FPS vs Frame Time in CS2: Why Smoothness Is More Than Average FPS
Learn the difference between average FPS and observed frame times in Counter-Strike 2.
If you want to understand why a 160-FPS game sometimes feels worse than a steadier one, look at the time between individual frames, not just the average number of frames per second. Both measurements matter, but they answer different questions.
The mathematical link
A perfectly stable 200 FPS corresponds to an idealized 5 ms frame interval. A stable 100 FPS implies 10 ms. But frame delivery is not perfectly constant in real gameplay.
That is why you cannot use average FPS to infer how frequently large frame-time spikes occur. A handful of very long frames might make movement feel inconsistent even if the counter remains impressive.
- 1000/100 FPS = 10 ms.
- 1000/200 FPS = 5 ms.
- A single real frame can be far above the average interval.
Measure the right event
Enable CS2’s frame-time HUD and watch a reproducible hitch. The value on the left is in milliseconds and the FPS number beside it is averaged.
Consider the moment of the hitch, not a random paused screenshot. Record the game map, workload and background activity. If both frame times and network missed ticks change, test each path rather than pick one in advance.
- Use the same resolution and display mode.
- Change one setting, then compare repeat runs.
Do not confuse percentile data with one spike
A credible percentile analysis needs frame-time samples gathered over a period and a defined calculation method. One example of a 30 ms frame does not reveal the “1% low FPS.”
Different benchmarking tools may also calculate percentile summaries differently. When comparing results published by others, make sure the method and test scene are compatible.
- Treat spikes as events, not automatic percentile statistics.
- Use a reliable measurement capture if you need rigorous comparisons.
Example: a frame-time distribution is missing
At 200 FPS, the ideal interval is 5 ms. If one screenshot captures a 28 ms frame, that frame took over five times the ideal average interval. You can express the comparison accurately—but you cannot infer how many other frames were late without collecting more samples.
The same applies to a claimed “1% low of 60 FPS” obtained from a single HUD screenshot. That percentile requires a defined frame-time series and method. If you need that level of evidence, capture a repeatable benchmark session rather than estimate it from averages.
Primary reference
Valve’s Source 2 frame-time description.
Put the numbers to work
Use the FPS & Frame-Time Calculator to turn these observations into a repeatable test or calculation. It works entirely in your browser.
Continue troubleshooting
Related questions often require a different measurement. These guides help narrow the next step:
Use the measurements above to form a testable hypothesis, not to guess which component is broken. Return to our free tools for another controlled check.
Open the toolkit ↗