わた

8000hz provided 1.55% higher performance (p=0.000129) than 1000hz in a tracking task for me even when causing 9.48% lower fps. A p-value that low means that if polling rate really was irrelevant, a difference this large or larger would be extremely unusual: 1 in ~7.7k comparable randomized tests. The "threshold" for statistical significance in science is usually 0.05, for context. The test was a within subject randomized blind trial done at 6400 dpi on ground plaza sparky v3. The reporting rate (hz) was changed between runs programmatically while keeping it hidden from me to minimize expectation bias (that's why it's a blind test). More testing incoming, especially at different dpi and in more tasks. It just takes a long time to do even a single test (1+hr as I play 60 runs in a row without a break), and will make sure I can reproduce with a different mouse too. BTW, as expected, the biggest delta in FPS is when using 6400 dpi, as "saturating the polling rate" makes sure that you are at your possible lowest FPS due to interrupts at any time when moving your mouse. What remains to be seen is if the higher performance is also only achieved at higher dpi or if it's also there at 800 dpi. Will post a video in the next few days going over everything in detail, stay tuned!