Here’s a question I get a lot from people building or upgrading a PC.
Their benchmark says 120 FPS, so why does the game still stutter in a firefight?
The answer is a number almost nobody looks at. Your average frame rate can read great while the game plays like garbage in the exact moments you care about.
Think of it like a road trip.
You can average 60 mph on a long drive and still spend twenty minutes stuck dead in traffic. The average doesn’t feel like anything. The traffic jam does.
FPS works the same way. And the number that catches the traffic jam is your 1% low.
Your average FPS is hiding the bad parts
Average FPS is the easy number. Count all the frames over some stretch of time, divide, done.
The problem is what averaging does to your worst moments. It blends them into thousands of good frames until they basically vanish.
Back to the drive.
Say a game runs a smooth 130 FPS for most of a minute, then chugs down to 25 for a couple of seconds when an explosion fills the screen. Your average might still read a happy 120-something.
That’s the open highway averaging out the pileup. The number looks fine. The trip didn’t feel fine.
A high average isn’t lying to you. It’s just answering a different question than the one you’re asking when a game feels choppy.
Average FPS tells you the game was fast overall, not that it was smooth in the moment that made you flinch.
Frametime is what’s really going on underneath
FPS is basically shorthand for something more exact called frametime.
Frametime is how long a single frame takes to draw, measured in milliseconds.
60 FPS means each frame takes about 16.7 ms. 120 FPS means roughly 8.3 ms.
Here’s the part that matters. Smoothness is about consistency, not the average.
On the road, traffic all moving at the same steady speed feels fine, even when that speed isn’t fast. It’s the stop-and-go that wrecks the drive.
Frametimes are the same deal. If every frame shows up in about the same amount of time, motion looks clean.
If they bounce around, say 8 ms, then 8 ms, then a sudden 40 ms, your eye catches that long frame as a stutter.
That’s why two PCs can post the exact same average FPS and feel completely different to play on. One’s cruising. The other’s tapping the brakes every few seconds.
Even frametimes feel smooth. Spiky ones feel like stutter, no matter how high the average climbs.
So what exactly is a 1% low?
Here’s the actual definition, and it’s simpler than it sounds.
Take every frame from your benchmark and line them up from fastest to slowest.
The slowest 1% of those frames is your worst 1%. Your 1% low FPS is the frame rate you were getting during those rough moments.
So a 1% low of 45 FPS means that during the ugliest 1% of the session, you were sitting around 45. That’s your traffic jam, measured directly.
One thing trips people up here.
The 1% low isn’t a random 1% sample of frames. It’s specifically the worst ones, which is the entire point of it.
You’ll also see 0.1% lows reported next to it. Same idea, just zoomed further into the slowest 0.1% of frames, so it catches the really nasty one-off hitches.
Your 1% low is the frame rate during the worst frames of the run, not a random slice of them.
Why 1% lows predict how a game actually feels
Your brain doesn’t clock the peaks. Nobody has ever noticed the one frame where everything briefly hit 140.
What you notice is the dip. The stutter when you whip around a corner. The hitch when a new area loads in. The drop when the screen fills up with effects.
1% lows measure those dips head-on. That’s why they line up with how a game actually plays far better than the average does.
Picture two setups.
One runs 120 average with 40 FPS 1% lows. The other runs 90 average with 75 FPS 1% lows.
On paper, the first one’s faster. So which would you rather play on?
The second one, every time. Its bad moments never get anywhere near as bad, so it feels smoother in your hands even though the big number is lower.
A high average with weak 1% lows is a stuttery game wearing a good score.
How to check your own 1% lows
You don’t have to take anyone’s word for your own numbers. You can pull them yourself.
The usual combo is MSI Afterburner plus RivaTuner Statistics Server. It’s free and it throws live stats on screen while you play.
CapFrameX is the other one I’d point you to. It’s built for exactly this, and it spits out your 1% and 0.1% lows plus a frametime graph you can actually read.
Some games ship their own benchmark tools that report lows, and Intel PresentMon does it at the system level for whatever you’re running.
However you grab the numbers, do it during real gameplay, not on a menu screen.
Play through a busy area, a big fight, whatever part tends to chug. That’s where the dips live.
And give it a couple of minutes. A ten-second capture is too short to catch the hitches you’re actually hunting for.
Benchmark the messy parts of a game, not the calm ones, or your numbers will lie to you.
What good and bad numbers look like
There’s no magic pass mark here. But the gap between your average and your 1% low tells you a ton.
On a healthy system, your 1% lows usually land around 60 to 75% of your average.
Average 100 FPS with lows near 70? That’s a smooth ride.
When that gap gets wide, something’s choking.
100 average with 30 FPS 1% lows means you’re eating regular stutters even though the headline number looks great.
This is why hardware reviewers always list 1% lows right next to the average. Two GPUs can trade punches on average FPS while one holds way steadier lows, and the lows are what you’d actually feel day to day.
A big gap usually isn’t the graphics card’s fault. It points at your CPU, your memory, or your storage.
The wider the gap between your average and your 1% low, the more the game stutters.
How to actually improve your 1% lows
So how do you close that gap?
Start with the junk running in the background. Browser tabs, Discord overlays, and update services all quietly eat CPU time and cause hitches, so close what you don’t need before you play.
Shader compilation is a big one in modern games. That stutter the first time you walk into a new area is often the game building shaders. Let it precompile at launch, or just run through the area once, and it usually smooths out.
Your CPU and RAM matter more for 1% lows than for your average. Slow or not-enough RAM, or a CPU that can’t keep the GPU fed, shows up as dips instead of a lower average.
Storage speed helps with traversal stutter, the hitching that happens as a game streams in new areas while you move. Moving a game off a hard drive onto an SSD often cleans that right up.
Last one. Watch the settings that lean on your CPU instead of your GPU. Draw distance, crowd density, and ray tracing all hammer the processor, and dialing them back can lift your lows even when your average barely budges.
Most of your 1% low problems trace back to the CPU, memory, or storage, not the graphics card.
The number to actually watch
Average FPS is fine as a rough gauge of raw speed. It just won’t tell you whether a game feels smooth once you’re playing it.
Your 1% lows will.
So next time you benchmark, don’t stop at the big average number sitting up top. Look at the lows underneath it.
That’s the one that tells you what the drive is actually going to feel like.