My QWERTY→Dvorak data
The data
A few years ago, I switched from QWERTY to Dvorak. My WPM on QWERTY was over 100 (maybe something like 120 max) before I switched, and my WPM on Dvorak is similar now. Here’s how my WPM changed over time during the period when I was switching over:
It looks pretty logarithmic, which is interesting.
Here’re the data as a table:
| Days since start | Max WPM | Min WPM |
| 0 | ||
| 1 | 10 | 10 |
| 2 | 12 | 12 |
| 3 | 18 | 14 |
| 5 | 22 | 20 |
| 7 | 21 | 21 |
| 8 | 34 | 22 |
| 9 | 39 | 30 |
| 10 | 43 | 31 |
| 11 | 46 | 37 |
| 12 | 49 | 41 |
| 13 | 47 | 45 |
| 18 | 59 | 47 |
| 19 | 59 | 53 |
| 20 | 58 | 47 |
| 26 | 63 | 56 |
| 29 | 63 | |
| 30 | 72 | 63 |
| 31 | 69 | 61 |
| 32 | 72 | 61 |
| 33 | 77 | 64 |
| 34 | 77 | 64 |
| 35 | 76 | 71 |
| 36 | 81 | 70 |
| 37 | 78 | 68 |
| 38 | 82 | 73 |
| 39 | 83 | 77 |
| 41 | 80 | 75 |
| 45 | 85 | 78 |
| 46 | 80 | 73 |
| 47 | 78 | 77 |
| 51 | 85 | 78 |
| 52 | 79 | 72 |
| 54 | 88 | 81 |
| 55 | 89 | 81 |
| 57 | 86 | 86 |
| 62 | 84 | 84 |
| 66 | 82 | 82 |
| 67 | 74 | 74 |
| 89 | 94 | 94 |
| 93 | 92 | 92 |
| 113 | 88 | 86 |
| 123 | 90 | 79 |
| 124 | 84 | 84 |
Background info and caveats
- I wasn’t intentionally trying to switch quickly, so my WPM didn’t increase as fast as it could have.
- I used learn.dvorak.nl (which is even available in Esperanto!) and tested myself with Monkeytype (importantly, Monkeytype by default doesn’t include punctuation and capitalization, which I’m now realizing matter and should’ve been tested).
- Switching took some transition time. I didn’t go cold-turkey; I waited until I had some proficiency with Dvorak (maybe 30 WPM on dvorak.nl?), and then switched to Dvorak on Discord at some point and typed messages excruciatingly slowly. I think both cold-turkey and non-cold-turkey approaches can probably work, depending on the situation.
General thoughts on keyboard layouts
I’m not convinced that Dvorak is actually better: I haven’t come across any really good studies (Wikipedia might be dubious on such a polemical topic, as it might be on the propadeutic value of Esperanto (though rereading this now, I actually have no idea how dubious Wikipedia is on either of this topics)) and I’m reevaluating a lot of my past beliefs. I also haven’t done the work to evaluate the other keyboard layouts worth considering. I can type around 80–100 WPM at least, so I’d need some pretty big potential gains to look into this more.
(Before you ask, I don’t have a knockdown argument for Dvorak over Colemak, Workman, &c.; I think I just knew of Dvorak most and didn’t preamble my keyboard layout switch with several hours of research starting with the Google query “best keyboard layout”. It is convenient that Dvorak, at least, is available out-of-the-box on both Windows and Ubuntu. I don’t really know the situation for other layouts, except that Ubuntu has many of the other alternative keyboard layouts too, and also a mostly translated Esperanto interface. Go Ubuntu!)
All the same, I find pretty plausible the claim (whose original source I’ve forgotten) that people who have bad typing habits in their current layout (usually QWERTY), in the process of switching to a different keyboard layout, often learn the new layout properly, thus increasing WPM and general QoL. In this case, switching to Dvorak or another keyboard layout might be a reasonable idea.
Things that make it a less reasonable idea:
- You often have to use keyboards whose layout you can’t switch.
- You often use shortcuts like ctrl+c, ctrl+v, ctrl+w, and ctrl+z and don’t want to have to learn new ones. (I think there might be ways to keep the QWERTY shortcuts, though.)
- You often use the curly braces ({}), which get moved to a position that seems slightly more inconvenient. (I use them a lot for LaTeX.)
- It gets harder to type on things like TTYs or GRUB where one seemingly can’t change the keyboard layout (unless maybe one can?).