My average FPS looks fine but the game still stutters — why?

Your 1% low is probably far below the average. The worst 1% of frame rates is what reviewers report alongside the average, because frame time spikes are what the eye actually sees.

The FPS number never moves above 60 no matter what I change — is my GPU dead?

Check VSync and any in-game frame cap first. VSync caps the frame rate at the display refresh rate and raises input lag, so a flat 60 on a 60 Hz panel is a setting, not a fault.

Do I need 144 FPS if my monitor is 60 Hz?

No: your panel can only display 60 images per second, so anything above that is wasted unless you plan to upgrade. Aim for a steady 60 with a 1% low near 45.

Does a high FPS number guarantee responsive online play?

No: lag is a separate axis measured in milliseconds. Below 100 ms is needed for playability and demanding play aims below 20 ms, regardless of your frame rate.

16.7 msframe time at 60 FPS

20 mstarget lag for demanding online play

1% lowthe 99th-percentile floor reviewers report

133 msphysical ping floor

Five-step FPS check

GA9Z — Gaming Performance Check | ga9z.net

A game that stutters on your PC may still show a high average FPS. This guide walks you through an ordered check that measures average FPS, the 1% low and frame time, then reads the result against your own monitor's refresh rate.

ga9z.net
GA9Z · Performance check ledger · FPS, frame time, lag
Measure FPS, frame time and input lag honestly on your own PC.

Ordered procedure

The five-step FPS check

133 msphysical ping floorThe lowest physically possible ping for a connection crossing the whole planet is 133 ms.

50 FPSsteady-light thresholdA modulated light source reads as steady to most observers above 50 FPS.

10–12images per second in human visionHuman vision processes 10 to 12 separate images per second.

Refresh rate against FPS target

Monitor refresh rate against a realistic FPS target
Refresh rate	FPS target	1% low to aim for	What changes in feel
60 Hz	60 FPS	about 45 FPS	Smooth baseline; tearing visible without VRR
75 Hz	75 FPS	about 56 FPS	Slightly lighter motion than 60 Hz
120 Hz	120 FPS	about 90 FPS	Noticeably clearer tracking in fast scenes
144 Hz	144 FPS	about 100 FPS	Standard for competitive play on mid-range GPUs
240 Hz	240 FPS	about 180 FPS	Only recent hardware holds the average here
360 Hz	360 FPS	about 270 FPS	Marketed to competitive players; 1% low matters most
Any VRR range	inside range	inside range	Tearing removed; sync follows the game's frame rate

Monitor refresh rate against a realistic FPS target

How to measure FPS in five steps

Run an ordered check: fix the scene, fix the settings, run 60 seconds, log three figures, repeat after any change.

Pick a repeatable scene — a fixed camera position or a scripted benchmark loop — so that run two measures the same load as run one. Change nothing else between runs except the variable you are testing.

Fix resolution and the quality preset before you start, and write them down next to each run. A preset swap mid-test makes the numbers incomparable.

Let the run last 60 seconds so that shader warm-up and background loading settle out of the average. Shorter runs swing on scene content rather than on your hardware.

Log average FPS, the 1% low and frame time from the same run. All three come from one pass; mixing figures from different passes hides the problem you are looking for.

VSync left on caps the frame rate at the display refresh rate and raises input lag, flattening the average.
An in-game frame cap pins the frame rate below your GPU's real ceiling, so the number never moves.
A thermal throttle drops clocks after a few minutes; a 60-second run may end before it bites, so watch clock speed too.
A recorder running in the background steals frames and adds stutter that your hardware would not otherwise show.

Terms behind the numbers

Frame time: The gap between frames in milliseconds. 16.7 ms equals 60 FPS, 11.1 ms equals 90 FPS, 6.9 ms equals 144 FPS and 4.2 ms equals 240 FPS.
1% low: The 99th percentile of frame rates during a run: the slowest one percent of frames expressed as FPS. A high average with a low 1% figure means visible stutter.
Refresh rate: How many times per second a display updates, in hertz. It sets the ceiling that VSync can reach and the range a VRR panel can track.
VRR: Variable refresh rate: the display follows the game's frame rate inside a stated range, for example 30 Hz through 144 Hz, and removes tearing there.
Input lag: The delay between an input and the visible response, in milliseconds. Below 100 ms is needed for playability; demanding online play aims below 20 ms.
Thermal throttle: A clock-speed drop applied when a component gets too hot. It shows up as a frame rate that decays over minutes rather than a constant number.

Questions readers actually asked

My average FPS looks fine but the game still stutters — why?: Your 1% low is probably far below the average. The worst 1% of frame rates is what reviewers report alongside the average, because frame time spikes are what the eye actually sees.
The FPS number never moves above 60 no matter what I change — is my GPU dead?: Check VSync and any in-game frame cap first. VSync caps the frame rate at the display refresh rate and raises input lag, so a flat 60 on a 60 Hz panel is a setting, not a fault.
Do I need 144 FPS if my monitor is 60 Hz?: No: your panel can only display 60 images per second, so anything above that is wasted unless you plan to upgrade. Aim for a steady 60 with a 1% low near 45.
Does a high FPS number guarantee responsive online play?: No: lag is a separate axis measured in milliseconds. Below 100 ms is needed for playability and demanding play aims below 20 ms, regardless of your frame rate.

How this record was assembled

Every figure on this page comes from published performance and display standards: refresh rates, frame-time conversions, lag thresholds and VRR release dates. No measurements were run for this page; the procedure describes how a reader runs their own.

Reference figures used here

Figures cited here follow published display and performance standards; no independent testing was performed for this page.