There is an old article on Yandell's site about Sampras' serve. It stated that Sampras' 1st serve had a high degree of vertical rotation (some topspin). It also has some side spin, so to keep it simple, think of it as spinning on a 45 degree angle forward and with side spin. To my knowledge, the only way to get vertical (topspin) on the ball is for the strings to actually be moving up at contact. You can also see Federer's 1st serve arcs downward due to spin as does almost every other top pro. You can see the ball bend in flight.
So, if the racket face is moving down at contact, how do you explain the vertical topspin? Wouldn't they be mutually exclusive?
Now, to the more practical side. You should visualize and practice hitting up on your serve. Some instructors say up and over and that's OK too. To me, if we open a thread and indicate hitting down is correct, there is a risk of screwing up a lot of beginning and intermediate players.
Now, to complicate things. If you are hitting up AND the racket face has a bit of forward tilt at contact; it would be like hitting up on the top 1/2 of the ball. I certainly can visualize where the resulting ball line of flight would be outwards or even slightly downwards even though the feeling or visualization is hitting up.
In tennis usage the term "pronation" for the serve is often used to mean arm rotation seen at the wrist. This usage was well established before 1995 when research proved that the service arm motion involved mostly internal shoulder rotation (ISR) with some pronation. When the arm is straight ISR and pronation produce the exact same effect at the wrist.
ISR occurs at the shoulder and rotates the the upper arm bone (humerus) and every arm part below it, forearm, etc.
pronation occurs at two joints at the elbow and wrist and rotates (actually it twists the bones in the forearm) the forearm and everything below it.
Videos show ISR before impact on the kick serve. But there may be less than for the slice or flat serves. There is a good reason why there is less. For the kick serve or topspin serve the racket has to be rising more than for the flat or slice serves. To get this extra upward motion the kick serve is done by having a smaller forearm to racket angle at impact. (I have posted this many times and anyone can use a protractor on a high speed video taken from behind to confirm it for themselves, see below Topspin vs Slice impacts). The forearm to racket angle is changing rapidly and that is the main factor that causes the racket to rise during impact. This all occurs roughly 5-10 milliseconds before the peak of the racket path. For the slice or flat serves the impact is closer to the peak of the racket path. Remember that the ball is on the strings for about 4 milliseconds so these events require clear high speed video to see. Identify these points in videos as word descriptions don't communicate well.
It is often said in instructions that the body 'stays sideways' for the kick serve. For high level serves, I'd estimate that servers don't rotate their upper bodies as much, as indicated by the chest orientations. Body orientation occurs simply so that the racket face (when viewed from above) is oriented to direct the serve into the court (side-to-side direction). Otherwise, because there is less angular rotation from ISR when impact occurs the racket face would be facing too much to the left - so the upper body faces more to the right to compensate.
These pictures illustrate the shoulder girdle/chest orientations for the kick serve vs the flat and slice serves. Toly composite from rare Fuzzy Yellow Balls overhead videos. Is the ball 'tossed over or behind the head for the kick serve'? Measure the distance from the hand to the tip of the racket, why the difference?
![]()
Notice the chest orientation to the side for the kick serve, compare. Notice the racket orientation at impact. It is as if the chest orientation for the kick has been used to make the racket face the court. Overhead views show these things clearly. But there are few overhead views for current ATP and WTA servers. What variety is out there?
![]()
This picture illustrates the racket motion, impact angles and heights for the slice and kick serves. Note how the racket is rising at impact for a topspin serve vs a slice serve. Toly composite images. A view I sometimes have - that she is using the same basic ISR motion but for the kick serve facing her chest more to the side and impacting the ball lower in the racket path.
![]()
Forearm-to-racket angles at impact. Toly composite image.
![]()
Thread - Kick serve wrist movement.
https://tt.tennis-warehouse.com/index.php?threads/kick-serve-wrist-movement-video.557005/
Keep in mind that these are high level serving techniques and that the majority of tennis players do not use these techniques.
Is Groth really throwing his racket up or hammering the ball down into the court like driving a nail?
.....
90% of the racket path to contact is up but I believe on a flat serve it actually peaks a little before contact and then goes down again a little.
The feel is definitely up but at some point the arm reaches the end of its length and the racket can't go up more and then rotates around the hand.
Basically up until the arm is fully extended, then the hand stops and the hand turns inward so the racket comes forward. The hand doesn't come down, just the racket.
He is not throwing the racket - He holds on to the racket and does not throw it.
Much too literal. We all know that the server does not release the racket. We can still refer to this as a "throw" or as an upward throwing motion.
This study suggests that you can hit down from an impact point 9 feet high or higher, if you take into account gravity and drag which impose a downward curving flight. Many taller players hit from much higher than this. John Isner is said to make contact 11.5 feet high and 4-5 feet inside the baseline.@Dan R
I'd take it a step further and say that most, if not all, of the swing after the racket head drop is an upward direction. Much more that 90%, possibly even 100%. Serve is often hit with a slightly closed face but the racket is still moving upward during contact.
Racket face is not moving down on the ball as it might for an overhead at the net. It is conceivable that very tall players who jump high might be able to hit slightly down on the ball -- but I'm not sure that this ever happens on the serve. At most, in extreme cases, the racket had stopped moving upward at contact time and is moving forward (rather than downward) during the 4 ms or so of contact. But not sure that this ever happens either.
Sampras serve swing path. Blue dots are data points that show the position of the racket tip (equal time intervals). Note that the racket tip peaks a little bit after contact. This is a clear indication that the racket is still moving upward during (4 ms of) contact.
Note that this 2nd image shows Sampras AFTER contact. Ball has already left the strings. I believe that I see the ball more than 6 inches off the strings already.![]()
![]()
Here is a research publication discussing impact and the 'angle of projection'.
A kinematic comparison of successful and unsuccessful tennis serves across the elite development pathway. David Whiteside, B. Elliott, B. Lay, M. Reid
https://www.researchgate.net/public...s_serves_across_the_elite_development_pathway
Figure 4 sums up some of the information. Projection angle, measured leaving the racket, is the angle between the ball's trajectory and the horizontal. Zero degrees is horizontal. This graph covers from +1° above the horizontal to -8° below the horizontal. It basically says for the serves covered that almost all have downward projection angles and that the faster serves have considerably greater downward projection angles. I don't remember more details about the types of serves for this graph. This is an excellent reference to be aware of.
![]()
The Rod Cross TW article on the kick serve has some projection angle discussions here and there in the article. One point that Cross discusses is the very rapid closing rate of the racket face.
http://twu.tennis-warehouse.com/learning_center/kickserve.php
Quote from the above Cross article on the kick serve
".......Despite the fact that the server swings up at the ball in a kick serve, the ball must be projected downward below the horizontal for a good serve. The serve angle, A, is typically between 2 and 6 degrees below the horizontal. ..."
For the second rotation produced by ISR, the rotation rates of the upper arm might reach about 3000°/sec. Then in 4 milliseconds of racket-ball contact the racket might rotate roughly 12° due to ISR.
The rotation rate of the racket from ISR is somewhat faster than the rotation rate closing the racket face. The rotation axis for the ISR is in the upper arm and is roughly vertical. The rotation axes for closing the racket face are roughly horizontal, at wrist, at shoulder, at hips, etc. . We could measure the closing rotation rate of the racket on the 6000 fps video.
If we simplify very complex ball-string interactions to a word - 'downward' - and ask whether the racket face is 'downward' for impact, how do we account for the significant closing of the racket face during ball-string contact? Do we have research such as the Whiteside, Elliott et al work on projection angle for the racket face angle? Is 'downward' defined for the very first initial contact of the ball and strings? If so defined is that 'downward' feature useful for the overall resulting 'projection angle'? Projection angle is an important and well known biomechanical measurement.
I think 'supination' is over-rated... you can basically have the palm facing the ball all the time and still hit bombs.
I find peace in your marshmellowAll these scribbles!
![]()
Crucial element(s)?
1. Extreme hip movement forward like 3' or so!
2. Extreme knee bend, almost like sitting on a chair ( 90 degrees or so)
3. Unique style of racket take back and up to trophy. Palm and racket face completely down and parallel to the ground, racket pointing forward.
You're basically saying he's just arming the ball then!None of the above. OK, the hip thrust does add pop, provided you stay sideways.
I would be willing to bet that Groth has internal shoulder rotators that are twice the size of a normal man. The Tennisspeed guy posted MRIs the Japanese did of an elite American sprinter and a Japanese sprinter. The American guy's hip flexors, which can only be gauged from an axial view, were two or three times as thick. With Fed, his serve is a product of otherworldly joint flexibility and sublime timing. Groth seems like more raw power.
With Fed, his serve is a product of otherworldly joint flexibility and sublime timing. Groth seems like more raw power.
You're basically saying he's just arming the ball then!![]()
The answer is ..he's talented.Ok, he's hitting it down into the court.
Second question: What's the secret to his fast serve? He is not the only strong tall guy among pros.
Crucial element(s)?
1. Extreme hip movement forward like 3' or so!
2. Extreme knee bend, almost like sitting on a chair ( 90 degrees or so)
3. Unique style of racket take back and up to trophy. Palm and racket face completely down and parallel to the ground, racket pointing forward.
Bonus video
If you want to understand what he is doing squeeze as much detail out of high speed videos as you want.
Words introduce images. Images and feelings are the best information that we have.
Throwing a ball up seems to be a good training simulation for ISR during the tennis serve. Video comparisons of the ball throw up show very similar ISR. See Pat Dougherty video.
I believe these types of tips are situational. Have a student with generally sound serving motion hitting only spinners and struggling flattening it out and pumping up pace? Make him toss farther into court and vizualize hitting top part of the ball.To add, Bollitieri tip is to imagine hitting the top part of the ball. This tip is said to encourage pronation.
I believe these types of tips are situational. Have a student with generally sound serving motion hitting only spinners and struggling flattening it out and pumping up pace? Make him toss farther into court and vizualize hitting top part of the ball.
But as a general, literal description of the serve mechanics (what I believe is @Chas Tennis favorite)? Or even worse - for someone having issues finding space to hit between the net and service line without slowing down the motion?
With regard to OP question of "hitting up" or down, it is complicated by the fact that the racquet is moving upwards and closing simultaneously.During the 5 milliseconds dwell time, there's a lot of stuff going on. It happens too fast. In this situation, I don't think high speed video (as in previous post) is all that helpful. Best to go by feel and imagery and visual cues even if that imagery does not correspond to what is physically happening on video.
Example: Bollitieri visual cue of "cracking an egg on top of the ball". That probably is not what is seen on video but that visual cue seems to achieve the desired result.
I want to bring to your attention a tip that I got from the great coach Nick Bollettieri. This player, she would reach high to the ball, so she was always getting up high but then she'd make contact with the ball on the back of the tennis ball and from that point then she'd try to push the ball down and so for as much extension as she was getting she was getting zero power on her serve.
So Nick had this really cool quick fix, and the way he described it was, treat the ball like an egg and when you go up with that high reach, to go hit the ball try to pretend like you're gonna crack the egg on top. And that was a good way to kind of coerce this player into getting pronation because there's no way to get that racquet up above the equator of the ball without turning that palm and forearm down.
It was like such an easy quick fix for her and all of a sudden I probably saw 15 miles per hour start to come off the racket on her serve.
![]()
That’s exactly what I was talking about - a situational tip.she was getting she was getting zero power on her serve.
It would be best if we could get some very high speed video of Groth hitting a 1st serve. The closest thing that we have is Aussie teaching pro, Adam Kennedy, hitting a 142 mph serve. Filmed at 6000 fps. Kennedy, like Groth, is pretty tall.
![]()
Racket head, although closed, appears to be moving slightly upward during contact -- tip still moving upward slightly before & after contact? At most, racket is moving forward rather than downward during impact.
Ball appears to be hit very close to the peak of the toss. Extremely slight, almost imperceptible, downward (falling) motion of ball just prior to contact. Ball does appear to continue falling after contact. This could be due to gravity (and the fact that that the ball was already falling) rather than Adam hitting down on the ball.
Note that the ball has some topspin after it has left the strings. This would happen if the the racket face is moving upward on the ball during contact.
The answer is ..he's talented.
Say, even if your 1,2, 3 were all true (the secret), would you still be able to do them?
The secret to everyone's success is their talent. Other people just can't understand someone else's talent/secret.![]()
Ok, he's hitting it down into the court.
Second question: What's the secret to his fast serve? He is not the only strong tall guy among pros.
Crucial element(s)?
1. Extreme hip movement forward like 3' or so!
2. Extreme knee bend, almost like sitting on a chair ( 90 degrees or so)
3. Unique style of racket take back and up to trophy. Palm and racket face completely down and parallel to the ground, racket pointing forward.
Bonus video
That's even better
They al have pre-pronations but the most extreme I've seen is Stan.Isner has this 'pre-pronation' move, almost the exact opposite of Raonic's 'pre-supination' move at the start of the motion.
I think 'supination' is over-rated... you can basically have the palm facing the ball all the time and still hit bombs.
also notice the contact is made high on the string bed, say the 5th cross from the top? this will produce more top spin due to the shape of the pocket and the slightly bigger gear effect (you can see the racket tip bent backwards) for top spin... the high contact also adds ~2 inches of extra length for more speed (which maybe canceled slightly by the smaller trampoline effect with the off-center hit)
you will notice the extra top spin, vs. say contact made in the dead center, the 10th cross.