The speed test that measures what games actually feel
Speed test says 500 Mbps and “excellent”. You still die to lag. Both are true — because a capacity test answers “how much can this pipe move?” and your game is asking a completely different question: “does every packet arrive, on time, every time?” LagScope is the speed test built for the second question.
Speedtest-style tests vs LagScope
| Ookla-style speed test | LagScope | |
|---|---|---|
| What it measures | Capacity — max Mbps over TCP | Consistency — UDP-like game traffic at 30 Hz |
| Test duration | A few seconds, best-case burst | Continuous — as long as you play |
| Server | Nearest, capacity-optimized | Real path to our game-side nodes (US / EU / BR) |
| Spikes | Averaged away — never shown | Counted, timed and logged one by one |
| Jitter | Not reported | Measured live, in ms, every second |
| Packet loss | Hidden — TCP silently retransmits | True UDP loss, like your game sees it |
| Bufferbloat | Not tested | 8-second stress test, latency under load |
| Result | A Mbps number and a grade | Per-game verdicts + timestamped ISP evidence |
What speed tests measure — and why it’s the wrong thing for games
A typical speed test opens several TCP connections to a nearby server chosen for raw throughput, transfers as much data as it can for a few seconds, and reports the peak or average. The “ping” it shows is a handful of samples to that same optimized server, averaged together. It’s a good answer to “how fast are my downloads?” — and almost no answer at all to “why did I just die behind cover?”
Games don’t need much bandwidth; they need consistency. A competitive game sends tiny UDP packets 20 to 60 times a second — your position, your inputs, your shots — and every one of them needs to arrive quickly and at a steady interval. The four things that break that are exactly the four things a capacity test can’t see.
The 4 things that actually kill games
- Lagspikes — your latency suddenly jumping from 25 ms to 300–800 ms for a second or two; that’s the freeze where you peek and die before the screen updates.
- Jitter — your ping on a rollercoaster (25, 90, 15, 60 ms); the game can’t smooth out arrival times it can’t predict, so players teleport and shots desync.
- Packet loss — bullets that never arrived: some packets simply never reach the server, so shots don’t register and movement snaps back. Test yours in 25 seconds.
- Bufferbloat — your router making game packets queue behind someone’s Netflix stream or a Windows update, so ping explodes exactly when the network gets busy. Here’s the fix.
A speed test averages the first two away, hides the third inside TCP retransmits, and never loads the line long enough to reveal the fourth. That’s the whole reason “speed test fine, still lagging” is the most common network complaint in gaming — we wrote the deep-dive on it.
When you still want a capacity test
Honestly: keep one around. If your 4K stream buffers, a download crawls, or you suspect your ISP isn’t delivering the plan you pay for, a capacity test is the right tool — that’s what it measures, and it measures it well. Just don’t let its “excellent” grade talk you out of a lag problem it was never designed to see. Different question, different instrument.
Frequently asked questions
Is there a better speed test for gaming?
For gaming you want a stability test, not a capacity test. LagScope probes your line the way a game uses it — small unreliable UDP packets at 30 Hz — and counts lagspikes, jitter, true packet loss and bufferbloat, with per-game verdicts (“fine for Fortnite — risky for Valorant ranked”) instead of a Mbps number.
Why is my ping high when my speed test is good?
The speed test measured to a nearby, optimized server and averaged a few samples. Your game server is farther away, and the average hides the spikes and loss happening on the way. A line can read 20 ms and “excellent” while spiking to 300 ms every few seconds in-game.
What is bufferbloat?
Your router making game packets stand in line behind bulk traffic — a Netflix stream, a Windows update. Nothing gets lost, everything just arrives late, so ping explodes when the network gets busy. A capacity test never loads the line long enough to see it; LagScope’s 8-second stress test does, and SQM/Cake fixes it.
Does Mbps matter for gaming?
Barely. Games use a few megabits per second — even 25 Mbps is plenty. What matters is stable latency, low jitter, zero loss and latency that stays flat under load. A stable 25 Mbps line beats a spiky gigabit line every match.