Might be worth going over a few things that haven't been covered since the early weeks of TREG which may help new - or even some more experienced - testers.
Of course, there is going to be an adjustment period. The key angle may not be what you are used to (that was an issue for me) and the reach is quite different (though in a good way).
As you go through the adjustment period, it helps if you know that the heck you are missing. Sure, a particular error may stick in your mind if you make it a lot, but many other errors may not, even if they happen a fair amount of the time.
One problem is remembering what the heck you missed. But another is HOW you missed it. So, the best solution I found for the most common stuff (words, not numbers or symbols) is to use this typing test:
It's apparently done for Dvorak layout, but I don't see anything in it that really affects other layouts so it should be fine for all who use English since it is common English words.
Tests are 1 minute, so you can do it quickly. For it's calculations, if you correct an error, it isn't counted at all. And if you make two errors in the same word, it only counts it as one.
But that is all secondary to the fact that at the end of a test, it tells you what words you got wrong AND how you spelled it.
I would take that list and, keeping a text file, enter that information. Something like this:
HN - 3
Which would mean I typed N instead of H 3 times. Or you could take the opposite order and use it to say you typed H instead of N 3 times. Whatever works best for you.
It may not be what you want to do from day one. After all, you'll probably be making all kinds of mistakes and lots of them. But once you feel a little comfortable with it, then this process can help.
Typically I'd do 20 tests (not necessary all at once) with all these errors listed, plus the total errors (also my WPM score). I also tracked "special" errors such as:
Spastic - sometimes I could tell my muscles just lost control for a moment.
Reverse - this is when I simply reverse two characters.
Nothing - When I failed to push the key down enough (which is related to being spastic since when it happens I can almost always tell my finger just locked up.
Wrong key - This is different than just getting the wrong character. With multiple characters on one keycap (3 or 6, not counting layers), some errors are hard to feel and thus you may not be sure if you goofed or the TB messed up. In the early weeks or months, it could be either!
But if you hit the wrong keycap entirely, you know the problem is you!
Over time, I was surprised to see that the single biggest error I got was getting nothing because my finger simply locked up! TextBlade can't help that!
Anyway, then I'd start another set of 20 tests and compare to see where I was and was not improving. In some cases, I would adjust my position slightly. For me, this was most common when dealing with the pinky. Especially the right pinky because of the angle I wasn't used to.
But, eventually, I could use this list to guide me in which key boundaries I should change. Don't want to do that right away. At least not for most (I do recommend, if you have trouble, that you might want to change ENTER and the apostrophe so it is less likely you'll get ENTER my mistake since in some websites, that will send a message! Same for P and Delete). But the others I'd wait to see how things develop.
You can't change all of the boundaries, but I think the ones which have caused the most trouble can be. There is still one I need and hope that will be added later.
I usually made note of boundary changes I made before the next series of 20 tests so I could go back and see if a change actually caused other problems.
It does take some effort to transcribe the errors into a text file, but I found it to be well worth it. I slightly reduced that time with each set of 20 tests simply by copying my error list from before (after removing the numbers). That way I only had to create entries if there was a new error.