Just completed ~80 hours of tutor time and ~3 months of usage on Typemax v1 and wanted to post a snapshot of my progress before making changes to Typemax. Sadly I reset my browser data a few weeks back and forgot that also cleared tutor data. So I have no data on tutor training from 60-80 hrs.
Twiddler Tutor (avg 76 WPM, max: 84 WPM):
I will be making some changes to the config as suggested by AlexBravo, mainly around reducing pinky load. I also want to start training tethered MCCs (untethered MCCs are not fast enough to boost WPM past ~75). Hoping the combination of the two will allow me to hit 85-95 WPM.
I decided to also demo some other typing tests that I've come to like, since they measure things the tutor doesn't:
Live Chat Typing Speed Test (53 WPM)
I like this one as a small step up in difficulty from tutor since it forces me to fix mistakes using backspace but is still just random lower case words, spaces and no symbols. Forcing me to fix mistakes causes a significant drop in WPM since when I make a mistake in tutor, tutor either stays at the mistyped letter until the correct one is typed (ignoring mistyped letters) OR more insidiously, notices I've typed the next few letters after the mistyped letter and it automatically marks the mistyped letter as a mistake and skips ahead. While this is great for initial training, it greatly exaggerates the average WPM overtime when compared to normal typing.
The Typing Cat Typing Speed Test [50 WPM]
I like this test as the most representative of a true WPM score since it has all the elements of real typing: capital letters, punctuation, some symbols, sometimes numbers and "natural" words occuring as they would in speech or written text (though sentences don't always make sense).
ZType (Lev 16 Score 2093)
My favorite typing game. Its tragic flaw is in not requiring the most commonly typed character, space. While the default mode doesn't have capitals, symbols or numbers, you can get these by loading custom text content.
Finally, I'm proud of these scores but after using my Twiddler part of each workday for quite a few months now, I can say definitively that these speeds aren't representative of my normal Twiddler typing. My normal typing is much slower. Getting this fast requires significant concentration and daily practice. But I think it's a worthy exercise to train for "speed runs" as a way to benchmark configs and versions. When doing speed runs it becomes obvious pretty fast when something in the config is suboptimal. The only downside of speed runs is that it gives new users the wrong impression of what a realistic WPM is for the config. My guess is that my normal typing speed is about ~75% of my speed run WPM (38).
Lance