I Respect the Thorough Nature of Your Inquiry, Sir
Here's what I received a while back:
Hi Bill,
In About What You'd Expect (October 2016) you ran two online reaction tests and noted that one was "probably more accurate" - .21-.23 on one, .25-.27 on the other. That gap is mostly the input path, and almost every one of these tests hides it behind a single number.
I built a new benchmark tool around that problem. Results are ranked separately by device class - mouse, trackpad, touchscreen - and the per-device offsets are labelled as extrapolation rather than measurement, because they are. The baseline itself is fitted from Human Benchmark's published statistics (n=81,000,000), which pools every input device into one figure; splitting it is my extrapolation, and the page says so. There is also an audio mode, with the caveat that Bluetooth output adds its own delay.
Not asking you to edit a ten-year-old post. But if you ever revisit reaction time, or test Eli again, this is a fairer instrument than the ones that print one number and stop.
No account, no ads, nothing to buy.
https://benchself.com/
Both the test and the website around it are excellent. The email was also a masterclass in explaining why this tool is different and providing comprehensive information.
If you were ever interested originally, it's a good time to revisit. My best 5-test run (average) was 184ms. Reaction time is still strong, even with my "advanced" age. It makes me wonder what it was like when I was young.
<< Home