I do boxing for several years in a gym. Besides physical training, it also includes sparring with an opponent. Since boxing, like most martial arts, relies on quick reflexes rather than thinking, I would like to automate my movements, in order to immediately react during sparring.

I was wondering, if I could train the sparring part at home, without any opponents, simply by visualization or mental practice.
By repeatedly mentally rehearsing the required movements for a given situation. This could mean working on how to defend a certain punch or combination, it could mean rehearsing ways to counter, or how to move before or after an attack.

Are there any studies about that form of practice?
How could that kind of training look like?

2 Answers 2


There are numerous studies showing that visualization improves performance, often nearly as well as actually doing the thing in question. It never gets better results by itself, but in practice you'd be adding visualization to your normal activity, not replacing your activity with it.

You can read more about this here:


What I think is lacking is a proper understanding about the limitations of visualization and how it can be used most effectively. If you have no experience with the thing you're visualizing, for example, it won't do you much good. The more experience you have with it, the more vivid your visualization of it will be. The more vivid your visualization is, the better your results will be.

Also, if the activity you're trying to visualize has a lot of dynamic variables, visualizing is going to be much less effective, because it requires a lot more brain power. So when you visualize something, you should be isolating it to just a very simple part, and something you have a good amount of experience with.

For example, with the 3-point throw from basketball, there aren't many variables which significantly effect your throw. It's just you, the ball, and the basket.

For kickboxing, you can take that idea and work on small pieces of your game. And I would specifically look at things that occur frequently, because those are the things you're going to be able to visualize most effectively.

Combinations are okay to work on. They're small enough that you can visualize them. But I'd say chop them up even smaller. Ask what the most important thing is about that movement. Usually it's what happens before you throw the first punch. You're looking at subtle "tells" from your opponent's body language and movement to determine how to move, and then the combo technique follows from there. So concentrate your visualization on this part. It will cue your subconscious muscle memory to react more quickly. That's going to give you the biggest bang for your buck as far as visualization is concerned, in my opinion.

Another question that hasn't really been answered by the research is whether or not visualization combined with doing the thing you're visualizing will make more improvements than either alone. I would imagine it's not going to hurt and probably would help, but I can't say for sure.

Hope that helps.

  • The problem is that you can't quantify the visualization for a study because it's so individual.
    – JohnP
    Commented Oct 19, 2019 at 0:42
  • @JohnP Yes, nobody can see inside of someone's brain to know what they're thinking about at any given moment. But you don't really have to for this experiment. You can just look at things you can observe, which are the results of the experiment. There will be some people who are terrible at visualization, while others are "gifted" at it. That doesn't matter. What matters are averages in the observed results. So I don't see too much that's wrong with the methodology in theory. Peer review would detect methodological problems. Commented Oct 19, 2019 at 1:07
  • No, I mean that there is no way to compare. Take person A, at a set point in performance. Train for 6 months using visualization and they improve Y percent. There is no way to backtrack and start from point A again and compare without visualization.
    – JohnP
    Commented Oct 19, 2019 at 5:39
  • @JohnP Again, that shouldn’t matter in determining what happens to people on average. It just needs randomized groups of people, with each group assigned different tasks (visualization only, action only, visualization plus action, and do nothing). Then you look at average results. You won’t be able to predict how any particular person will perform, but you can know the probability. All pretty standard science stuff. Commented Oct 20, 2019 at 23:17

Don't try to automate your movements. You will never be able to perform spontaneous actions by becoming an automaton (even though such 'movements' may look automatic). Read Bruce Lee's teaching of water and Intercepting Fist. He knew how to immediately intercept the opponent's attack. He did it with sharp intelligence.

In mental training, it is assumed that one can perform thought experiments. Only experienced martial artists can do good quality mental training. Why? Thought experiments are only possible after you ammas a large number of real-life experiences. After years of real-life impressions, you can start to visualize realistic scenarios. Moriji Mochida could do such visualizations.

"... it took me 50 years to learn the basics of kendo using my body. After I turned 50, I started the real discipline. I wanted to perform kendo using my mind and spirit. When you reach 60 your lower body weakens. I used my mind to try to reverse the drawbacks. When you're 70 your whole body weakens. That's when I trained my mind to be imperturbable. With a still mind, the mirror inside you reflects the opponent's mind" - Moriji Mochida

Source: YouTube

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.