frankfurt, latency is the sum of every serialisation delay (turning the packet into bits on the wire), every propagation delay (electrons or photons crossing fibre), every queueing delay at each router, and the final processing delay at the destination. You can\u2019t measure one-way latency without synchronised clocks at both ends, which is why almost nobody does — instead we measure round-trip and divide by two.
Ping
Ping is the round-trip-time reported by an ICMP echo request and reply (the original ping from BSD). In gaming context “ping” usually means “round-trip latency to the game server” regardless of which protocol measured it. RTT and ping are interchangeable for our purposes.
Jitter
Jitter is the variation in latency over time. If you ping a server 20 times and the round-trips are 23 / 25 / 23 / 80 / 24 / 21 ms, your average is fine but your jitter is poor — that one spike at 80 ms means a player or a vehicle warps. There are two common ways to quote jitter: the IETF RFC 3550 definition (a smoothed inter-arrival variance used inside RTP), and the simpler Cisco-style “mean absolute deviation from the average”. Network tools usually report the latter; our widget below does the same.
Packet loss
Packet loss is the percentage of packets your endpoint sent that never came back. A small amount is normal — Wi-Fi loses packets routinely, congested links drop the rest. Game protocols are built to handle a few percent transparently via prediction and rollback. The crossover where it becomes noticeable is around 1 %; beyond 3 % games start visibly stuttering.
Measure yours, from this page, right now
The widget below pings a tiny asset on this domain 16 times in sequence and reports the same four numbers a CLI ping -c 16 would. The route from your browser to us isn\u2019t the same as the route to your game server, but the quality of your access network is — if you see consistent loss or large jitter to us, your game will see it too.
We ping /favicon.ico on this domain 16 times at 220 ms intervals and compute the same stats a CLI ping would. This measures the path your browser takes to us — handy for spotting problems in your access network, but not a substitute for a real game-server ping.
Two things worth understanding about this measurement: browser fetch includes TCP and TLS setup overhead the first time it talks to a host, so the very first sample is usually slower than it needs to be. We also can\u2019t send ICMP from JavaScript (no raw sockets), which means HTTP round-trip is a slight overestimate vs the same path measured with real ping.
What 60 ms actually feels like
The same ping is a luxury in one game and a death sentence in another. The widget below lets you drop in a latency number — yours from the probe above, or one of the connection presets — and see where you sit on the comfortable / playable / unplayable curve for each game category.
Pick a connection profile or drag the slider to a number from your last speed test. The bars below show comfortable / playable thresholds for each game category; the marker shows where you sit.
A common myth worth disposing of: your monitor refresh rate doesn\u2019t directly hide ping. A 240 Hz monitor still draws what the game tells it to draw 50 ms ago. A high refresh rate makes you faster at reacting, but the round-trip to the server is the same as on a 60 Hz panel.
Fixing each one
High average latency
Average latency is mostly geography. The speed of light in fibre is about 200 km/ms; nothing on your end can beat that. What you can do: connect to a game server in your region (a server in the same city is ≤ 10 ms; the other side of the continent is 60–80 ms; the other side of the planet is 200+ ms). Wired Ethernet instead of Wi-Fi saves 1–10 ms. Closing background bandwidth hogs on your network saves a few more.
High jitter
Jitter is almost always a Wi-Fi or congestion problem. Switch to Ethernet first; that fixes 70 % of cases. If you must use Wi-Fi, prefer 5 GHz over 2.4 GHz (less interference) and channel auto-pick — many home routers ship with everyone defaulting to channel 1 or 6. The other common cause is bufferbloat: large packet buffers at your modem or router's upstream link that fill up under any load and balloon latency. Routers with FQ-CoDel / CAKE queueing (most OpenWrt builds, all modern eero / Ubiquiti gear) eliminate it.
Packet loss
Loss above 0.1 % usually comes from one of three places: a flaky cable / DSL / docsis line on the access side, congestion at a peering point upstream from your ISP, or a misbehaving Wi-Fi radio. Pings to your gateway (192.168.1.1 or your router's LAN IP) tell you whether the loss is local — if the gateway is clean and the upstream isn\u2019t, the problem is past your modem. ISP tickets get traction faster with three days of MTR output than with one snapshot.
Tools that help diagnose
- MTR (My Traceroute) shows per-hop latency and loss along the path. Single most useful tool for “is it me or my ISP”.
- Waveform Bufferbloat Test grades bufferbloat by loading the line and watching ping under load — gives you an A–F.
- Game-specific tools. Riot's Hextech Repair, Steam's server-pick reports, and your console's built-in network test all give you the path the game itself will see.
Throughput too?
Latency and bandwidth are independent — a gigabit fibre line with high jitter is worse for gaming than a 100 Mbps cable line that's steady. If you want to test the throughput side of your connection, run our Speed Test alongside the probe above. If your speeds are fine but games still feel bad, you almost certainly have a jitter or loss problem rather than a bandwidth one.