Sampras Serve:
New Filming Protocols
Classic Motions

John Yandell

I can't speak for everyone on Tennisplayer, but speaking for myself, I have yet to tire of watching the great Pete Sampras play--especially watching him hit that wonderful, silky bomb we call his serve.

Recently we had the chance to film Pete again, playing an exhibition with Sam Querry at the Tiburon Peninsula Club in Marin County, California, just over the Golden Gate Bridge from San Francisco. It was a tremendous event organized by our friends on the tennis staff at TPC, Steve Jackson and Brandon Coupe (both Tennisplayer subscribers).

Read on to find out more about the event and what and how we filmed. But if you just can't wait to read the analysis of Pete's serve, skip down about 6 or 7 paragraphs to the "Analysis" subhead below.

Over the weekend at TPC, there was a college alumni competition featuring players including Roscoe Tanner, Paul Goldstein, Scott Davis, Justin Gimelstob, Eric Taino and other former collegiate stars, some going back a few decades.

Then Sunday afternoon Pete showed up to provide a final burst of star power. Over 2,000 people watched as he and Sam split the first two sets and then Sam won a third set super tie breaker.

It looks like Pete is starting to get ready for the exo series he has planned with Federer starting in November. How did he look? Overall he seemed a little rusty and he missed some balls, but he still looked very athletic and moved as fluidly as ever. And the serve? See for yourself in this gorgeous new footage. It's still there. (The backhand might be another story, but more on that in another article.)

As we've discussed in the Forum (Click Here), Brian Gordon was here with us for a week to do a massive quantitative filming for his research and his PhD dissertation. Prior to the Pete exo, we were up in St. Helena in the wine country, filming about 25 top Division 1 college players at a team event called the Napa Classic, run by our friend Doug King, the tennis director at the Meadowood Resort. This was big time college tennis with teams from Cal, Southern Cal, North Carolina, and defending champion Georgia.

But Sunday we made it back down to Tiburon to film Pete. What did Brian find out studying the Sampras serve? Well, after he spends a few hundred hours in a dark room with his data, we'll know the answer to that, and he'll be sharing the info with us on Tennisplayer.

In the meantime, though, we thought you'd like to see some of our new qualitative high speed footage shot by our technical editor, Mr. Aaron Martinez. You'll be seeing more of this type of footage on Tennisplayer and we're curious to know what you think about it. In addition to the animations, we've embedded some sequences as click through movies right in the article. Just click on anyone of them, and you can study Pete's motion frame by frame for yourself, just like in the Stroke Archive.

The Analysis

What did I see when I looked at Pete? One of the first big series of analytic articles I did was on Pete's service motion. (Click Here.) Sometimes as players get a little older you see their motions evolve, change, or deteriorate. Sometimes players exaggerate tendencies they already have. Sometimes a tendency can become a significant flaw, and problems develop. This was what was so interesting about my work with Johnny Mac on his serve in the early 1990s. (Click Here.)

But there were no real problems in Pete's serve, at least that I could see. To me it looked as good as ever. And that's pretty friggin' good. In our previous studies we found that although Pete's velocity tended to stay between 120mph and 130mph, his phenomenal spin rates--averaging over 2500rpm on his first delivery with a huge topspin component--were what made his ball so heavy and almost impossible to return. (Click Here.)

The Tiburon footage shows the unique constellation of elements that make his serve is still in place. Let's take a look at a few of the key points as seen in this new footage.

Racket Path

For starters take a look at the depth of his racket drop. The racket falls along his right side and the face of the racket is on a plane that is perpendicular to his torso. Notice how his upper arm is almost in line with his shoulder. This shows his tremendous shoulder flexibility.

The racket drop position is a goal most players can achieve. But not too many players can do it by externally rotating the arm backward in the shoulder joint as far as Pete.

Now watch the action of the hitting arm frame by frame going up to the ball. The elbow extends. Then 5 or 6 frames before the contact, watch the hand and racket start to turn into the shot. As this rotation progresses the wrist moves from the laid back to the neutral position at contact.

Now the rotation continues out into the followthrough, with the arm and racket rotating primarily as a unit from the shoulder joint (with some probable additional elements of independent forearm rotation). The extent of Pete's arm rotation, typically referred to as "pronation," is still incredible, easily 90 degrees after contact.

Next we see the famous early elbow bend. So many people and analysts have jumped on this as a signature move. The fact is that most players also bend the elbow at some point in the followthrough, it's just not as early as in Pete's motion.

Why this happens has been a topic of debate for over a decade. Some analysts have claimed it has to do with the "tremendous" forward wrist snap at the hit, but the footage shows that this is anything but an accurate description of what is actually happening. As the clips all show, both the bend at the elbow and any release of the wrist occur after the contact. (Want to read more on the Myth of the Wrist in the serve? Click Here.)

Ball Position

What I think is really happening has to be understood in the context of Pete's unique contact point. Pete's first delivery is really best understood as a 125mph kick serve.

Compared to most other top servers, his ball position is further to his left, closer to the edge of his head at contact. This means that his motion with his hand and racket is more radically upward. Look at the point when Pete's arm straightens out, roughly at the point of maximum pronation. Compare this to the Roddick serve in the same animation.

Look how much earlier Pete's arm straightens out and how much sooner he completes the rotation. It's much higher off the court than Roddick's, actually above the line of his shoulder.

My opinion is that the elbow bend is just the natural consequence of everything that happens before it in the motion. There is very little muscle tension in the arm. The arm moves upward to the hit. The arm straightens out and pronates.

From that position, there is really no where else for the hand and racket to go except down and back across the body in the continuation of the followthrough. So the elbow bends, the hand drops and the motion comes back the other way.The racket continues outward, downward, and to the left. It's just happening sooner in Pete's motion than for most players.

Again this is one of those critical cause and effect issues and I think trying to consciously create or imitate this position is a mistake. If the other elements of the serve are in place, it'll naturally happen at the right time and you definitely shouldn't try to force it to occur.

Torso Rotation

Pete Sampras may start his motion with his shoulders basically square to the baseline, but watch what happens during his windup. He turns off the ball as far as any player in the history of the game except John McEnroe. With McEnroe, the massive body turn was hard to miss since he started with his shoulders almost parallel to the baseline.

With Pete, the turn happens during the motion, so it isn't as obvious. But in my opinion it's key to understanding that massive combination of velocity and spin. Turning so far away in the windup, Pete naturally gets tremendous leverage when he rotates back the other way into the ball.

As we've seen in the previous articles on his motion, this turn and forward body rotation are directly related to his stance. At the start of his motion, his rear foot is offset, behind and to his left of the front. The toes of the back foot are turned open to the baseline around 30 degrees.

We all know how Pete starts in the ready position for his serve with the front toes raised up in the air. But as his arms drop, he foot turns and moves down so that it is flat on the court and close to parallel to the baseline. (Slightly more open usually in the deuce court.)

His torso also immediately starts to rotate away from the net. At the maximum turn, it's probably rotated something around 60 degrees from its starting position--we'll get a measurement from Brian Gordon on this eventually I can hope.

The amount of turn is directly related to his stance. If you draw an imaginary line across his toes, and another imaginary line across the front of his chest, they are essentially parallel at the completion of the turn. The angle of his stance corresponds to, and to some extent actually determines the amount of his body turn.

Legs

We also know that Pete has tremendous knee bend. But it's important to understand how that occurs. First pay close attention to the timing of his leg action. The coiling of the legs really only begins after he has released the toss.

Note that as the toss arm is going up, he is standing almost straight up and down with just a little flex in the knees, and the weight appears equally distributed between his feet. Now as the tossing arm continues to extend, he begins to drop his weight, bending the knees and then shifting the majority of the weight onto the left front foot.

The completion of the knee bend coincidences with the completion of the turn. And with the full extension of the tossing arm. At this point he is really coiled and ready to explode.

The legs, hips and shoulders all uncoil as he launches upward into the ball and outward over the court. The launch means that when he makes contact he is about 2 feet into the court, but it's also important to understand where the contact point is in relationship to his body. The ball is tossed in front, definitely, but note that the actual position of the racket when it strikes the ball is just at the front edge of the body--roughly even with the front of his face.

So in relationship to his body the contact point is slightly in front. But as we saw above is also quite far to his left. This means that as his body rotates into the hit, the torso is still quite closed in relation to the baseline at contact. This angle is probably something like 60 degrees in the add court, and maybe more like 30 degrees in the deuce. (Brian will correct me here I'm sure.)

This explosive full body motion, combined with the left ball position, are two factors that account for the incredible weight of shot or "heaviness" Pete produces. As our studies showed, his ball is spinning with over 5,000rpm by the time his opponent makes contact. And due to the topspin component, his serve bounces significantly higher than serves that may be as fast, but have less total rotation or less topspin.

Landing

A final point to look at is his landing. Like all top servers, Pete lands on his front foot, with the rear leg kicking back. The interesting point is that, although Pete has served and volleyed as much as possible over the course of his career, he typically lands only a shoe length or so inside the court. By contrast a player like Richard Kracijek, another great serve and volleyer, landed much further inside the baseline.

Pete also lands relatively balanced, with his body tilted slightly forward but not a lot of bend from the waist. Compare this to a player like Roddick, who actually lands a foot to two feet further inside the baseline than Pete and jack knifes over much more at the waist.

This landing probably accounts for Pete's somewhat lower kick back with the rear leg. His leg barely reaches parallel with the court. Roddick's kick back is much higher. Interestingly, Pete's landing is actually similar to Roger Federer.

So what about the serve and volley? In the old days the conventional wisdom was that you had to reach the service line to be in good position for the first volley. That rarely happens in the modern pro game, unless the opponent hits a weak floating return. Most the returns are coming back at 80 or 90 mph. And that is a major factor accounting for the lack of serve and volley play.

In the modern game, the first volley is usually hit about half way between the service line and the baseline. It's a function of now fast the ball comes back to the server. And it makes the first volley very difficult.

Pete's footwork pattern is interesting here. He lands only a foot inside the baseline on relatively good balance. But watch the next step. He takes a huge stride forward with the right, rear foot. This may be why he limits his kick back as well--to facilitate taking this forward step as rapidly as possible.

The split step then follows the cross step. So it's really three components: a left foot landing slightly inside the baseline, a big cross step, and then a split. And then, hopefully, a successful first volley. On a weaker return of course Pete's happy to close further, but we are talking about the limitations and difficulties of a tactical style resulting from the ball speed of the modern game.

So that's it for the look at Pete's serve. We also have some incredible footage of his forehand that I'll be talking about in the future, as well as a very interesting look at his backhand side, including how it was always a little different from the other top one-handers and what has happened to it over the years. Stay tuned.