Latency for Gaming: What Ping You Need and How to Test

Quick answer: in online games, latency (usually called ping) is the time a small message takes to travel from your device to the game server and back, measured in milliseconds (ms). As general guidance, under 50 ms feels instant in almost every genre, 50 to 100 ms is fine for most players, and above roughly 100 to 150 ms fast games such as shooters and fighting games begin to feel unfair. Jitter (variation in latency) and packet loss (messages that never arrive) often cause more visible lag than a slightly higher average. This guide explains how games hide latency, what ranges to aim for by genre, and how to test your connection.

Important: ping is not the same number as GlobeLatency's HTTPS latency. The ranges on this page describe round-trip time to the server you actually use (what ping, or a game's own ping display, shows). GlobeLatency times full HTTPS requests from the browser, which include DNS, TCP and TLS setup on a fresh connection (roughly three round trips) plus server and browser time, so its numbers are usually higher than ping to the same server, often 1.5 to 3 times on a cold connection. Judge GlobeLatency results by the colour and the comparison with the world average, and use these ranges for ping or in-game latency. See ping vs HTTPS latency.

Ping, latency and lag: what is the difference?

Latency is delay. Ping is most often the round-trip form of that delay: your packet goes out, the server answers, and the elapsed time is the ping. A game's on-screen "ping" is typically measured inside the game's own protocol, so it can include the server's processing time. The ping command in a terminal uses ICMP echo messages instead, and a browser tool such as GlobeLatency uses HTTPS requests. All three are "round-trip" numbers, but they are not interchangeable (see the section on in-game ping versus HTTPS latency below).

Lag is the player's experience, not a single measurement. You can experience lag from high latency, from jitter, from packet loss, from a slow client (low frames per second), or from an overloaded server. When a player says "I'm lagging", the cause could be any of these. For a general introduction, see what is latency, and for the difference between delay and capacity see latency vs bandwidth vs throughput.

How online games deal with latency

Almost every real-time multiplayer game uses a client-server design: the server holds the authoritative state of the match, and each player's client sends inputs and receives updates. Because the information takes time to travel, games use several techniques.

Tick rate

The server does not simulate continuously; it advances the game in discrete steps called ticks. The tick rate is how many steps per second. Some competitive shooters run servers at 64 or 128 ticks per second; many large or mobile games run at 20 to 60. The interval between ticks is 1000 ÷ tick rate:

Tick rate	Interval between updates	What it means
20 per second	50 ms	The server state can be up to 50 ms stale before network delay is added
30 per second	33.3 ms	Common for large-scale games
64 per second	15.6 ms	Common in competitive shooters
128 per second	7.8 ms	Finer granularity; benefits depend on the game

A low tick rate and high ping add together. With a 50 ms ping and a 20-tick server, an event may reach you roughly 25 ms (half the round trip) plus up to 50 ms (waiting for the next tick) after it happened, before your screen even draws it.

Client-side prediction

If your character waited for the server to confirm every keypress, you would feel your own ping on every movement. Instead the client predicts the result immediately and corrects it if the server disagrees. That is why your own movement feels smooth even at 80 ms, while other players seem to teleport or rubber-band when the correction arrives.

Interpolation and lag compensation

Other players are drawn slightly in the past, blended between received updates (interpolation), which hides gaps between ticks. Many shooters also use lag compensation: when you fire, the server rewinds other players to where they were on your screen at that moment, within a limit, then decides whether the shot hit. This makes aiming feel fair for the shooter, but it means a low-ping player can be hit "after" reaching cover on the high-ping player's screen. The practical effect: high ping hurts you mostly through peeker's advantage, delayed reactions to enemies, and corrections, not through everything feeling slow.

What is a good ping for gaming? Ranges by genre

The table below is general guidance, not a rule from any one game. Individual games differ in tick rate, netcode and tolerance.

Genre	Comfortable	Noticeable	Often frustrating	Why
Competitive FPS (tactical and arena shooters)	under 40 ms	50 to 80 ms	above about 100 ms	Tiny timing windows; aim and peeking depend on fresh state
MOBA and RTS	under 60 ms	70 to 120 ms	above about 150 ms	Commands are queued and abilities have wind-up, so a little delay is hidden
MMO and RPG	under 100 ms	100 to 200 ms	above about 250 ms	Abilities usually have cooldowns and animation times
Fighting games	under 30 ms with rollback netcode	40 to 80 ms	above about 100 ms, or any unstable link	Frame-exact inputs (a frame is 16.7 ms at 60 fps); delay-based netcode suffers most
Racing	under 60 ms	80 to 150 ms	above about 150 ms	Cars are predicted well, but contact between cars shows corrections
Turn-based and card games	almost any	above 300 ms	only when disconnects occur	Latency has little effect on play

For context across all uses, see what is a good latency.

Worked example: a 60 ms player versus a 15 ms player

Suppose you have 60 ms ping and an opponent has 15 ms, on a 64-tick server (15.6 ms per tick). Roughly, your actions reach the server 30 ms after you press the button, the opponent's after 7.5 ms. Add the tick wait (up to 15.6 ms) and interpolation delay on each side. The difference between you is on the order of 45 ms plus rounding, about three frames at 60 fps. That is enough to lose some close duels, but it is smaller than typical human reaction-time variation of a few tens of milliseconds, which is why a competent player on 60 ms still competes well while a player on 200 ms with frequent spikes does not.

Why jitter and packet loss matter more than the average

Jitter is the variation in latency from one packet to the next. Packet loss is the share of packets that never arrive. Games send many small updates per second (often 20 to 128 packets each way), so a bad stretch is felt immediately.

• Stable 70 ms: prediction and interpolation work well. Most players adapt within a minute.
• Average 40 ms but bouncing between 15 and 140 ms: interpolation buffers underrun, other players stutter or teleport, and shots register inconsistently.
• 1 to 2 percent loss: visible as rubber-banding, because the client must guess through missing updates. Loss that comes in bursts is worse than the same average loss spread evenly.

Never judge a connection by the average alone. Look at the best and worst values, and at a high percentile (see percentiles p50, p95 and p99).

Choosing a server region

Distance sets a floor. Light in fibre travels roughly 200,000 km/s, about 5 microseconds per kilometre one way, or about 1 ms of round-trip time per 100 km of cable path. Real cable routes are longer than straight lines, and every router adds a little.

Cable path distance (one way)	Minimum round trip from physics alone	Typical real-world round trip
500 km	about 5 ms	roughly 8 to 20 ms
2,000 km	about 20 ms	roughly 30 to 45 ms
6,000 km	about 60 ms	roughly 80 to 110 ms
12,000 km	about 120 ms	roughly 150 to 220 ms

The "typical" column varies by route; treat it as an order of magnitude. See undersea cables and the speed of light for why. Practical rules: let the game pick the lowest-ping region, do not assume the geographically nearest region has the best ping, and if you play with friends overseas, choose the region that minimises the highest ping in the group.

Wi-Fi versus Ethernet for gaming

Wi-Fi is shared radio. Every packet can be delayed by contention with neighbours, microwave ovens, walls and your own other devices, and lost frames are retried invisibly, which shows up as jitter. Ethernet has none of that. Typical differences, which vary a lot by home:

• Good 5 GHz Wi-Fi, close to the router: perhaps 2 to 10 ms extra latency and slightly more jitter than Ethernet.
• 2.4 GHz Wi-Fi in a crowded building, or through two walls: often 10 to 50 ms extra with random spikes of 100 ms or more.
• Powerline adapters: usually between the two, strongly dependent on wiring quality.

If you cannot run a cable, use 5 GHz or 6 GHz, sit in line of sight with the access point, and avoid doing large downloads on the same network while playing. More detail is in Wi-Fi, Ethernet, fibre, mobile and satellite latency.

NAT type, CGNAT and connection problems

Most home networks use NAT (network address translation): many devices share one public IP address, and the router rewrites addresses. Consoles often label this "Open", "Moderate" or "Strict" NAT. It mostly affects whether you can connect directly to other players in peer-to-peer matches, and which lobbies you can join, rather than raising latency in server-based games.

CGNAT (carrier-grade NAT) means your ISP shares one public IPv4 address among many customers, so you cannot forward ports. Symptoms: strict NAT type that no router setting fixes, trouble hosting games. Solutions include asking the ISP for a public IP, using IPv6 if the game supports it, or choosing dedicated-server games. Details are in IP addresses, IPv4, IPv6 and NAT. NAT type is not a latency measurement, but a Strict type can force relayed connections that add delay.

In-game ping versus HTTPS latency

Game traffic is usually UDP, small packets, sent straight to a game server. A browser test uses TCP and TLS requests to a web server. Consequences:

• A browser HTTPS number is usually a few milliseconds to tens of milliseconds higher than ICMP ping, because it includes DNS, connection setup and web server response (see ping vs HTTPS latency).
• The game's own ping may be measured to a specific machine whose route differs from a web CDN near you.
• Web destinations in your region tell you about your access network (Wi-Fi, home router, ISP), which is shared by the game. They cannot tell you whether the game provider's servers are healthy.

Use the browser test for relative information: is my connection stable, is my latency to region A lower than to region B, did my change improve it.

How to test your connection for gaming with GlobeLatency

1. Close downloads and streaming on other devices, and connect by cable if you can.
2. Open the free latency test, select the Gaming destinations (a category or template of well-known game-related servers where available) plus a few nearby regions, and raise the passes to 10 or 20 so that worst-case and 95th percentile values mean something.
3. Read the average for each destination, then the best and worst pass. A worst pass more than three times the average is flagged as unstable.
4. Save and compare. The report link lets you re-run the same test after a change, such as switching from Wi-Fi to Ethernet. See running a test and reading your results.
5. Before a long session, open the monitor and watch a few destinations for several minutes. It re-measures continuously and shows the 95th percentile and the no-reply percentage over recent readings. The live map shows the same data geographically.

Remember GlobeLatency measures HTTPS round trips from a browser, so numbers will be somewhat higher than an in-game ping, and a background tab may be throttled by the browser. See accuracy and limitations.

Common mistakes and myths

• "A faster plan lowers my ping." Only if your line is saturated. Gigabit versus 100 Mbit makes little difference to the latency of a quiet link.
• "Lowest ping always wins." Below about 30 to 40 ms, other factors (frame rate, input device, server tick) matter more.
• Testing once. Evening congestion and a neighbour's microwave do not show up in a single quick test.
• Ignoring bufferbloat. A slow upload while someone uploads video can add hundreds of milliseconds. See bufferbloat explained.
• Using a VPN "to reduce ping". Usually raises it; see VPN and latency.
• Comparing different tools directly. Compare like with like.

What to do next

If your ping is worse than the table suggests, follow the how to reduce latency checklist in order, starting with Ethernet. If the problem is intermittent or only affects some servers, use troubleshoot high latency. If you have evidence the problem is outside your home, read talking to your ISP.