I rank somewhat high on typeracer: http://www.seanwrona.com/typeracer/profile.php?username=kanzure http://www.seanwrona.com/typeracer/profile.php?username=kanz.

I rank somewhat high on typeracer: http://www.seanwrona.com/typeracer/profile.php?username=kanzure http://www.seanwrona.com/typeracer/profile.php?username=kanz... Maybe I can provide some advice. The most important thing is to type very quickly. Heated real-time chatroom activity (read: flamewars) are also helpful for honing this skill... But chats might not be your thing. One other thing that might help is going to conferences and typing transcripts. Here's a few that I have done: http://diyhpl.us/wiki/transcripts/ http://diyhpl.us/wiki/transcripts/ For example, this recently included a Bitcoin event where my typing was originally meant to be a community resource for after the event... but during the event this "resource" became a supplement to and alternative to live translation into Chinese (someone decided to throw it up on a projector while I was typing- no pressure): http://diyhpl.us/wiki/transcripts/2016-july-bitcoin-developers-miners-meeting/cali2016/ http://diyhpl.us/wiki/transcripts/2016-july-bitcoin-develope... A long time ago I thought of myself as a keyboard snob, but the truth was that I hadn't tried that many keyboards. After experimenting with some expensive keyboards, I have found that for me the actual keyboard doesn't make too much of a difference, unless it's an exceptionally bad keyboard. I would recommend looking for keyboards that support n-key rollover, which would allow you to experiment with plover if you ever wanted to go in that direction for typing. I use the default qwerty keyboard layout. It might be more productive to experiment with building alternative keyboards rather than changing keyboard layouts. Chording might be an interesting direction to pursue. Skip to using lots of MEMS accelerometer sensors and gyroscope sensors on a glove and get real-time motion capture of tiny finger motions. Use a few accelerometers per finger. You could spend a few hours making weird (easy) motions with your fingers, and then those state transitions could then be mapped to different symbols. You could also use a 3d model of the human hand, forearm and shoulder to compute the exact range of feasible and repeatable motions based on muscle anatomy, then sample from that range and assign symbols to different transitions. Typing quickly isn't everything. It doesn't help you figure out what to say.