Same Person, Different Monitor, Different Score
Play the Reaction Time game at home and then at a friend's place, and you may notice the score shifts slightly. Your condition feels about the same — so why? One surprisingly big reason is monitor refresh rate. Reaction-time tests time you from the moment a stimulus appears on screen. But a monitor doesn't redraw the screen continuously; it refreshes on a fixed schedule, so a small random delay gets inserted between "the game triggered the stimulus" and "it actually reached your eyes."
Refresh Rate and Frame Delay
A monitor redraws the screen at intervals determined by its refresh rate (60Hz, 120Hz, 144Hz, 240Hz, and so on). 60Hz means 60 redraws per second, or one frame roughly every 16.7ms. 144Hz is about every 6.9ms; 240Hz is about every 4.2ms.
The problem is that a stimulus can be triggered between refreshes. If the game fires the stimulus just after a frame has been drawn, you have to wait until the next refresh for it to appear — up to one full frame of delay. At 60Hz that delay is a random value between 0 and 16.7ms. At 240Hz it's between 0 and 4.2ms. The higher the refresh rate, the smaller the variance of that delay.
The Numbers in Practice
The same person, in the same condition, can record a slightly slower average on a 60Hz monitor than on a 144Hz or 240Hz one. Because the delay is random, on average about half a frame gets added. Half of 60Hz's frame is about 8.3ms; half of 240Hz's is about 2.1ms. That difference — roughly a few to 10ms — can show up in your average.
You might shrug off 10ms. But in Reaction Time, 10ms is not a negligible unit. Against your own stable baseline, 10ms is a swing comparable to a real condition difference. And that's the average — on any single trial, up to a full 16.7ms can land at once, so individual results wobble even more.
Input Latency and Browser Timers
Refresh rate isn't the only factor. Input latency also shakes the result. After you click the mouse, the signal has to travel through the computer, reach the browser, and be recognized as a "response." Mouse polling rate, USB handling, the OS input queue, and the browser's event-processing delay all add up, and they differ from device to device.
Browser timer precision is another variable. The timers browsers use (such as performance.now()) are high-precision, but the event loop and rendering pipeline can introduce a tiny gap between when the stimulus was actually triggered and when the timestamp was recorded. This is why web-based reaction-time tests tend to run slightly slower and with more variance than lab measurements taken with dedicated hardware.
Why It Differs from Lab Measurements
When cognitive-science research measures reaction time, it uses high-refresh-rate CRT or specialized displays, dedicated response devices, and precision timers. The moment of stimulus onset and the moment of response are pinned down at the hardware level, so device-induced delay is controlled down to the millisecond. A web game, by contrast, relies on a general-purpose monitor, an ordinary mouse, and browser timers — so even the same person usually records a bit slower than a lab figure.
So if a study reports 250ms and you score 300ms, you're not necessarily slower. Different measurement environments put the numbers on different scales. The relative change within the same environment is what carries meaningful information, not the absolute number.
Conditions for a Fair Comparison
When comparing reaction-time results, keep this in mind:
- Measure on the same device. Matching monitor refresh rate, mouse, and browser is the minimum for a fair comparison. If you want to compare scores with a friend, at least match the refresh rate.
- Use the same refresh rate. Don't mix 60Hz results with 144Hz results. If you have a high-refresh monitor, keep that setting consistent across comparisons.
- Take multiple trials and average. A single result is heavily influenced by random frame and input delay. An average of 5 or more trials is far more stable.
- Don't fixate on absolute numbers. Lab figures and web-game figures are on different scales. The gap from your own baseline is the more honest signal.
The conclusion is simple. Reaction Time results are device-dependent. Refresh rate, input latency, and browser timers affect the result on the order of a few to 10ms. So judging "who's faster" requires measuring in the same environment, and any comparison that doesn't meet that bar should be taken as a rough reference at best.