Contributions to Racquet Head Speed Prior to Contact

• Elbow Extension: 35%
• Wrist Flexion: 24%
• Upper Arm Twist (internal rotation): 17%
• Trunk Twist: 10% (-3% from hips; 13% from shoulders)
• Forearm Pronation: 5%

If you are wondering at this point why the analysis was cut off prior to contact, it is because measurements taken at contact are subject to potentially large errors for reasons beyond the scope of this article. Still, mathematical extension and visual inspection of the available range of joint motion in Figure 3 allows conjecture with reasonable confidence that the segmental motion contribution in the 1/100 second between Figure 3 and contact would come from continued elbow extension, wrist flexion, upper arm internal rotation, trunk twist, and forearm pronation. However, as the elbow is nearly fully extended in Figure 3, the forearm contribution to racquet head speed from elbow extension, for many servers, would decrease rapidly in the final milleseconds before contact.

For those who possess the geek gene that I seem to have inherited, the chart below shows the primary contributors throughout the phase from racquet low point to near contact by body segment motion (with predicted contact values).

Based on this quantitative data we can conclude that John Yandell pretty much had it right, although his article leaves a few important things unrecognized and unexplained.

If you read his article on the "Myth of the Wrist" he states the swing pattern was a function of racquet position at the end of the drop, elbow extension that straightens the arm, and internal rotation of the hitting arm. Additional information provided from my quantitative analysis shows one should also consider seriously the twisting rotation of the trunk and motion generated from the shoulder joint complex. This allows a greater understanding of how the motion is really produced.

Further insight into serve mechanics is gained by exploring the main subject of John's article, the role of wrist motion. Wrist motion in fast, complicated activities like the serve is difficult to discern. To really follow the specific motion of the wrist, one would require a view of the motion along the precise axis of the joint movement. As this axis moves in space, a camera would need to move in perfect concert with it. Regardless of the frame rate of the camera, this viewpoint is logistically impossible to obtain.

If you followed my results you may have noticed that throughout the upward swing phase, the motion of the wrist is a significant determinant of the speed of the racquet. Wrist motion also effects the path of the swing. John’s article, while acknowledging some motion, discounts wrist motion as a causative factor, particularly around the contact point. His argument is that any motion of the wrist is a reaction to the motion of other segments and not a conscious effect of muscular activity.

Basically I agree with this argument. I also agree the benefit of the age old advice of snapping the wrist is probably insignificant although additional research is needed to determine this conclusively. There is little doubt, however, the motion in the wrist joint is actually extensive and a significant contributor to racquet head speed. This is very clear in my results and the results of other researchers. Regardless of the cause, the rotation of the hand and racquet will affect racquet head speed and path.

The preceding discussion may seem troubling. How can wrist joint motion occur without conscious muscular activity? To see how this could happen we need to look at one of those terms that gets thrown around loosely in discussions of tennis technique: “torque.”

In practical terms, “Torque” is simply the turning or rotational effect caused by a force. In the case of the serve, applying torque to a body segment tends to cause that segment to rotate. But torque can be applied as a result of direct muscular action, or as a result of the motion of adjacent body segments. The latter form of application is termed "motion-dependent torque." This is a major source of torque (and therefore segment rotation) in rapid motion of linked segments (i.e. upper arm, forearm, and hand/racquet in the serve).

One could anecdotally argue that much of the wrist joint motion measured during the approach to contact during the serve is passively accomplished through motion-dependent torque. This joint motion, combined with that actively caused by muscular activity early in the upward swing, leads to impressive joint rotation speed leading into contact. So impressive, that any last ditched effort to increase it at contact by consciously “snapping the wrist” would likely be fruitless.

The same argument could be expanded to the elbow extension. The data indicate that forearm rotation (due to elbow extension) is critical to the swing pattern characteristics, particularly in the second half of the upward swing phase. Yet, it is likely that part of this rotation is due to motion-dependent effects during the serve. In studies done in throwing, as an example, a nerve blocking agent that rendered the triceps muscle (elbow extensor) unusable was given to subjects. The results showed minimal differences in throwing attributes, compared to throws without the nerve blocking agent. This and other research indicates that the rapid elbow extension seen in the latter stages of a throw has little to do with muscle activity and a lot to do with the motion of adjacent segments.

The point is, because the action of a joint and associated rotation of an adjacent segment may not be driven exclusively by muscle contraction, in no way diminishes its importance in defining the motion. It may, however, limit its effectiveness as a coaching focus. This is where I agree with John: you can't train a motion-dependent torque in the same way you can train a muscle driven joint torque. Instead, the training (coaching) should focus on the ultimate cause of the motion-dependent torque, which is muscular torque applied lower in the chain of linked segments (at the shoulder joint or in the trunk or in the legs for example).

To conclude, I’m reminded of something I learned in a computer science seminar: "Data do not imply their own explanation." The ability to measure and calculate millions of numbers will not further the understanding of tennis strokes. Biomechanics alone is not the holy grail of sport understanding, and there are some issues the current methods of biomechanics can't address.

It’s also important to understand that biomechanical analysis itself is subject to a degree of error, that the analysis is time consuming and expensive, and that to date it has been difficult for academics, coaches, and players to work together productively. But the methods of biomechanics combined with knowledge and experience of tennis experts is a vehicle through which substantive advancement can be made in tennis understanding and communication. This collaboration can only help the tennis community.

TennisONE is an informational and instructional Website. Your feedback is important to us. Let us know what you think and what you’ve learned by emailing us here at TennisONE. 

Brian’s obsession with the great game of Tennis has been a life long constant. He played his first tournament at age 8 and has competed ever since at the state, sectional, and national level, including two NCAA Division III championship tournaments. He has 12 years experience as a working teaching pro. Brian’s stated passion is to merge Tennis Science and Tennis Practice into an inseparable package.

He is the Director of Tennis Biomechanics for Gamasports Technologies, and the founder of Bio-Trak Training Systems which will soon offer 3D stroke analysis at low cost to players at all levels.

Feel free to contact Brian directly via email at:
bgordon@gamasports.com