1. Sports
Send to a Friend via Email
You can opt-out at any time. Please refer to our privacy policy for contact information.

Neural Networks in Table Tennis - Learning New Techniques

Theorizing about my practice...


Neural Networks in Table Tennis - Learning New Techniques

In theory, theory is the same as practice - but it practice it isn't...

© 2010 Greg Letts, licensed to About.com, Inc.
Over the years that I have played table tennis, there have been many occasions where I have thought that I had the answer to a particular question, only to find out many years later that I was wrong or had overlooked other solutions. So these days I try to stay open to other points of view, and be more willing to admit that I sometimes simply don't have a clue.

I'm encountering one such issue at the moment, and since the situation is interesting I thought I'd throw it out there for better minds than mine to theorize about. I'm afraid there is a fair bit of scene-setting required in order to explain where I am baffled, so bear with me please.

The Situation

Readers may remember that back in November 2010 I mentioned that I had radically altered the grip I was using, which resulted in correspondingly large changes in my forehand and backhand technique. Since this was the first time in many years that I have attempted to learn techniques that were very different from my normal strokes, I naturally paid a lot of attention to the problems I was having along the way.

For the first month or so, I made a lot of mistakes with my basic technique, and quite a few of these mistakes were large errors, such as accidentally using my old swing path, getting the racket angle completely wrong, or even reverting back to my old technique now and then. I didn't actually make a lot of errors in choosing the correct category of stroke though - I rarely blocked when I wanted to counter-attack, for instance.

It's now four months since I changed grip (March 2011), and the number of technical mistakes that I make has steadily reduced in frequency as time has gone by. It's now pretty uncommon that I make a large mistake with my swing path or racket angle. It's even rarer that I revert to my old strokes, although this occurs a little more often on my backhand than my forehand side. And interestingly enough, my old conventional grip now feels quite strange if I attempt to play a normal forehand!

However, I have noticed that the number of times I choose the wrong type of stroke has increased. The two most common examples of this are when my opponent attacks the ball with a better shot than I expected, and I am suddenly put under extra pressure. If my opponent attacks to my backhand, I frequently block the ball with my long pips even though I really want to drive the ball instead. Similarly, if my opponent attacks strongly to my forehand, I often block the ball back instead of attempting a counterloop, which is the stroke I really want to play. It's not a question of balance or recovery, it's mainly just playing the wrong stroke by instinct.

As a table tennis geek, I naturally spent a little time surfing the web trying to find out how long it takes to learn a new habit, and what are the roles of the conscious brain vs the unconscious brain when quick decisions have to be made in sports. And while there isn't exactly a wealth of information out there, I learned that the old theory of it taking 21 days to learn a new habit is now largely debunked (it seems that learning a new habit to replace an old one can take anywhere from 66 to 245 days), and that it's possible that when quick decisions are called for, our unconscious neural nets in our brain instinctively make a decision and send it to our conscious brain and our muscles, so that the decision arrives in our conscious thoughts just before the motor action occurs. This gives us the incorrect impression that we consciously made a decision and acted, when actually we reacted and then become conscious of our reaction.

So What Does This Mean?

I'm not sure exactly, but what I'm thinking is that when I first changed my grip, I started to develop a new subconscious neural network related to the new grip and the stroke techniques involved. Naturally enough, this neural network had many gaps to begin with, resulting in frequent errors when I attempted to play strokes. On the other hand, I was concentrating hard on learning the new grip and stroke techniques, so I did not have many occurrences of accidentally slipping back into my old neural network when choosing what stroke to play. So most of my errors were those of execution, not selection.

Now, four months later, I've developed my new neural network to the point where I am comfortable with the new grip, and my techniques are quite solid and consistent. The mistakes I do make in regards to technique tend to be relatively minor errors in my grip position and racket angle. It's pretty rare that I revert to my old strokes, and when I do it tends to happen more on the backhand than the forehand, which I'm guessing occurs because my new backhand technique is closer to my old backhand technique than my new forehand technique is to my old forehand technique, so maybe it's easier for my subconscious to skip back to my old neural pathways on the backhand side.

It would also explain why I'm now making a few more errors in my shot selection. Since I have improved my new neural network, I don't have to concentrate as much to use it successfully. This means I can spend more time thinking about other things, such as what my opponent is doing, what spin and speed is on the ball, where is it going, and so on. But when I suddenly come under extra pressure, the lesser amount of concentration that I have on using my new neural network (driving with my backhand long pips or counter-looping with my forehand) is not enough, and I go back to my old neural network, which I have used for many years, and I block the ball.

That all makes some sort of sense, but the one thing that baffles me is when I revert back to my old neural network and block the ball instead of counter-attacking it, you would expect that I would block the ball using my old technique as well - resulting in me missing the shot completely, since my racket angles are completely different between blocking the ball with my old and new grips.

But in fact, almost every time this occurs I block the ball with the correct racket angle, meaning that I am using the old neural network for choosing my stroke, but the new neural network for deciding how that stroke should be played.

Anybody got an good explanation?

©2014 About.com. All rights reserved.