NBN Co fibre optic cable

What Is Latency and Why Does It Matter for Gaming and Video Calls?

Most people judge an internet connection by one number: speed. The MTN or Airtel bundle advertises megabits, the speed test app reports megabits, and the assumption follows naturally that a bigger number means a better experience.

Then you join a Zoom call on a connection testing at 40 Mbps and everybody talks over each other because your voice arrives a beat late. Or you play an online match and lose every exchange despite reacting first. Or a livestream stutters on a line that downloads a large file perfectly well. In each case the speed number is fine and the experience is not, which tells you the number was measuring the wrong thing.

The thing it was not measuring is latency: the delay between sending a request and getting the first response back. Bandwidth is how much data can move at once. Latency is how long it takes to start moving. For anything interactive — a call, a game, a remote desktop, a trading app — latency is what you actually feel, and it is the number worth paying attention to. This guide explains what causes it, what counts as good, why Nigerian connections have the particular latency profile they do, and what genuinely reduces it.

Bandwidth and latency are different things

A useful way to think about it: bandwidth is how many lanes the road has, latency is how long the journey takes.

A ten-lane motorway that runs the long way round a city can carry enormous traffic and still get you there late. A single-lane shortcut gets one car there quickly but jams the moment a second car appears. Neither one is simply “better” — they solve different problems.

Downloading a large file is a bandwidth problem. Once the transfer begins, the initial delay is irrelevant; what matters is how much data flows per second. A video call is a latency problem. The amount of data is small, but every packet must arrive promptly or the conversation breaks down.

This is why a connection can score well on a speed test and still feel unusable for calls, and why upgrading to a bigger data bundle often does not fix the problem people bought it to fix.

How latency is measured

Latency is measured in milliseconds, and usually reported as ping or round-trip time — the time for a small packet to travel to a server and back.

Rough guide to what the numbers mean in practice:

  • Under 20 ms — excellent. Feels instant for anything.
  • 20 to 50 ms — good. Comfortable for competitive gaming and calls.
  • 50 to 100 ms — acceptable. Fine for calls, noticeable in fast-reaction games.
  • 100 to 200 ms — poor for interactive use. Conversations develop that awkward talk-over rhythm.
  • Over 200 ms — bad. Calls become difficult, real-time gaming is largely lost.

There are two related numbers that matter as much as the average, and often more.

Jitter is the variation in latency. A connection steady at 90 ms is far more usable than one bouncing between 30 ms and 250 ms, even though the second might average lower. Voice and video encoders can adapt to consistent delay; they cope badly with unpredictable delay. Most call quality complaints are actually jitter complaints.

Packet loss is data that never arrives. Even small amounts — one or two per cent — cause audible dropouts, because the missing pieces must either be re-sent, which adds delay, or skipped, which removes sound.

What actually causes latency

Latency accumulates at several points between you and the server. Understanding which one dominates tells you what is worth fixing.

Physical distance. Signals cannot exceed the speed of light, and in fibre they travel somewhat slower. Lagos to a server in Lagos might be a few milliseconds. Lagos to Europe is roughly 70 to 120 ms on a good path. Lagos to the west coast of the United States can exceed 200 ms. No amount of bandwidth changes this — it is geography.

Routing. Traffic rarely takes the straight line. A request from Lagos to a Nigerian service can still be routed through Europe if the networks involved do not exchange traffic locally. This is one of the largest and least visible causes of high latency in Nigeria, and it is why local peering and local hosting matter so much here.

The last mile. The connection from your building to the provider. Fibre adds little. Fixed wireless adds more. Mobile data adds a variable amount depending on signal strength and how congested the cell tower is.

Congestion. Every link along the way has finite capacity. When it fills, packets queue, and queuing is delay. This is why latency worsens in the evening even though nothing about your equipment has changed.

Your own Wi-Fi. Frequently the single largest contributor and the one most people never check. A weak or congested Wi-Fi link can add tens of milliseconds and a great deal of jitter before your traffic even reaches the router.

The device and the server. Processing time at both ends adds a few milliseconds, and an overloaded phone or an overloaded game server adds considerably more.

Why Nigerian connections have the latency they do

A few structural factors shape the experience here specifically.

Most services you use are hosted abroad. Social platforms, streaming services, game servers and cloud tools are largely hosted in Europe or North America, with comparatively little hosted regionally. The physical distance sets a floor on latency that no local upgrade can beat. If the server is in Frankfurt, you are not getting 20 ms from Ibadan.

Local peering has improved but is uneven. When Nigerian networks exchange traffic locally rather than routing it through Europe and back, latency drops dramatically. This has got better over the years, but whether any given path stays local depends on the networks involved.

Mobile data dominates. A very large share of Nigerian internet use is over MTN, Airtel, Glo and 9mobile rather than fixed fibre. Mobile adds latency and, more importantly, adds jitter — signal strength, tower congestion and handover between cells all vary minute to minute.

Power interruptions. Routers and ONTs that restart repeatedly re-establish connections, and a link that has just come back up frequently performs worse for a while.

Content delivery networks help more than people realise. Large services cache content closer to users. This is why a YouTube video can start quickly while a video call to the same country feels laggy — the video is served from a nearby cache, the call must reach the other person in real time.

How to reduce latency

Working from largest typical gain to smallest.

Use a cable if the activity matters. Ethernet removes Wi-Fi’s variability entirely. For a gaming console, a desktop, or a computer used for important calls, a cable is the single most effective change available. It costs very little and eliminates a whole category of problems.

Fix your Wi-Fi before blaming your ISP. Move closer to the router, or move the router somewhere central and unobstructed rather than in a corner behind a cupboard. Use the 5 GHz band for devices that are near the router — it is faster and far less congested than 2.4 GHz, though it does not travel as far or through walls as well. Wi-Fi placement fixes more “slow internet” complaints than any bundle upgrade.

Reduce the number of active devices. Every device streaming, backing up or updating in the background competes for the same link. A phone silently uploading photos to cloud storage during your call is a real cause of jitter.

Choose servers close to you. In games that let you pick a region, choose the nearest one. For video calls, the platform usually decides, but some let you influence it.

Enable QoS on the router if it has it. Quality of Service prioritises latency-sensitive traffic over bulk downloads. On a shared home connection this can transform call quality without changing your bandwidth at all.

Test on mobile data as a comparison. If your fixed connection is unusable and a phone hotspot on the same activity is fine, the problem is your line or your router, not the internet in general. That single test saves a lot of guessing.

Be realistic about a VPN. A VPN almost always increases latency, because your traffic takes a longer path. The occasional exception is when your provider’s default routing is unusually bad and the VPN happens to take a better route. Test rather than assume.

Restart the router occasionally. Unglamorous, and it genuinely resolves accumulated connection-state problems.

What you cannot fix

Some honesty is useful here.

You cannot beat the speed of light. If the server is on another continent, there is a floor below which you will not go, and no bundle, router or setting changes it.

You cannot fix the provider’s routing from your end. If traffic to a service leaves Nigeria and comes back, that is a decision made in network agreements far above your account.

You cannot make a congested cell tower less congested at seven in the evening.

For anything where latency is critical and the audience is Nigerian — a business tool, a game server, a trading application — the real fix is hosting closer to users rather than optimising the connection at the user’s end.

The short version

Speed tests measure how much data can flow. Latency measures how quickly it starts. For calls, gaming, remote work and anything else interactive, the second number is the one you feel.

Under 50 ms is comfortable, over 150 ms is painful, and consistency matters more than the average — a steady 90 ms beats an erratic connection that occasionally hits 30 ms.

If your calls are choppy, check your Wi-Fi and your background devices before upgrading anything. Plug in a cable where it matters. Accept that distance to the server sets a floor. And remember that the bundle advertising the biggest megabit number is answering a question you may not be asking.

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *