Reaction Time vs Input Latency: Why These Two Numbers Are Not the Same Thing
Launch Companion Diagnostic Tool
Test your hardware right now in your browser with zero install:
Open Reaction Time TestYour reaction time and your system's input latency are completely separate values, and mixing them up will send you chasing the wrong fix. Reaction time is a biological delay: the interval between your visual cortex registering a stimulus and your motor cortex firing the signal that contracts your finger flexor digitorum muscle. Input latency is an electrical and firmware delay: the time between that muscle movement closing a microswitch and the game server acknowledging the resulting packet. One lives in your nervous system; the other lives in silicon, wire, and code.
What Reaction Time Actually Measures
Human simple reaction time to a visual stimulus sits between 150 ms and 300 ms for most adults. That range is not random noise. It breaks down into discrete biological stages: photoreceptor activation in the retina takes roughly 20-40 ms, signal propagation along the optic nerve to V1 cortex adds another 20-30 ms, motor planning in the premotor cortex adds 50-100 ms, and neuromuscular transmission from spinal cord to finger takes a final 10-20 ms. None of those stages are touched by swapping a USB cable or enabling 1000 Hz polling.
Fatigue, caffeine, sleep debt, and ambient lighting all shift your reaction time by measurable amounts. A tired player running on five hours of sleep will often test 40-60 ms slower than their rested baseline. That delta is larger than the latency difference between a 125 Hz and 1000 Hz mouse. If you want an honest baseline, use the Reaction Time Test across at least 20 trials and discard the top and bottom 10 percent of results before averaging.
What Input Latency Actually Measures
Input latency is the sum of every electrical and computational delay between hardware actuation and rendered frame output. It is not one number; it is a chain of numbers that add together.
- Switch actuation to USB packet: A mouse running a 1000 Hz polling rate sends a new HID report every 1 ms. At 125 Hz that interval stretches to 8 ms. The microswitch itself, whether it is an Omron D2FC-F-7N, a Kailh GM 8.0, or a TTC Gold 80M, closes in under 1 ms mechanically, so the switch is rarely the bottleneck.
- USB host controller processing: Adds roughly 0.1-0.5 ms depending on chipset and USB controller driver stack. Plugging into a rear motherboard port versus a front-panel header can change this slightly because front-panel headers often route through an internal hub.
- Driver and OS input stack: Windows Raw Input path is faster than legacy DirectInput. On Windows 11, the HID stack processing time sits around 0.5-1 ms under normal scheduler load.
- Game engine input sampling: Engines that sample input once per frame at 60 fps introduce up to 16.7 ms of additional latency on top of everything above. Uncapped or high-framerate modes reduce this significantly.
- Render pipeline and display: GPU render time plus display panel response time plus any display processing like overdrive or local dimming control. A 1 ms GtG IPS panel with 4 ms display processing still adds 5 ms total. Display latency and reaction time are frequently confused because both are measured in milliseconds and both affect how quickly a stimulus reaches your eye.
You can measure your mouse's polling consistency directly with the Mouse Polling Rate Test to confirm whether your device is actually hitting its rated interval.
Comparison: Typical Values Side by Side
| Component | Typical Delay | Improvable by User? |
|---|---|---|
| Human visual reaction time | 150-300 ms | Yes, through practice and rest |
| Mouse at 125 Hz polling | Up to 8 ms | Yes, upgrade polling rate |
| Mouse at 1000 Hz polling | Up to 1 ms | Already near floor |
| USB host controller | 0.1-0.5 ms | Marginally, by port selection |
| OS HID input stack (Windows 11) | 0.5-1 ms | Minimally, via Raw Input API |
| Game engine at 60 fps | Up to 16.7 ms | Yes, raise framerate or uncap |
| Display panel GtG + processing | 1-10 ms | Yes, disable display processing modes |
| Total system input latency | 20-40 ms typical | Partially |
The takeaway from that table is blunt: even a fully optimized system still adds 20 ms of latency before your action appears on screen. Your reaction time floor is 150 ms. The biological component is roughly six to eight times larger than the hardware component for most players. Spending $200 on a 4000 Hz polling mouse matters far less than sleeping eight hours before a tournament.
How to Isolate Each Number in Practice
To get a clean reaction time measurement, you need a test that removes as much system latency as possible from the equation, or at least holds it constant. Run the Reaction Time Test on the same machine, same display, same browser each session so that the hardware latency is a fixed offset. Your score will still include display latency and click registration delay, but since those are constant across sessions, the delta between sessions reflects genuine biological change.
To isolate input latency, you need external hardware measurement: a high-speed camera at 240 fps or higher pointed at both your mouse button and the screen simultaneously, or a purpose-built latency tester like the LDAT or a microcontroller-based photosensor rig. Software-only methods cannot measure their own delay accurately because they rely on the same pipeline they are trying to measure.
For keyboard input specifically, switch actuation point matters. A linear switch with a 1.2 mm actuation like the Gateron Yellow LP bottoms out faster under hard keypresses than a tactile with a 2.0 mm actuation, which affects the moment the electrical circuit closes. Check your keyboard's registration consistency with the Keyboard Test to confirm keys are registering on first press without bounce or dropout.
Frequently Asked Questions
Does lowering display latency improve my reaction time score?
It lowers the total stimulus-to-action loop time, which makes your scores on browser-based tests appear faster because the visual cue reaches your eye sooner. Your underlying biological reaction time does not change. The improvement is real in competitive play but comes from reducing system delay, not from training your nervous system.
What is the difference between input lag vs reaction time in competitive games?
Input lag is the hardware and software delay between your physical action and the game registering it. Reaction time is how long it takes you to decide to act after seeing a stimulus. In a real firefight both are in play simultaneously, but only reaction time can be trained. Input lag is engineered down through hardware choices.
Can a 4000 Hz polling rate mouse meaningfully help?
At 4000 Hz the polling interval is 0.25 ms versus 1 ms at 1000 Hz. That 0.75 ms reduction is real but sits well below human perception thresholds and below the jitter introduced by game engine frame timing at anything under 300 fps. It matters at the absolute top tier of professional play where every fraction counts, but it will not compensate for a 250 ms average reaction time.
Why do my reaction time scores vary so much between sessions?
Biological variability is the dominant cause. Hydration, caffeine timing, ambient light level, time of day relative to your circadian rhythm, and hand temperature all shift your motor response time. Hardware jitter from an inconsistent polling rate or a worn microswitch is a secondary factor. If your scores vary by more than 30-40 ms session to session, rule out hardware first with a polling rate check, then look at your pre-session conditions.
Does display latency reaction time testing work the same on all monitors?
No. A monitor with 2 ms of display processing latency and a monitor with 15 ms of display processing latency will produce systematically different scores on any browser-based reaction test because the visual stimulus appears at different times. Always run baseline tests on the same monitor if you are tracking improvement over time.
Is 200 ms a good reaction time?
200 ms is a solid result for a simple visual reaction test with no anticipation. Elite FPS players in controlled lab conditions typically score between 150 ms and 190 ms. Anything under 150 ms on a browser-based test is almost certainly the result of anticipation rather than genuine reaction, since the biological floor for simple visual reaction is around 140-150 ms even in trained athletes.
Launch Companion Diagnostic Tool
Test your hardware right now in your browser with zero install:
Open Reaction Time Test