My local tai chi motion caputre facility

I just discovered there is a tai chi motion capture facility only about half an hours drive away from me at the University of Portsmouth! It seems that it is open to the public in principle (by arrangement), so I'm sorely tempted to go down there myself! Maybe this will be a realistic way in to start scientifically analysing my movements. And a lot of fun of course :-)

Obviously this is just a general motion capture facility, but it seems that one of the lecturers there is a fan of tai chi so sets his computer science students the task of building a virtual tai chi trainer system using motion capture and computer graphics. A very cool project no doubt. You can see a video here.

Perception of weight distribution in tai chi stances

In tai chi we make a significant conscious effort to control the distribution of our weight in postures and stances. This led me to wondering, that although we talk about having 70% of our weight on one leg (for example), how accurate and exact is this 70% figure? I intend to carry out some tests myself, but in the meantime I did a quick search for scientific literature on the subject and discovered this particularly relevant journal paper entitled "Perception of unilateral lower extremity weightbearing during bilateral upright stance".

The abstract of this paper is below, but in summary it seems that humans can judge the weight distribution to within a 7% error for asymmetrical stances [i.e. if you feel like it's 70% of you weight, it is likely somewhere between 65% and 75% of your weight]. It seems reasonable to assume that with tai chi training, this error margin is reduced, as it is a conscious, learned and developed skill say to a 3% accuracy [i.e. if you think it's 70% of you weight it is likely somewhere between 68% and 72% of your weight] - which is pretty accurate.

What this means is that this 70% figure is not just an arbitrary feeling that our teachers are talking about (as it would be if the average percentage error was high implying humans were bad at judging weight distribution, like say >20%), but a real scientific quantification of what we should be doing. Hence when we talk about moving 70% of our weight onto one leg, not only should this feel like 70% of weight is on that leg, but practically it should also be 70%. Hence it makes sense to play around with bathroom scales at home to actually measure how good your stances are and improve your perception of weight distribution for better control.

Abstract of "Perception of unilateral lower extremity weightbearing during bilateral upright stance"
"The primary purpose of this study was to describe the error in 61 healthy subjects' perceptions of weight-bearing at three target levels during bilateral upright stance. The secondary purpose was to describe the effects of age, sex, lower extremity dominance and target weightbearing level on the error in perceptions of weightbearing. Weightbearing was determined while subjects stood on digital scales. They adjusted their weight in an attempt to bear 25, 50, and 75% of their weight through a designated lower extremity. Three trials were allowed at each weightbearing target, and the results were averaged. Each subject's error in perception of weightbearing at each target level was determined by taking the absolute value of the target percent weightbearing minus the mean actual percent weightbearing. The mean errors at the 25, 50, and 75% targets were 7.3, 3.3, and 7.7%, respectively. The magnitude of the error was unrelated to age. An analysis of variance showed that error was not dependent on sex or whether the dominant lower extremity was used for making judgements. The error did differ between target levels. Clinicians cannot assume, based on the findings of this study, that individuals can accurately judge the percent weightbearing they are placing through one of their lower extremities during bilateral upright stance."

Note I disagree with the final sentence of this abstract, as I believe that this weight distribution perception is the key skill that tai chi practitioners are developing, and hence their ability to judge it is significantly better than the normal population. You can hypothesise an extrapolation of this fact from the numerous scientific studies demonstrating that there are benefits to doing tai chi for improving balance. I would argue that improving your balance is primarily driven by improving the accuracy of your weight distribution perception.

Further thoughts on Tai chi for Kinect

Tai Chi on Fitness-Evolved supposedly tracks
over 1 million points of your body

I've been delving a bit deeper into Microsoft Xbox Kinect's Tai Chi system that I talked about before.The interactivity of the system is really what inspires me, but I realise now that it also has a hidden benefit: visual feedback. What is powerful about the system is that it shows a digital representation of your own position on the screen in real time. This is placed right next to your "instructor" so you can immediately spot the difference and make the corrections both consciously and subconsciously. In my time training in tai chi, I have found that that there are significant advantages to either watching yourself in a mirror, or videoing yourself and watching it back. After all, in tai chi we are trying to rewire our properception, and unfortunately what the brain tells you is not always correct. You may think that you are standing upright, and feel like you are, but look at yourself in the mirror/video and you suddenly realise you are leaning over to one side. These video analysis techniques are widely used in professional sport coaching - and of course, top sportsmen only do things if they are worthwhile! The kinect system allows you to do this in real time, which I think is great. Indeed, this review by FindingSanctuary highlights this as one of its key positives:

"When you’re playing Fitness-Evolved it actually projects you into the game! You see yourself stood next to a virtual instructor. What’s more amazing is that as you follow the instructor through the movements Kinect tracks you’re movements and offers advice."

As that review discusses, whether or not it's good at teaching you tai chi, is almost secondary. What this system is doing is popularising tai chi, and inspiring people, who might not otherwise do so, to "have a go". This game is not designed for a serious practitioner, (and indeed looking at these videos of the exercises, it seems that it is really chi gung rather than tai chi), but that is to miss the point. If these 15 minute xbox kinect tai chi workouts help to relax some people, then that is a benefit in itself. Tai chi snobs may turn their noses up, but really it is a way of demystifying these internal arts, and bringing it to the (western) masses. Hopefully some of these people will then go on to join classes and of course the more people who want to practice tai chi the better!

Another fascinating article I found over at the well respected wired.com, was about the research potential of tai chi and kinect. I have previously talked about how mood can be affected by posture, which is where a lot of the relaxing benefits of tai chi come from. The wired article extrapolates on this principle, to show how kinect systems, which encourage you to adopt certain postures, can in turn affect your mood. Essentially you can design games that subconsciously make you feel different ways. Fascinating. What's more, is that this has potential to be used in scientific research, as I can explain by outlining a little scenario:

Researchers could design two different chi gung/tai chi/yoga workout programs to be performed using the kinect system. Volunteers could carry out the exercises daily for a period of weeks with the data recorded. You could also ask the volunteers a series of psychometric questionnaires throughout the trial. At the end of the trial what you would then have would be data on how people's emotions and moods were affected. You would also have data on how well/accurately the volunteers performed the tai chi moves. It would then be possible to compare the two groups of volunteers who each did different exercise routines to determine which one was better/worse.  

Serious chi gung practitioners will say that each of the different exercises is connected to a different meridian/element in Chinese theory and can help to balance and strengthen the associated aspects (including emotions) of that meridian/element. The kinect experiment I outline above, would provide empirical data on how different chi gung moves affected emotions to test these assertions. I would hope such testing would be welcomed by the chi gung community.

Martial arts training aids

Tai chi is traditionally a formal and prescriptive practice. However, there are lots of scientific insights that can be gained from other disciplines and sports science and incorporated into training regimes for improved performance. Just because "that is the way it has always been done" does not mean that the best way, and as good scientists it is our duty to investigate them.

Anyway, I'm not really going to talk too much about that here, instead, given that background, here are a couple of modern training tools that I use that I think have helped me improve my Tai Chi. Equally they would benefit anyone who does any form of martial arts

1) Swiss ball. As I have a computer-based job in the real world, I obviously spent a lot of time sitting in front of it. Sitting immobilises your spine and causes all of your balance reflexes to become dulled over time. By sitting on a Swiss ball you engage in "active sitting" that maintains your balance skills continually strengthens your back.

2) Poi. Poi is basically a performance art similar to juggling. The movements require you to continuously spin pendulums around your body, which with practice can form hypnotising patterns (experts even set them alight for fire-poi dispalys). However from an exercise perspective, I'm not sure there's anything better for toning your arms and shoulders. As you have to continually keep the ball end of the poi spinning, there is a centrifugal force that you must counteract with your arms. As the spin direction of this force changes you must therefore exert this force in a continually changing direction, which work the muscles for dynamic strengthening in all directions. Anyone who practices Tai chi will recognise this need to project force outwards in all directions is a fundamental tenet of tai chi. Not only is poi strengthening, but the diverse rotation planes mean that your shoulder and arm joints become significantly looser and more flexible. I really do recommend poi to anyone who would like to loosen their shoulders, elbows or wrists.

3) Contact juggling. I do not do this myself, but it's my intention to learn. Lots of Tai Chi is about visualising a ball and controlling it. Indeed I understand there are a whole sets of Tai Chi exercises and forms that use a ball. You just have to watch a contact juggling video to see the obvious parallels.

4) Core Balance Discs. These inflatable discs are slightly oval in shape and therefore to stand on them requires quite a lot of balancing skill as it makes you very unstable. I have two of them, and by putting one under each foot and then performing chi gung or silk reeling exercises, it really takes the difficulty to a new level. It's amazing how much your understanding of the movements improve (by identifying where you are unbalanced or weakest) as you wobble around all over the place. Best practice privately, at least to begin with, as graceful it is not.