Why Your Ping Is High and Games Lag

Rubber-banding, hit registration problems and sudden spikes are usually not bandwidth. Here is what actually causes them, how to measure which one you have, and what genuinely helps.

You have a fast connection. The speed test says so. And you are still rubber-banding across the map, losing fights you clearly won on your screen, and watching the ping counter spike into the hundreds for no visible reason.

Bandwidth is almost never the problem. A competitive shooter uses a few hundred kilobits per second — less than a music stream. What games need is consistency: small packets arriving reliably and on time, every time. That is a completely different property from throughput, and it is the one nothing in a speed test measures.

### Measure what games actually care about The connection test reports packet loss, jitter and latency separately, in both directions. Those three numbers explain nearly every lag complaint. Run the connection test →

The four things that cause lag

Different symptoms come from different faults. Matching yours to the right one saves a lot of pointless setting changes.

1. Distance — the one you cannot fix

Latency has a floor set by physics. Light in fibre covers roughly 200 km per millisecond, and real routes are never straight lines. A round trip from London to Virginia will not beat about 75 ms no matter what you buy.

If your ping to a nearby server is 20 ms and to a distant one 180 ms, nothing is broken. Choosing a closer server is the only fix, and it is the reason server selection matters more than any hardware upgrade.

What does not help: more bandwidth, a "gaming" router, or a gaming VPN — with one narrow exception covered below.

2. Packet loss — the rubber-banding one

Games send position updates continuously over UDP, without retransmission, because a movement update that arrives late is worthless. When packets go missing, the client and server disagree about where things are. The server corrects you, and you snap backwards.

That correction is exactly what rubber-banding is. Teleporting enemies, shots that clearly connected registering as misses, and being killed after reaching cover are all the same underlying problem seen from different angles.

Games tolerate a little loss and interpolate through it. Past roughly 1% it becomes noticeable; past 2–3% competitive play stops being viable. Bursty loss is far worse than the same percentage spread evenly — forty consecutive missing updates is a visible teleport, while the same total scattered thinly is invisible.

Diagnosing it is the subject of what causes packet loss, but the short version for gamers: test wired before anything else.

3. Jitter — the one nobody talks about

Jitter is variation in latency between consecutive packets, and it explains the single most frustrating experience in online gaming: a connection that looks fine on paper and feels wrong to play on.

Stable connection          average 30 ms
  30  31  29  30  31  30  29  ms

Jittery connection         average 30 ms
   8  55  12  70  15  48  11  ms

Both average 30 ms. The first is predictable and your aim works. The second makes hit registration feel arbitrary, because the server is receiving your inputs at wildly varying delays and reconciling them against other players whose timing is steady.

This is why "my ping is fine but the game feels off" is a real observation rather than imagination. Ping counters show an average and hide the variation entirely.

Under 5 ms of jitter is good, 5–15 ms is playable, above 30 ms is disruptive regardless of what the average says.

4. Bufferbloat — the "someone started a download" one

If your game is fine until a housemate opens Netflix or a console starts updating, this is almost certainly it.

Oversized buffers in consumer routers queue packets rather than dropping them when the link is busy, so latency climbs from 20 ms to 300 ms and stays there until the transfer finishes. Your bandwidth is not exhausted — your queue is.

It is measurable in a minute and genuinely fixable at your end, which makes it the highest-value thing on this page. The bufferbloat guide covers testing and the fix in detail.

Measuring yours

Three tests, about ten minutes total.

1. Baseline. Run the connection test over Ethernet with nothing else running. Note packet loss, jitter and latency. This is your connection at its best.

2. Under load. Start a large upload and run it again. If latency multiplies, you have bufferbloat.

3. Wi-Fi comparison. Run it once more over Wi-Fi. If loss or jitter appears that was not there on the cable, wireless is your problem.

Then check the path to something in the region your game servers live:

Windows          tracert 1.1.1.1
macOS / Linux    mtr 1.1.1.1

Or use the MTR tool, which probes continuously and catches intermittent faults a single traceroute misses. Look for loss that begins at a hop and continues through every hop after it — loss at one hop that clears afterwards is just that router deprioritising pings, not a fault.

What actually helps

In descending order of effect, which is roughly the reverse of what gets recommended in forums.

Use Ethernet. Not advice anyone wants, and still the single largest improvement available to most players. Wi-Fi adds jitter and loss even when signal strength looks perfect, because the medium is shared and retransmission happens invisibly below the application. If a cable is genuinely impossible, use 5 GHz, sit close to the access point, and accept a ceiling on how good it can get.

Fix bufferbloat. Enable SQM on your router if it supports it. This is the fix that most often turns an unplayable evening connection into a fine one, and it costs nothing.

Pick the closest server. Manually, where the game allows it. Automatic selection frequently optimises for population rather than latency.

Stop background traffic. Cloud backups, game updates downloading in the background, another device streaming 4K. Pausing these during play is crude but effective.

Check for interference. If you must use Wi-Fi, move to a clear channel on 5 GHz. Neighbouring networks on 2.4 GHz are a common cause of periodic spikes that correlate with nothing you are doing.

What does not help

Being direct about this, because these are heavily marketed.

More bandwidth. Going from 100 Mbps to 1 Gbps changes nothing about latency, loss or jitter. A game needs a fraction of either. This is the most commonly sold non-solution in the category.

"Gaming" routers. The label is mostly marketing. What matters is whether the firmware supports proper queue management — a cheap router running OpenWrt with fq_codel will outperform an expensive one without it. Buy for SQM support, not for styling.

Gaming VPNs, usually. A VPN adds a hop and normally increases latency. The narrow exception is when your ISP routes badly to a particular destination and the VPN happens to take a better path — real, but uncommon, and worth testing before paying for it. Measure with and without rather than assuming.

Most "network optimiser" software. Registry tweaks and TCP tuning utilities have negligible effect on modern systems, and games do not use TCP for gameplay traffic anyway.

Closing background applications, beyond ones actually using the network. A chat client uses no meaningful bandwidth. The problem is transfers, not process count.

When it is not your connection

Some things look like network problems and are not.

Server-side issues. If everyone in the match is complaining simultaneously, the game server is the problem. Nothing at your end will help.

Frame rate confusion. Stuttering caused by GPU or CPU limits feels similar to network lag but is unrelated. If your frame rate drops at the same moment things stutter, look at the game's settings rather than your router.

The return path. Routing is asymmetric, and your outbound path can be clean while the route back is congested. Your own traceroute cannot see that direction at all — testing the return path is the only way to check, and it explains a proportion of "my connection is fine but this one server is unplayable" cases.

ISP peering. If one particular game is bad while everything else is fine, the problem may be at the interconnect between your ISP and that game's hosting provider. An MTR to the game server during a bad period, compared with a good one, will usually show where. If it is consistent, it is worth reporting with evidence.

Frequently asked questions

Why is my ping so high when my internet is fast?

Speed and latency are unrelated measurements. Bandwidth is how much data moves at once; ping is how long a round trip takes, and it is governed by distance, routing and queueing. A gigabit connection to a server on another continent will still have high ping.

What is a good ping for gaming?

Under 30 ms is excellent, 30–60 ms is good and competitive play is comfortable, 60–100 ms is playable for most genres, and above 100 ms puts you at a real disadvantage in anything reaction-based. Consistency matters as much as the number — steady 60 ms beats an average of 40 ms that swings between 10 and 90.

Why do I rubber-band even with low ping?

Rubber-banding is caused by packet loss, not latency. Position updates go missing, the server corrects your position, and you snap backwards. Low average ping with rubber-banding almost always means loss or severe jitter rather than distance.

Does a wired connection really reduce ping?

It reduces latency slightly, but the larger benefit is consistency. Wi-Fi introduces jitter and intermittent loss even with a strong signal, because the medium is shared and retransmissions happen below the application. The improvement in how the game feels is usually much bigger than the change in the ping number.

Will a gaming VPN reduce my ping?

Usually it increases it, since traffic takes a longer path through the provider. The exception is when your ISP routes poorly to a specific destination and the VPN's path is better. It is worth measuring with and without before subscribing — treat any guaranteed improvement as marketing.

Why does my ping spike when someone else uses the internet?

Bufferbloat. Oversized buffers queue packets when the link saturates, so latency climbs sharply the moment anyone starts a large transfer. Enabling Smart Queue Management on your router usually fixes it — see the bufferbloat guide.

Can packet loss be caused by my own PC?

Yes. A failing network adapter, an outdated driver, a damaged cable or aggressive power management on the adapter can all cause loss before traffic ever leaves your machine. Checking interface error counters — netstat -e on Windows — will show whether errors are accumulating locally.

Is 1% packet loss bad for gaming?

It is at the threshold where most people begin noticing. Whether it matters depends heavily on distribution: 1% spread evenly is often imperceptible, while 1% arriving in bursts produces visible teleporting. Anything sustained above 2% will affect competitive play.