Important parts of strokes can be learned very quickly after attempting, first try or even unexpectedly. Other parts especially that occur afterward in a stroke are much more difficult to change.
Frederick Alexander, of the Alexander Technique, said that when we try to learn a new athletic technique he thought that this is what is going on -
My interpretation of Alexander's discussion - Say that you are trying to perform a new complex motion, a new stroke technique. When the motion involves both old and new subconscious instructions to your body, that is, nerve signals to many muscles, learning the new motion becomes difficult. When actually attempting to perform the new stroke, the subconscious mind is uncertain at certain instants about the new nerve signals for the new motions. Some of the old unwanted nerve signals are likely to be used at these instants, spoiling the new motion after that instant. I recommend Alexander's book to those interested in some concepts that may apply to learning new strokes. ?
For the OP's friend, she may have learned the proper techniques as "a young girl" with its nerve signals, similar to riding a bike. The nerve signals are still functioning for her tennis strokes.
Moon walking would be an example of something that is very difficult to change. Not to mention, I can't even understand it....
Amateurs use more upper arm swing at the shoulder joint for tennis strokes and not much body turn. (Look over several courts of tennis players to see this.) Nerve signals for an upper arm swing from the shoulder using deltoid and pec are very different than nerve signals to turn the upper body using leg and trunk muscles. The body turn is new and maybe the old shoulder motion nerve signals aren't easily substituted by errors of the subconscious. Going from swinging from the shoulder to a stroke based on body turning can be very quick for that important part of the motion. After studying Djokovic, I was able to do a body turn forehand first try.
