Fascinating! The obvious question is how much of what we believe we know on a variety of topics from politics to food choices are "hard-wired" into our brains making it virtually impossible to change our perspectives?

It reminds me of the old Mark Twain quote:

"It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so."

I've never really bought into the "muscle memory" thing as a concept. I think it is more brain related, personally. I know, for instance, that the brain retrains itself over time to accept new paradigms. For instance, when someone wears "inversion glasses" (glasses that make everything appear upside down), the brain will eventually retrain the person and flip the perspective back to normal view, even when wearing the glasses!

It is not about knowledge or understanding. It is about automatization.

Riding a bike is a complex thing and when it is automatized, it is controlled in a diffent part of the brain, I think the cerebellum, than the thinking and contious controlling, which is done in the cortex. The cortex is very smart and can do almost anything, but it is slow while the cerebellum is fast and automatized.

This is why you can do complex things very smooth, because they are saved as an autonomous program, that functions without thinking, like walking, running, swimming etc. programmed by thousands of repetitions. When these programs run, and then the slow cortex wants to interfere, the system collapses, because the signals from the cortex come much too late. This is what happened to this bicycle rider. The system collapses and he falls.

Through many repitions (around 5000, depending on the complexity), you must create a new program, that then runs autonomous from the cerebellum again. That took him 8 months, his son 2 weeks.

If you want to change an autonomous program, let us say to refine or change your technique, his has to be done in slow motion, so that the cortex can controll the movement until it is then stored in the cerebellum again. If you execute it too fast, the cerebellum will take over again and you will do the old techniques, because the cortex then is too slow to give the correct signals at the right time.

Then it took him 20 minutes to reactivate the "old" program of riding a bike. If he would now continue to ride the normal and the reverse bike every day, the time for the switching of those two programs will get shorter, until he has both programs paralell, so he can ride either of the bikes randomly. The programs will run at demand. Then he has established a second program, that did not overwrite the first but is paralell available.

I teach Modern Arnis, a martial art with complex movements, of which some are autonomous "programs". Some very similar to the other, but still different. So I deal with and teach the above problematic every day.