Andrei Rublev:
Serve

Analyzed by John Yandell

Andrei Rublev can rocket groundstrokes with anyone. His first serve, while not at the high end of pro velocities, can be a weapon. But compared to most elite pro players, his second serve can be a liability, causing him lost points, double faults, and, at times, obvious frustration.

His second serve ball speed can be in the low 90s or high 80s but sometimes drops in the 70s or even high 60s. There are many women servers with higher second serve mphs.

Why is that? If we look at Andre's service motion in high speed video, we see he has a great, flexible shoulder. His internal shoulder rotation and the rotation of his arm and racket in the upward motion–so called pronation--is equal to the best servers.

But Andre has technical problems with his stance and the movement of his feet during the motion. These stem from his extreme pinpoint stance in which he drags his back foot up and then around his front foot.

What are the problems? I think his stance is affecting his leg drive to the ball. The movement of his back leg is also affecting the timing of his torso rotation. At contact he is significantly over rotated compared to the best servers, with his shoulders parallel or close to parallel to the net.

A Little Background

We know from researchers such as Brian Gordon, Ben Kibler, and Bruce Elliot that the use of the legs is a critical factor in high level serving. From maximum knee bend, great servers immediately begin the uncoiling of the legs in the drive up to the ball.

Andrei doesn't do this. He pauses at maximum bend for around 2/10s of a second. We know from research that holding a fully loaded position reduces the ability of the muscles to explode with maximum force.

Kibler has shown that the ideal is for both feet to push off together, and that this is best accomplished from a platform stance. It can work from a pinpoint stance with the right positioning of the feet. This is with the back foot directly behind the front foot. In both cases, the knee bend in both legs is maximized at the same time, and the push with both feet starts simultaneously or very close.

The drive with both legs is critical to the upward acceleration of the back hip and this in turn is critical to the rest of the motion and the development of maximum racket speed.

But in extreme pinpoint stances in which the back foot comes around and to the right side of the front foot, the contribution of the leg drive can significantly reduced. The extreme pinpoint stance, for reasons that aren't fully understood, is common in women's pro tennis, with Venus Williams being an obvious example. But it's more unusual on the men's side.

Knee Bend

So first how is the stance affecting the push off from the knee bend? If we look at a great platform server such as Roger Federer we see that the upward explosion in both legs starts immediately after the knee bend is complete. There is no pause. The bend and the upward thrust are a continuous motion.

Compare that to Andrei. His front knee bend is complete before he finishes dragging the back foot up. So he has to hold that position while he is still bringing the back foot up and around.

From this position his back foot leaves the court first, and continues to move forward and around. Look again at the Fed sequence and see that the back foot leaves the court a few fractions of second sooner than Andrei. It also stays behind his front foot instead of coming further around like Andrei's.

Could these relatively minute differences really have an effect? I think so. They are also the source of the other major problem in Rublev's serve, his overotation.

Great servers typically make contact in roughly the middle of the torso rotation. This means the shoulders are still partially closed to the baseline. The angle is around 45 degrees in the add court, and somewhat less in the deuce.

But Andrei's shoulders are already fully open or very close at the contact point. His back foot is almost even with his front foot. That motion appears to be dragging his hips and shoulders with it. Compare that again to Federer or any great servers and see that the rear foot is much further behind at the same point in the motion.

Solutions

So extreme pinpoint stance, delay at the knee bend, and too much torso rotation at contact. Is there a solution? That's what my friend Gary Swain, who is also Rublev's agent, asked me. So Gary this article is for you to ponder.

I think there are a couple of possible solutions, both starting with changes in the stance. In considering them, there is one other thing about Rublev's back foot to note first. He initially sets up his feet a little wider than shoulder width. In fact at the start of the motion his stance isn't that dissimilar to Federer.

But then he rocks way back on his front foot so his front toes are up in the air. At the same time he moves his back foot backward on a diagonal a foot or more. That makes his actual starting stance super wide and it looks awkward.

This modified back foot position is what forces him to drag the foot up There is no way to get to the knee bend otherwise.

So here is the first possible solution. What if he just kept his feet where they were in his initial set up and didn't slide the back foot further back? What if he then just dropped his weight down roughly equally on both feet and then kept them in place in a platform stance similar to Fed?

That change would definitely affect the rhythm of the motion, but possibly that change alone would be enough to change to the timing of the coiling and uncoiling of the knees. He might also find that he could actually increase the knee bend in both legs before uncoiling.

This change might also immediately improve the timing of his torso rotation, so that his shoulders were angled to the baseline instead of parallel.

And there is related possibility with this option. With his current stance, Rublev can't really rotate his torso back away from the ball very much in the windup. With the platform that factor could be increased. Almost all great servers rotate the shoulders back away from the ball more than Andrei.

Potentially those changes could a big positive effect on his mph. Not only on his second serve, but maybe on his first serve as well.

To make this work, he'd also have to pay attention to the timing of the kick back with the rear leg so that it went back sooner and didn't come around the way it currently does. If it continues to come around as much as it is now, likely the torso rotation pattern would not change.

Second Option

The other option would be just to modify his pinpoint stance. Instead of bringing his back foot around to the right side of his front foot, he could use a player like John Isner as a model.

He could bring his back foot up right behind his front foot, so both feet were roughly parallel to the baseline with his front foot in fact actually touching the back foot like John does.

With this option, he wouldn't be able to increase his rotation back away from the ball in the windup on the Fed model.

But he could synchronize the push between the legs so both legs were pushing and leaving the court at the same time. From that position he could get the delay out of the front leg knee bend. He might also be able to deepen the knee bend in both legs.

These would be interesting experiments to try on court. When Gary and I first started talking about Andrei's serve I said I would love to go on court with him for an hour with a compact high speed video camera and a pocket radar gun and try out one or both options. Gary what do you say?