See what ping, jitter and packet loss ranges are good for online gaming, then diagnose lag spikes on console, PC and cloud gaming without relying only on download speed.
Gaming latency guide
A smooth gaming connection is not only about download speed. Ping controls reaction delay, jitter controls consistency, packet loss causes stutter, and bufferbloat creates lag when the connection is busy.
Use this page as the benchmark. It gives practical target ranges, then shows how to test consoles, PCs and cloud gaming without sending users away to a separate generic guide.
Recommended tests
A good average can hide jitter, packet loss or large spikes when another device starts using the connection.
Target ranges
For gaming, consistency matters as much as the headline number. A stable 40 ms connection can feel better than a 20 ms connection that spikes to 150 ms whenever someone uploads, streams or downloads.
| Gaming metric | Excellent | Good / playable | Poor / investigate |
|---|---|---|---|
| Idle ping | Under 20 ms | 20–50 ms | 80–100 ms+ or regular spikes |
| In-game ping | Under 30 ms | 30–60 ms | 100 ms+ in fast games |
| Jitter | Under 5 ms | 5–15 ms | 20 ms+ bursts or uneven spikes |
| Packet loss | 0% | Occasional tiny blips only | Repeated loss or 1%+ during play |
| Loaded latency | Less than 30 ms above idle | 30–80 ms above idle | 100 ms+ above idle when line is busy |
Interactive risk profile
Choose the game style, local connection and server region. The result estimates connection latency and compares the likely range with the tolerance of that gaming workload. Confirm it with the LinkSpeed Ping Test because Wi-Fi interference, provider routing and peak-time congestion can change the live result.
This is a diagnostic estimate, not a speed-test result. Game type changes the risk threshold; connection and region change the expected latency and variability.
Likely range: 11–16 ms
A wired connection to a nearby server should suit fast competitive play, provided jitter and packet loss stay low.
Run a live test before changing equipment.
Measure live gaming latency and jitterPlatform and NAT checks
Different platforms expose different parts of the connection. Check the platform screen first, then use the LinkSpeed tools to separate local Wi-Fi, packet loss, loaded latency and external routing problems.
Use the PS5 connection test for NAT type and basic throughput, then check latency inside the game where available. A Type 3 result is restrictive, but it does not by itself prove Double NAT. Pause console downloads before judging match latency.
Open the Xbox network statistics and NAT screens. Check latency, packet loss and whether the console reports Double NAT detected. A stable average with repeated packet loss can still cause rubber-banding or delayed hit registration.
Close launchers, cloud sync, update tools and VPNs, then compare the game's server ping with the LinkSpeed Ping Test. Use an FPS counter as well: frame-time spikes can feel like network lag even when ping stays stable.
Cloud gaming adds input travel time, server processing, video encoding, bitrate variation, local decoding and display delay. Reduce stream quality only when it stabilises the bitrate; extra download speed alone will not repair jitter or packet loss.
Lag spikes
Lag spikes are usually a stability problem. Work through device load, server selection, household traffic and router queueing before assuming the broadband package is too slow.
Separate network lag from frame stutter.
On PC, use the game’s FPS counter or performance overlay. Smooth ping with frame drops points to PC performance, while rubber-banding, packet loss or sudden in-game ping jumps points to network stability.
Run live latency tests while the fault is happening.
Use the LinkSpeed Ping Test for idle latency, then use the Bufferbloat Test while other devices are active. If the loaded result jumps hard while idle ping is fine, router queueing is the likely cause.
Check night-time behaviour.
If ping only spikes in the evening, compare tests at different times. Evening-only issues often point to local contention, ISP routing, busy streaming hours, game patches or automated cloud backups.
Check the selected game server region.
A distant server can create 80–150 ms gameplay even on a strong UK line. Do not rely on auto-select if the game lets you choose a region manually.
Stop background traffic before changing provider.
Pause launchers, console updates, CCTV uploads, cloud sync and livestreaming. If gameplay immediately improves, prioritisation, SQM or upload management is the fix.
Beyond your home network
A fast fibre line cannot guarantee the lowest in-game ping. The selected region, the path between the ISP and game network, congestion on that path and the server's update model all affect responsiveness.
Border Gateway Protocol selects a policy-compliant route between networks; it does not guarantee the geographically shortest route. Direct peering between an ISP and a game platform can reduce handoffs, while an indirect transit path can add distance, congestion and inconsistent delay even when the selected game region looks correct.
A route can also change during an outage or capacity event. This is why one title may suddenly gain 30–60 ms while ordinary browsing and a nearby speed test remain normal.
The outbound packets shown by tracert or traceroute may follow a different network path from the returning game traffic. A clean outbound trace therefore cannot prove that the reverse route is equally direct. Return-path congestion, trans-oceanic detours or different transit providers can raise in-game ping without appearing in a one-way trace.
Use repeated traces, game-region comparisons and ISP evidence rather than treating one traceroute as a complete map of both directions.
Traditional fixed-tick servers process actions at discrete intervals. Sub-tick systems, including Counter-Strike 2's input model, can record when an action occurred between server updates instead of assigning it only to the next whole tick.
Sub-tick timing does not cancel network instability. Jitter, loss and packet reordering can still make updates arrive late or unevenly, forcing interpolation or correction and producing poor hit registration or rubber-banding even when the average ping looks low.
Test more than one region, compare quiet and peak times and check whether the problem affects one title only. If wired local tests stay stable but a specific route remains poor, the fault may sit beyond the router.
SQM can control local queueing and a better access point can stabilise Wi-Fi. Neither can make a distant server physically closer or repair congestion inside an external game network.
Route diagnosis
A route trace shows the responding hops on the outbound path. It is useful evidence, but it must be read carefully: routers often deprioritise or block diagnostic replies, and one trace cannot reveal the complete return path.
Replace the example host with a game-server IP or hostname only when the game publisher exposes one.
tracert -d linkspeed.co.ukThe numeric option avoids slow reverse-DNS lookups while you compare hop latency.
traceroute -n linkspeed.co.ukHop 1 is normally the local router. A consistently high first hop points to Wi-Fi, local Ethernet, router load or a powerline/mesh path. The next few hops often cover the provider access network, but hop numbering varies by technology and ISP.
A permanent latency increase that begins at one hop and continues through every later responding hop is more meaningful than one isolated slow reply. Network names may identify an ISP, exchange or transit operator, but do not assume that hops 4–7 always represent the same part of every route.
If one intermediate hop shows 150 ms or packet loss but the following hops and final destination return to normal, that router is probably rate-limiting diagnostic traffic rather than delaying forwarded game packets.
Best evidence: capture several traces during the fault and at a normal time, alongside the game's server region, wired ping, packet loss and loaded-latency results. Do not publish private WAN addresses, account identifiers or full router screenshots when asking for help.
Practical title-by-title targets
These are practical target ranges rather than hard pass/fail limits. A stable result near the upper end can play better than a lower average with jitter, packet loss or repeated spikes.
Modern network protocols
Native IPv6 gives devices globally routable addresses and can avoid the IPv4 NAT layer used in many homes. In some setups this simplifies peer connectivity and removes one source of NAT traversal, but the console or game platform still controls firewall and session behaviour.
The route chosen by the ISP, peering quality, server region and congestion normally matter more than the protocol label. Test both paths only when the platform supports them properly; use the route that is stable rather than assuming IPv6 is always faster.
Some providers share public IPv4 addresses between customers. Local port forwarding cannot bypass provider-side CGNAT. A public IPv4 option or correctly supported IPv6 may improve inbound and peer-to-peer connectivity, but neither guarantees lower match latency.
IPv6 does not mean exposing consoles or PCs directly. Keep stateful firewall protection active and use platform-supported automatic rules rather than opening broad inbound access.
Advanced Windows checks
Registry tweaks are not universal latency fixes. Most real-time games use UDP for gameplay, while Nagle-related settings apply to TCP. Test over Ethernet, update the network driver and record a baseline before changing Windows.
Create a restore point and export any registry key before editing it. Change one setting at a time, restart when required, repeat the same test and restore the original value when there is no measurable benefit or another application behaves differently.
For a specific active network-interface key under HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\Interfaces\, advanced users sometimes create DWORD values TcpAckFrequency=1 and TCPNoDelay=1.
These values affect TCP behaviour, not UDP gameplay traffic, and Windows applications can set their own socket options. Treat this as a controlled troubleshooting experiment for a known TCP-based workload, not a site-wide recommendation.
The value NetworkThrottlingIndex is located under HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Multimedia\SystemProfile. Setting it to hexadecimal ffffffff disables that throttling mechanism.
Record the original value first. Disabling it can change multimedia scheduling behaviour, so only keep the change when repeated tests show an improvement without harming streaming, audio or other workloads.
Disable third-party VPNs, traffic shapers and vendor “gaming” utilities temporarily; update the Ethernet or Wi-Fi driver; turn off background launchers and cloud sync; and compare the same game over Ethernet. These checks are easier to reverse and often expose the real cause.
Console network configuration
PPPoE, VPN tunnels and other encapsulation can reduce the usable path MTU. Fragmentation or black-hole MTU faults may interrupt sessions, but manually lowering a console MTU does not normally reduce ordinary ping.
On Windows, start with ping 1.1.1.1 -f -l 1472 and lower the payload until it succeeds, then add 28 bytes for the IPv4 and ICMP headers. Use the measured result as evidence before changing the PS5 setting. Values such as 1454 or 1473 are test points, not universal UK defaults.
UPnP lets a console request temporary mappings and is normally the simplest option on a trusted home network. It can fail when two routers are active, firmware is faulty or stale mappings remain.
For manual forwarding, first reserve a stable local IP for the console or PC, then forward only the ports documented by the platform or game publisher. UDP 3074 is commonly associated with Xbox and some games, but required ports vary; do not expose broad ranges without a specific need.
NAT status mainly affects inbound matchmaking and peer connectivity. A correctly opened port cannot shorten the physical route to a server or fix provider congestion, Wi-Fi interference or bufferbloat.
Do not combine broad DMZ exposure, UPnP and overlapping manual forwards for the same device. Use one controlled method, keep router firmware current and remove rules that are no longer required.
No easy Ethernet?
Ethernet is the cleanest gaming connection, but it is not always practical. These options are not equal to a direct cable, but they can be much better than weak Wi‑Fi through several walls.
Use a mesh node connected by Ethernet where possible, then cable the console or PC to that node. This avoids repeating everything over wireless backhaul.
Powerline can help where Wi‑Fi is poor, but results depend heavily on electrical wiring. Test latency stability, not only download speed, before trusting it for competitive games.
If your home has suitable unused coax cabling, MoCA can provide a wired-style route between rooms. Do not connect equipment into a provider network without the correct isolation and setup.
A short, clean 5 GHz or 6 GHz link can be good for gaming. Keep the router visible, avoid crowded channels and do not force a console to cling to weak Wi‑Fi at the edge of coverage.
Router settings
Quality of Service tells the router which devices or traffic types should be treated as important. It can help when a console or gaming PC competes with downloads, TVs, CCTV uploads or cloud backups.
A simple First-In, First-Out queue treats a large upload and tiny game packets as one line. When the broadband uplink is saturated, newly arriving game packets sit behind bulk data already stored in the buffer, so delay grows even though no packet has technically been lost.
FQ-CoDel separates traffic into flows and actively controls persistent queue delay, preventing one transfer from monopolising the buffer. CAKE builds on fair queueing with integrated shaping and practical handling for hosts, overhead and Diffserv classes.
Neither algorithm identifies “gaming” by magic. Their strength is structural fairness: small interactive flows do not have to wait behind an entire bulk transfer.
Run a wired speed test, then start with shaping limits around 90–95% of the measured download and upload rates. Retest with the LinkSpeed Bufferbloat Test and adjust gradually rather than copying another line's values.
SQM will not fix a distant region, external provider routing, weak Wi-Fi signal or a game-server fault. Use it when latency rises specifically while the connection is busy.
Live diagnosis
Run a continuous ping while the connection is idle, then repeat while uploading and downloading. A large increase under load points to queueing delay and gives you a before-and-after check when tuning SQM.
ping 1.1.1.1 -t
ping 1.1.1.1
Record the idle baseline.
Watch response times for 20–30 seconds while no large upload or download is running.
Load both directions.
Start a speed test or controlled upload and download. Note the typical rise and the worst repeated spikes rather than one isolated result.
Compare, tune and repeat.
Enable SQM or lower its shaping limits, then repeat the identical workload. The aim is a much smaller increase under load without cutting more throughput than necessary.
Stop the command.
Press Ctrl + C to finish and show the summary statistics.
Longer diagnostic captures
Use these commands to collect repeatable evidence for an ISP or game-support ticket. Test a stable public target and, where available, the actual game endpoint. Packet loss to an intermediate router is only meaningful when it continues to later hops or the final destination.
Writes a continuous timestamped log to the current user's Desktop. Stop it with Ctrl + C.
ping 1.1.1.1 -t | ForEach-Object { "{0:yyyy-MM-dd HH:mm:ss.fff} - {1}" -f (Get-Date), $_ } | Tee-Object -FilePath "$env:USERPROFILE\Desktop\gaming_ping_log.txt"A 0.2-second interval helps expose short jitter bursts. Some systems restrict rapid unprivileged pings; use the normal interval when the command is rejected.
ping -c 100 -i 0.2 1.1.1.1Combines route discovery with repeated probes. Allow several minutes for the statistics phase to finish.
pathping linkspeed.co.ukmtr may need to be installed first. The report mode sends 100 cycles and then exits.
mtr -rwzc 100 linkspeed.co.ukInterpretation rule: loss that appears at one hop and disappears later is normally reply-rate limiting. Investigate loss that begins at a hop and persists through subsequent hops and the destination, especially when it matches the time of the in-game fault.
DNS myths and useful settings
Changing DNS normally does not reduce the ping shown once a match is connected. DNS resolves hostnames before the connection starts; gameplay packets then travel directly to the selected server IP. A better resolver can still improve reliability, login lookups, launcher responsiveness and content downloads.
| Provider | Primary IPv4 | Secondary IPv4 | Useful for |
|---|---|---|---|
| Cloudflare | 1.1.1.1 | 1.0.0.1 | Fast general-purpose resolution and broad anycast coverage |
| Google Public DNS | 8.8.8.8 | 8.8.4.4 | High availability and global reach |
| Quad9 | 9.9.9.9 | 149.112.112.112 | Security filtering against known malicious domains |
Open Settings > Network & internet, select the active Ethernet or Wi‑Fi connection, edit DNS server assignment, choose Manual, enable IPv4 and enter the preferred and alternate addresses. Then run ipconfig /flushdns.
Go to Settings > Network > Settings > Set Up Internet Connection, open Advanced Settings for the active network, change DNS Settings to Manual and enter the primary and secondary addresses.
Open Settings > General > Network settings > Advanced settings > DNS settings, select Manual and enter both IPv4 addresses.
Open System Settings > Internet > Internet Settings, select the saved network, choose Change Settings, set DNS Settings to Manual and enter the two resolver addresses.
Do not mix providers. Use the primary and secondary addresses from the same service. If connectivity fails, return DNS to Automatic and check for typing errors or router-level DNS restrictions.
Test while fixing
Keep this page as the main gaming ping guide. When you change a router setting, move a mesh node or pause background traffic, run the LinkSpeed Ping Test and then use the Bufferbloat Test to check whether latency is still stable under load.
Common fixes
Evening spikes are usually tied to local ISP congestion, busy household uploads, cloud backups, game patches or streaming traffic. Compare wired idle and loaded-latency tests at quiet and peak times to separate local queueing from provider congestion.
Use Ethernet where practical, pause background uploads, select the nearest game region, disable unnecessary VPNs and enable Smart Queue Management if the router supports it. Retest idle ping, jitter, packet loss and loaded latency after each change.
A low average ping can hide packet loss, jitter bursts or loaded-latency spikes. Frame-rate drops can also feel like network lag, so compare the game's FPS counter with live ping and packet-loss results.
Cloud gaming is more sensitive because each input travels to a remote system and the rendered video must return to the player. Stable latency, low jitter, consistent bitrate and fast local decoding all matter.
Avoid running two routers in routing mode. Use bridge or modem mode on the ISP gateway where supported, or put the second router into access-point mode. If the upstream address is carrier-grade NAT, ask the provider about a public IPv4 address or supported IPv6 connectivity.