Process of learning new layouts

I’ve had somewhat of a revelation in my learning of my self-designed letter arrangement layout.

Until recently I’ve been typing “on-the-fly”, just typing symbol by symbol as soon as I was confident in what I was doing. It had gotten me to somewhat of a plateau in my typing speed progress, at about 50% of my former QWERTY speed. Just recently, I have started delaying typing a word while training myself until I was ready to “burst” through most of it, and it seems to be making a heap of difference to my progress, at least subjectively. So i visualise the “shape” of a word on the keyboard, execute the movement, think about the next one, execute etc.

I don’t type any faster by doing so (slower, but not a lot) but I am more accurate, and it feels like by doing so, I am giving myself better training in muscle memory of common n-gram combinations rather than only increasing my familiarity with the layout. It also makes me feel a little less frustrated, as by doing so, I can type small words and fragments at about my former QWERTY speed.

Will have to see over time if this truly improves my speed (and I don’t have a valid null hypothesis to test it against so it’s certainly not scientific). But it would make sense on the surface, as the muscle memory of n-grams is probably the major factor in most people’s typing speed, so practicing that fluidly makes a lot of sense.

Has anyone else experienced something similar in their process of learning new layouts or keyboardio itself? Or any other unanticipated helpful techniques?


My research is in motor control, and what you’re doing makes perfect sense. Our sensory systems are a bit on the slow side, so part of developing motor programs is the “feed forward” part. We start moving with the assumption that the feedback we receive will be what we expected. If not, we tweak things a bit, which is the learning part.
Another part is stringing movements together. These “phases” of the motor program are often delineated by sensory events, like a finger contacting a key.
Then there is the motor imagery part of learning. By visualizing what you are about to do, you are actually using parts of your motor system. The same parts as if you’re doing the movement, except the “output” is turned off so you don’t actually move. This is why people (such as athletes) can visualize movements and improve their outcomes without ever doing the task.
TL;DR carry on! There’s a scientific basis for what you’re doing.


Heh, I’m actually a neuroscience student nearing the end of undergrad, though tending towards the sensory side more than motor. But my knowledge of similar things did guide my intuition on this, I think :slight_smile: I’ve only done a little learning on motor control, but that little did help. One of my last subjects next semester is on neuromotor control and in more depth, should be interesting comparing that to this experience!