A couple of weeks ago I went to the Pacific Life Open in Indian Wells. It's got to be the greatest tennis venue for the true tennis fan. You have the opportunity to see and stand right next to the top men and women players in the world. It's awesome to see how hard these players hit the ball. From the grunting of Nicolas Massu to the shrieks of Sharapova, you're so close that you can hear every sound they make and many times there's just a short fence that separates you from the biggest stars of the game.

But let me get back to the topic of this month's column on getting more speed for your tennis serves. To start, here's a little quiz.

Can anyone out there find a flaw with this player's serve? She has quite a problem when you consider she hits her backhand harder than her serve. So hypothetically, if you could change something, what would it be? More knee bend? More arch in her back? Less forward tilt of her spine at the fully loaded, externally rotated shoulder position (picture #4)? Is there anything that can be done to improve this serve? I'll give you my answer later.

Roddick serve style

Since Andy Roddick has the fastest serve in tennis, it is natural to have many players attempting to copy his distinctive style. But before you do that, consider that his style isn't really so different in the critical phases of the serve. Look at pictures 6, 7, and 8 and you'll see what I mean.

If you look carefully at the 7th picture, you'll see that Andy's shoulder reaches the same scratch the back position (maximum external rotation) as anyone else.

In fact if you compare his hand position in relation to his elbow, his hand is lower indicating a greater amount of external rotation than Sharapova's. So while it resembles a baseball catcher throwing a baseball rather than a pitcher with a full windup, he does achieve similar positions in the most critical phases of the serve.

Still not convinced? Here's the USTA research on the Roddick serve.

USTA Research Summary - The Biomechanics of the Abbreviated and Traditional Serves
A research study, titled "A biomechanical comparison of the traditional serve and the abbreviated serve (the Roddick serve)." was recently completed by Dr. Robert Shapiro and his research group from the University of Kentucky. They investigated the movement patterns, the forces experienced by the shoulder and arm, and looked at muscle activations in 28 experienced tennis players as they performed an abbreviated or traditional serve.

Their research essentially found there are very few differences between the two serve styles:

• Peak velocities of the shoulder were similar in both serve styles.
• Peak elbow velocities were similar in both serve styles.
• Maximal external rotation obtained during the "cocking phase" was not different between the two serves.
• The muscle activity was higher in the infraspinatus muscle during the abbreviated serve; This muscle is one of the four muscles that make up the rotator cuff.
• Ball velocity was the same for both styles of serve.

From this, the investigators drew several coaching implications.

1. The abbreviated serve does not seem to place any additional biomechanical load on the shoulder of tennis players who perform the motion in a skillful manner.
2. The abbreviated serve does not seem to provide any additional power above what can be gained from the traditional serve.
3. Extra conditioning of the muscles that make up the rotator cuff may be needed, since these muscles are activated to greater levels during the abbreviated serve.

Extra care should be taken in teaching young or unskilled players the abbreviated serve, making sure the backswing incorporates appropriate trunk and hip rotation and that the shoulders stay in line with one another as well as the torso.

Isn't that interesting? No additional power is obtained by using the Roddick style service motion. Could it be that Andy is just more powerful and faster than anyone else is? Hmmmm.

Where do high speed serves come from?

A good study on the serve by Dr. Bruce Elliot (Elliott et al.,1995) has provided coaches with an appreciation (Table 2) of the role of individual segments in developing racquet velocity.

Table 2 - Approximate Contributions to Impact Racquet Velocity (%)

The above table shows that a number of segments, if coordinated correctly, lead to an effective stroke. Such information, however, must be treated very carefully as the action of some segments (such as the legs in the service) play and important role prior to impact, while others act in a way that enable other segments to operate more effectively. The high percentage attributed to the upper arm, particularly in the service action, in part reflects energy transferred up the kinetic chain from the lower extremities and trunk.

As in baseball throwing, the external/internal rotation of the shoulder is extremely important for serving. And as is the case with many sports, they are still in the prehistoric era by training the external/internal rotation of the shoulder using rubber bands or tubing. As I've said before, there's no way to properly train the rotational movements of the shoulder using a linear device as in the picture below.

As to training the second most important factor in the high speed serves, palm flexion. Several players at the Pacific Life Open were throwing a weighted football, which also works the external/internal rotation of the shoulder. It's a lot better than doing rubber band exercises.

What about the trunk and the lower body?

But Elliot's study does not take into account the energy or force developed by the trunk and the lower body. Several other studies have been completed with this in mind. These two are very important if you can read between the lines.

Olivier Girard, et al did a study called The Lower Limb Activity during the Power Serve in Tennis: Effects of Performance Level. They found that the vertical forces and coordination in lower extremities during the tennis serve were different between players of different skill levels but with the same Pmax (maximum power).

The other important, relevant study, Network: Trunk Biomechanics in Tennis, was done by Dr. Rafael Bahomonde. His hypothesis stated when generating high speed serves, the magnitude of the ground reaction forces was greater and that the EMG activation timing in the lower extremities was different in the highest skilled players vs. players with less skill.

Simply put, higher skilled players apply greater ground reaction forces, which then translates into faster loading of the trunk, shoulder and arm muscles, which in turn facilitates higher racquet speeds and therefore higher serve velocities.

They are able to apply more force to the ground in a shorter amount of time when compared with players of lesser abilities. This is similar to the faster runner applying more force in that short period of ground contact time that the foot touches the ground.

How to measure this?

Since average people don't have Jugs guns, force platforms or EMG measuring equipment in their basements, what can you do to measure relevant aspects of the swing without sophisticated equipment? First, measure racquet speed (you'll need a Swing Speed Radar from Sport Sensors). In order to measure racquet head speed at impact you must mount the radar at the height your racquet head will swing during the serve. You can also have a tall person hold the Speed Radar while standing on a chair. Then, do "dry swings" without a ball (if you hit a ball, the radar may pick up ball speed). All things being equal, the higher your racquet speed, the higher your serve velocity will be.

To measure improvement, take a baseline reading and keep trying to increase that speed. Ball speed can be measured as well but you'll need an expensive Jugs or Decatur radar gun that baseball coaches use to measure pitching velocity.

Time to impact

Next, you can use a video camera to shoot videos of yourself doing serves. If you can play it back on a VCR with a step forward function, you can measure your time to impact. Here's how it works. Each step forward of your VCR equals .033 seconds given that the video shoots 30 frames per second. Next, count the frames from maximum external rotation, multiply by .033 and you've got your time to impact. Obviously, the shorter the time to impact, the better. It represents how fast you can accelerate your body, limbs and ultimately the racquet head to the ball.

The pros

Andy Roddick's service motion is so fast that he can get maximum external rotation (MER) of his shoulder despite his abbreviated serving motion and explode in the other direction. He takes an incredible .22 seconds to go from maximum external rotation to impact. I think that many who copy this style are not able to get to this MER while using the abbreviated style. But only a video will show whether you do or not.

Maria Sharapova, with one of the best serves in the women's game takes .24 seconds to reach impact. Still blazing fast.

Oh, and the answer to the quiz about the woman pictured earlier? It's Elena Dementieva, who lost in the finals to Sharapova, she takes an eternity to get to impact. She takes .35 seconds to reach impact. Does that give you an idea of what her problem might be? It might not be a technique problem. Could it be that it's just a slowness problem?

If that were the case, I'd have her try to speed up her training routines so she can learn to explode more quickly during the serve. Slow medicine ball drills will be of little use for a serve like this. Yet, this is what she was doing in warm ups.

Instead, she could be using a jump chain to get a stronger, faster leg drive to develop more ground force and the SpeedChain to help her external/internal shoulder rotation increase speed. There's no reason for Elena to have her serve eaten for lunch by the top players and it may be the reason she may not win Grand Slams despite having arguably the best groundstrokes in women's tennis today.