1% lows explained: the FPS stat that actually predicts stutter
Two builds both average 120 fps. One feels glassy-smooth; the other stutters every few seconds. The average can't tell them apart. The 1% low can.
Average FPS is the number everyone quotes, and it's the most misleading. It smooths over the brief, ugly moments — the hitches when you turn a corner, open the map, or a new area loads — that are exactly what make a game feel bad. Those moments hide inside a healthy-looking average, which is why two PCs with identical headline numbers can deliver completely different experiences.
What is a 1% low?
Record every frame in a benchmark run, sort them from slowest to fastest, and look at the worst 1%. That bottom slice is your 1% low — effectively the floor your experience drops to during the rough patches, your stutter floor. (Tools differ slightly on the exact method: some average the slowest 1% of frames, others take the frame at the 1%-of-total-time mark, per CapFrameX's method. The takeaway is the same either way.)
You'll also see a 0.1% low in detailed benchmarks. That zooms in further, to the worst 0.1% of frames — the rare, deepest stalls. It's noisier from run to run, but it exposes the single ugliest hitches that even the 1% low can average away.
Why does a high average still feel choppy?
Because your eyes don't see an average. They see each frame as it arrives. What actually matters is frame time: how many milliseconds pass between one frame and the next. A steady 120 fps means a new frame every 8.3 ms, like a metronome. But if a single frame takes 50 ms to arrive while the rest take 8, that one long gap reads as a visible hitch, even though it barely moves the per-second average.
Stack up a few of those long frames per minute and the run still averages 115–120 fps, but it feels like it's snagging. That's the disconnect. FPS counts how many frames you got in a second; the 1% low tells you how bad the worst ones were. Smoothness lives in the second number.
From our calculator · real numbers
Take one build — a Ryzen 5 7600 with an RTX 4070 at 1440p — and our FPS Calculator estimates two very different stories depending on the game:
Same PC, same settings. The esports title posts a much higher average but a far wider gap, because it leans hard on the CPU. The headline FPS would never tell you that.
Why does the CPU usually decide your 1% lows?
1% lows are often a CPU story. When a new area streams in, an explosion spawns dozens of physics objects, or the game compiles a shader on the fly, it's the processor scrambling to prepare the next frame in time. Miss that deadline and you get a long frame — a dip in the 1% low — even if your GPU is barely breaking a sweat. It's no accident the swingy example above was the CPU-heavy esports title.
This is also why memory matters more than people expect. The CPU leans on fast RAM to feed those bursts, so slow or untuned memory shows up as frame-time spikes rather than a lower average. A stronger gaming CPU paired with quick memory tightens the dips, which is how a part swap can make a game feel smoother even when the headline FPS barely moves. If you're weighing a CPU, our CPU tier list ranks them on gaming strength, not raw core counts.
1% low vs average FPS: which should you trust?
Use both, but weight the 1% low for how a game will actually feel. The average tells you whether your hardware is broadly fast enough for a target frame rate; the 1% low tells you whether it'll hold that frame rate when things get hectic. A steady 50 fps with a 45 fps 1% low almost always feels better than a 60 fps average that keeps collapsing to 25.
The simplest read is the gap itself: the closer your 1% low sits to your average, the more consistent it feels. Many builders treat a 1% low above roughly two-thirds of the average as comfortable, and anything that routinely falls below that as stutter you'll notice — no matter how good the headline number looks.
How do you measure your own 1% lows?
An average FPS counter won't show them — you need a tool that records every frame's timing:
- CapFrameX. Free, built on Intel's PresentMon capture backend; records average, 1%, and 0.1% lows and draws the frame-time graph for a session (setup guide).
- MSI Afterburner + RivaTuner (RTSS). The classic overlay; its built-in benchmark logs highest, average, 1%, and 0.1% lows to a file you can read back.
- In-game benchmarks. Many titles (Cyberpunk 2077, Shadow of the Tomb Raider, and others) report 1% lows in their own benchmark runs — the quickest way to get a number with no extra software.
How do you improve your 1% lows?
- Turn on your memory's XMP/EXPO profile. One of the biggest free gains for frame-time consistency — reviewers have measured 1% low improvements of around 15% over default JEDEC speeds in CPU-bound games (TechSpot).
- Close background apps that steal CPU time. Browsers, overlays, and chat apps compete for the same cores that feed your frames.
- Stay inside your VRAM budget. Running out forces the game to swap textures over the PCIe bus mid-scene, which causes brutal, repeating hitches.
- Cap your frame rate slightly below your peak. A steady ceiling evens out delivery and stops the big swings that hurt the 1% low.
If a part swap is on the table, our guide to fixing a bottleneck walks through which upgrade actually moves the needle. And our FPS Calculator shows the 1% low right next to the average, with the math in our methodology — the headline number alone never tells the whole story.
Frequently asked questions
What is a good 1% low?
A 1% low that stays close to your average — roughly two-thirds of it or higher — feels consistent. The smaller the gap between average and 1% low, the smoother the game; a wide gap is what you feel as stutter.
Is a 60 fps average fine if my 1% low is 25 fps?
Not really. That 35 fps gap is exactly what you feel as stutter. A steady 50 fps with a 45 fps 1% low usually feels smoother than a 60 fps average that keeps dropping to 25.
How do I check my 1% low fps?
Use a frame-time capture tool: CapFrameX (built on Intel PresentMon) or MSI Afterburner with RivaTuner both record average, 1%, and 0.1% lows from a gameplay session. Many built-in game benchmarks report 1% lows too.
Does RAM speed really affect 1% lows?
Yes, often more than it affects the average. Enabling your memory's XMP/EXPO profile tightens frame-time spikes; reviewers have measured 1% low gains of roughly 15% over default JEDEC speeds in CPU-bound games.
What's the difference between a 1% low and a 0.1% low?
The 1% low reflects the slowest 1% of frames; the 0.1% low zooms into the worst 0.1% — the rare, deepest hitches. The 0.1% low is noisier but exposes the single ugliest stalls.
Curious what your build's stutter floor looks like?