Hacker News new | ask | show | jobs
by blackearl 1631 days ago
It says he competed at 267wpm in the article. The monkeytype site totally bans 300+ so I think that's still very impressive.
3 comments

"with 76% accuracy" - I'm surprised typing competitions even allow scores below some accuracy level, e.g. 95% or so. Missing every fourth letter means it's basically gobbeldygook, and if you were actually typing anything where you even needed to be remotely accurate you'd spend at least 2x as much time revising.
The accuracy measurement is flawed, because the firmware of the device is deleting chorded letters and retyping the chorded word, which artificially reduces accuracy
If it's happening at the firmware level why wouldn't it just resolve the word internally and output only the completed word to the PC?
Yes. It’s actually artificially increasing accuracy.
Yes but the accuracy percentage is low because the software’s essentially guessing what he’s typing (not unlike swype keyboards + autocorrect on a phone though implemented differently). It would be way slower to correct errors and type at 100% accuracy because each error has to be corrected based on some cognitive process (like looking at autocorrect suggestions for example) which is far far slower than correcting a typo in a QWERTY keyboard. I can type around 110 WPM with 95% or higher accuracy and I can also feel the majority of typos I make and correct them without active thought.
Yes, but at only 76% accuracy. If you were to include the time spent on going back and correcting mistakes, I bet he would be down in the 50-100 wpm range, if that.