Innovabase › Lag and dropouts
Lag, stutter and dropouts on an Australian connection
The word "lag" gets used for four entirely different faults. Delay between pressing a key and seeing the result, movement that jerks rather than flows, characters that teleport backwards, and sessions that end without warning all have separate causes — and treating them as one problem is why people change settings for a week without improving anything.
Short answer
Steady delay is usually distance to the server and cannot be reduced. Delay that jumps around, movement that snaps backwards, and dropouts are usually packet loss or jitter on your own connection, and the first thing worth testing is a cable instead of Wi-Fi. If the problem appears only between about seven and eleven at night, it is congestion, and that is a conversation with your provider rather than a setting on your machine.
Four faults, one word
| What it looks like | What it is called | Usual cause |
|---|---|---|
| Everything responds late, but smoothly and consistently | Latency | Distance to the server, or a slow path to it |
| Response time swings between fine and terrible | Jitter | Congestion or interference on your own link |
| Characters jump back to where they were a second ago | Packet loss with correction | Dropped data between you and the server |
| The picture stutters while the connection seems fine | Frame rate, not network | The graphics processor or thermal throttling |
| Session ends and returns you to a login screen | Disconnection | A link that drops, or a session timing out |
The fourth row catches people out regularly. If the world stutters but other players continue to move normally and chat keeps arriving, the network is fine and the machine is struggling. That is a graphics settings problem, and lowering resolution or effects fixes it where no amount of network work will.
The part you cannot fix, and why
Signals in optical fibre travel at roughly two-thirds of the speed of light in a vacuum, and no equipment changes that. A round trip between an Australian home and a server in Europe or North America therefore has a floor measured in the high tens or low hundreds of milliseconds before a single router is considered. Add the routers — and there are many on a long path — and the practical floor rises further.
This has a useful consequence for troubleshooting. If your latency to a game server is high but stable, sitting near the same figure minute after minute, you have most likely found the floor rather than a fault. Upgrading a plan, changing a Wi-Fi channel or reinstalling the game will not shift it. What does shift it is the server you connect to, where a game offers a choice; some do, many do not.
Variation is the signal that something is wrong. A connection that reads well for thirty seconds and then badly for ten has a fault worth finding, even if its average looks acceptable. Averages hide exactly the behaviour that ruins a session.
Wi-Fi is the usual culprit
Home Wi-Fi is a shared radio medium in a building full of other radios. Microwave ovens, cordless phones, Bluetooth devices, the neighbours' networks and the walls themselves all degrade it, and the degradation shows up as exactly the pattern games are worst at tolerating: occasional lost packets and unpredictable delay rather than a clean slowdown.
A one-off test settles it. Connect the device to the router with an Ethernet cable, play for twenty minutes, and see whether the character still snaps backwards. Where a cable is impractical, the next best options in rough order are moving closer to the router, moving to the 5 GHz band if the device supports it, and adding a wired access point nearer the room you play in. Powerline adapters are a mixed proposition and depend heavily on the wiring in the building; they are worth trying only if they can be returned.
One detail that is easy to miss: a household where several devices stream video while someone plays is a household where the game is competing for the same upstream capacity. Testing with everything else switched off for ten minutes is a crude experiment but a conclusive one.
Evening congestion
Fixed-line broadband in Australia is sold with a distinction between the speed of the connection and the capacity the retail provider buys to carry it. When a provider buys too little for the number of customers using it, performance degrades during the hours when everyone is home. The pattern is unmistakable: fine at two in the afternoon, poor from dinner time onwards, fine again late at night.
If that describes your experience, the cause is not inside your house, and the evidence for that is worth collecting before you make the call. Run the same latency test at three separate times of day for several days and write the results down. A provider that is told "it is slow at night" has a support script for you; a provider that is shown a table of measurements from consistent times has a fault to investigate. The Australian Competition and Consumer Commission publishes material about broadband performance and what a service is expected to deliver, which is useful background if the conversation becomes difficult.
Packet loss and jitter
Games send small packets constantly and rely on nearly all of them arriving. Losing even a small fraction produces visible correction, because the game briefly predicted your position from stale information and then had to reconcile it with reality. That reconciliation is what the sudden snap backwards is.
Loss can occur anywhere along the path, but the parts you can influence are the first few: the device's own wireless adapter, the cable, the router, and the line into the property. A router that has been running for a year without a restart is a reasonable early suspect — memory leaks and full connection tables are real and they produce intermittent loss. Restart it, and if the fault returns on a predictable schedule afterwards, the router itself is worth replacing.
Ageing physical infrastructure produces the same symptom. Corroded connections, a damaged lead-in cable or water in a pit will show as loss that worsens in wet weather — a pattern worth noting on your record, because it is the sort of detail that moves a fault report forward.
Things on the machine itself
Not every cause is on the network. Background updates are the most common local offender: an operating system or store pulling several gigabytes saturates the connection and produces exactly the symptoms above. Checking for active downloads takes seconds.
Power management is the other frequent one. Laptops running on battery reduce processor and graphics clocks aggressively, and the result is stutter that disappears the moment the machine is plugged in. Thermal throttling does the same thing after twenty minutes of play rather than immediately, which makes it easy to misattribute to the server filling up at peak time.
Measuring instead of guessing
- Establish a baseline when things are good. Run a continuous ping to a stable address for a few minutes on a quiet afternoon and note the typical figure and the spread. Without a baseline, later numbers mean nothing.
- Repeat the same test while the fault is happening. Same target, same duration. Compare the spread as well as the average; a rising spread points to jitter, a rising average to a longer path.
- Test wired and wireless back to back. Twenty minutes each, in the same session if possible. This single comparison resolves a large share of cases.
- Test on mobile data. Tether the device to a phone. If the fault disappears entirely, it is in your fixed connection, not the game.
- Test a second device on the same network. If one machine is affected and another is not, the problem is that machine rather than the line.
- Record times and dates. Three or four data points across different times of day turn an impression into evidence.
Do accelerators and VPNs help?
Products marketed as game accelerators route your traffic over a private network in the hope of a better path than the public internet provides. Occasionally that works, where the default route is genuinely inefficient. More often it adds a detour, because the direct path was already reasonable.
What such a product cannot do is beat physics. No service reduces the time light takes to cross the Pacific. Treat any claim that a product will dramatically reduce ping to a distant server with the scepticism it deserves, and if you try one, test it against a measured baseline rather than an impression. A free trial with a measurement either side of it is the only honest way to evaluate one.
What to watch out for
Downloadable "network optimisers" and "ping boosters" that require deep system access are a category worth avoiding entirely. Adjusting settings a game exposes is fine; installing an unknown program with privileged access to fix a problem you have not yet measured is not. General guidance on assessing software before installing it is published by the Australian Cyber Security Centre.
Taking it to your provider
When the evidence points outside your home, the report that gets action contains measurements, times, and what you have already eliminated. Say that you have tested wired and wireless, tested a second device, tested on mobile data, and restarted the router — and give the figures. Ask for the fault to be logged, and ask for a reference number, because a logged fault with a history behind it is treated differently from a first call.
If the service is not delivering what was sold and the provider will not resolve it, complaints about telecommunications services in Australia are handled by the Telecommunications Industry Ombudsman, and the ACCC's consumer material at accc.gov.au explains the consumer guarantees that apply to services as well as goods.
My ping is fine but the game still feels delayed. What else could it be?
Display latency and frame rate both add delay that no network test will show. A television in its default picture mode can add a noticeable amount; most have a game or low-latency mode that removes much of it. A frame rate below the display's refresh rate adds its own delay on top.
Would a faster plan help?
Rarely, for games specifically. Games use very little bandwidth; what they need is consistency. More capacity helps only if the connection is genuinely saturated by other activity in the household, in which case the real fix may be scheduling large downloads rather than paying more each month.
The game is fine on my phone's mobile data but poor on home broadband. What does that prove?
That the game, its servers and your device are all working. The fault is in the fixed connection or the equipment attached to it. That is a strong piece of evidence to lead with when you contact your provider.
Does the game's own server choice matter?
Where a game lets you pick a region, it matters more than anything on your side, because it changes the physical distance the data travels. Whether a particular game offers that choice is something to check on the vendor's own site rather than assume.