Discussion in 'Tennis Tips/Instruction' started by 3kids, Jul 20, 2017.
Love this coach's short videos! Each video is packed with lots of good info.
Sadly i dont understand what he means by seperation. Well i think I do in that the line of the hips should be different than the shoulders but i dont see it in the vids of the serves. They look like they have no seperation to me or not a lot. And maybe i missed it but what moves first? Is it the shoulders?? He said hips but i think that was to start the stroke and not create the separation????
Imagine viewing a server from above. Imagine a line joining the two shoulders. Imagine a second line joining the two hips. Viewed from above a server first turns the line of the shoulders back and the line of the hips back. Then turns them forward.
Turn Back. When the shoulder line rotates farther back than the line between the hips, the angle measured from above view is 'separation'. If the camera is located at the level of the player it is more difficult to see and estimate that angle than if the camera is above the server. But you can usually see typical maximum separation angles. Ask yourself if the shoulders are turned back farther than the hips. Look at yourself in the mirror with one eye (so that no stereoscopic from two eyes helps you).
Turn Forward. When the forward serve motion starts, the hips turn first keeping or increasing the separation created in the take back.
This separation also occurs especially for the forehand. Watch Dojokvic. He has done too much twist in the past so he is probably not a good model forehand to copy for many players
Illustration of shoulders-hips separation on the forehand. Fuzzy Yellow Balls overhead video.
Stretch Shorten Cycle. When there is an angle between the shoulder line and hip line, the upper body is rotated back farther than the hips. This lengthens and can stretch trunk muscles and can be used in a stretch shorten cycle. That's the end purpose, to use stretched muscles for more racket head speed.
Obviously, twisting the trunk and spine can be an injury risk if done too far, too often, too fast or if you have a back weakness or history of injury.
Yes, he's talking about the hips "moving first" when he's talking about the execution of the stroke -- not the preparation/loading part.
I certainly didn't watch the full video, but if these are the high-level servers you're referencing, here are the snapshots of the moments after they've driven the hip to generate separation from the shoulder line. For both, the hip is somewhere in the general neighborhood of 45 degrees relative to the baseline, and the shoulders are closer to 90. (There was a third vid, where the separation and video angle were both poor.)
Here, both servers are using pinpoint stances, and "step-through" mechanics to open the hips. When doing it that way, they begin the serve and toss the ball with their hips and shoulders pretty much in line with one another. It's only during that forward stride that they move the trailing leg into position such that driving up from the legs automatically extends the hips at an angle ahead of the shoulders. (As long as you didn't uncoil the shoulders prematurely.) This is the exact mechanical technique you previously used on your pinpoint serves.
When serving platform, that movement is taken away from you, and you have to create that separation during the takeback phase instead. That's what my response to you (the Becker/Sampras one) in the other thread was about -- generating this same source of power on serves where the feet don't move.
So, in the pinpoint (at least this more traditional step-through pinpoint) the hips "moving first" entails moving them into position ahead of the shoulder line, AND extending them by driving the legs. In the platform, it's just about extending them, and the "moving them into position" part gets taken care of differently (at the shoulders).
Incidentally, this is why I feel the step-through pinpoint has been unfairly blacklisted by modern coaches.
It's true that a pinpoint that doesn't move up the rear foot as far or across the body allows some more torque, if you get all of your lower body mechanics nailed perfectly. But at the top levels (7.0), we're talking about the difference between 100% effectiveness and 95% effectiveness (where "effectiveness" is some combo of max potential mph's and rpm's).
Amateurs are often dealing with serves that are around 50% in both departments -- an utterly failing grade. So often, they shouldn't be worried about "maximizing" so much as adopting a method that lets them get to a passing grade. I think the step-through pinpoint is the easiest full body serve to master, and remains a great choice at any level where you're not worried about whether or not you can take a set off of Federer or Nadal. It was perfectly adequate for Roscoe Tanner, Ivan Lendl, Stefan Edberg, Pat Rafter, Marat Safin, and thousands of others with excellent serve games. It still holds up today.
That's a very good explanation regarding difference between platform and pinpoint hip/shoulder separation! Makes sense to me but any coaches that have looked into this want to chime in? Perhaps @JohnYandell ?
My son has a platform serve and yes he creates the separation during the takeback (or coiling, loading, whatever you want to call it). His visual cue is to show his back to the net during this phase.
I am really having difficulty seeing the 45 degree separation angle from that pic -- hips and shoulder look pretty much in line . ..But as Chas mentions maybe overhead view is necessary to see it. But I definitely cannot even close to estimating separation angle from those pics.
Dunno. Especially on the guy, it stands out clear as day to me. Guy's dingus is practically facing the net. His nipples are straight at the side fence.
I thought TTPS was the only one who says this.
He stole it from me.
The overhead view is better, most accurate, especially for small separation angles, but not "necessary". I would guess the separation and is less than 45 d, 20 d girl, 30-40 Miryni,?? But it is best to avoid the issue.
Videos can be interpreted better than pictures. The time of the picture in the service motion has to be known.
I believe that as the camera is placed higher the estimate of separation would probably get better. Avoid estimating distances directly toward or away from the camera, Miryni's left hip and right hip move toward or away from the camera, least accurate. There are also a lot of angles in the serve and a general body rotation to make it more complicated in 3D. Camera view from behind would work better.
If interested you can get a racket, the same camera angle relative to the court, and pose in front of the camera. Give a thought to the effect of wide angle lens distortions......
Maybe you can reproduce the picture with a mirror and then measure the angle of your shoulders and hips. ? Use one eye?
Get an action figure that rotates at the waist and look at it. ?
He's talking about the kinetic chain. In that context, the same movement that starts the stroke creates separation between hips and shoulders. Players from the past, especially the women, were better at this technique than present players.
When I stand at the baseline, with feet and chest pointed staright to the net, it feels like I can rotate shoulder slightly more than 90 d away from the baseline and rotate hips about 45 d away from the baseline -- this is with the toes pointed straight towards the net. If I point both my toes towards right net post, I can get much more hip rotation -- 90 d away from the baseline.
Overhead view of Salazar in semi-open stance, back foot roughly parallel to baseline, front foot roughly pointed to right net post. I have drawn a line through the shoulders. but it is impossible to draw a line through the hips as both hips are are not clearly visible?, thus making even a rough estimate shoulder/hip separation estimate impossible... One cannot extrapolate the hip position from the feet position, because the hips can rotate significantly many degrees (more than ~30 d?) while maintaining the same foot stance.
Also, is Salazar's shoulder rotation anglebest measured relative to the baseline? Or relative to his semi-open feet position?
Best Angle Estimate and Uncertainty Estimate for Above Camera View
(Ignoring some accuracy issues, such as the effect of camera lens. Crude measurement by eye.)
The shoulders line looks clear, a very good estimate for the shoulder direction.
For overhead views I like to measure all angles from the baseline.
Uncertainly is always part of measurement. For example, we estimate/measure an angle to be 45 d with an uncertainly of +/- 15 d. In this case, we are estimating an angle as best we can and also roughly estimating our uncertainty using our eyes on a 2D video image.
Print the picture of Salazar's overhead forehand. Get a thin straight object like a needle. Salazar's body near his left hip is visible, consider the hip is somewhere near the middle of the upper leg. The right hip is obscured, but it is not impossible to estimate its location. But the uncertainty in knowing the location is much higher than for the left hip. Assume that his buttocks are about the same distance back from each hip, use the direction of a line across his butt? - it might even be a reasonable estimate for the hip line.??
1) Best Angle Estimate. Place the needle on the printed picture, hips direction as best you can, using middle of the left hip.
2) Uncertainty Estimates. Change needle direction away from best angle until you know its direction does not look right for the hips. Increase and decrease the angle a little until you are uncertain if it looks right. When you first decide the direction is not right - measure that angle with a protractor (from baseline). Now move the needle to the other side of the best angle and repeat.
We now have a best needle direction to represent the hips and two measurements roughly indicating uncertainty for this crude estimate.
Let's say 40 degrees looked best and as you move the needle to the right it looks reasonable to 28 d. but nor farther, 12 degrees uncertainty. As you move it to the left it looks reasonable until it reaches around 49 d. but not farther, 9 degrees uncertainty.
For this eye estimate from a 2 D video our 'measurement' might be stated
We think the angle is about 40 d +9/-12 degrees.
Now if you work with the Salazar picture for a while you will find that the measurement is accurate enough for your purpose or it is not.
Probably a less accurate estimate, like the one below, would not be very useful -
We think the angle is about 40 d +20 /-28 degrees.
My quick eye estimate -
The shoulder line looks very accurate, 1-2 d ?
The hip line estimate I guess would be +/- 5 d ?? Playing with the needle would tell us.
The uncertainty of measuring angles for the Miryni picture would be much larger because court lines & fence are not well placed and then for 'separation' those line directions have to be visualized from above. The uncertainty would be larger than for the Salazar picture because of the camera angle.
If your interest is in maximum angles - I don't understand the optimal angle for 'separation' for strokes under any circumstances. I have never believed that that maximum possible angle of any joint motion is best for tennis strokes.
Aim for 45 and 90. At least if you're using that sort of pinpoint stance. The idea and the execution matter. The precise measurements do not matter in the slightest sense.
That would be separation of 45 d, correct?
Where did you get that information?
I believe he is saying have the shoulder line rotated 90 degrees from the baseline and hips rotated 45 degrees from the baseline, as in the Miryni step up pinpoint position.
Maybe the ideal serve stance is an open one for ideal trunk coiling just like the open stance forehand. Kyrgios' hips and torso face the net at the beginning of the serve much more than say Federer. A.Zverev's is very similar. They seem to have a quick, brief 'separation' only at the trophy point.
Actually now after watching a few more slow mo videos of Kyrgios and Zverev I see that there is not much separation at all. They seem to have bloody sheer power gifted to them.
Boris Becker used a sligjtly open stance.
Are you sure?? He lines up like that but his front foot pivots and his feet end up parallel to the baseline in the vid Sinjin linked in my kick serve thread
About 45 degrees, for Kyrgios. Never analyzed Zverev, but did NK frame by frame.
Criminally easy to see in the clip below.
Nick Kyrgios - Slow Motion Serves - Indian Wells 2015
It looks like zero here:
It is useful to print a scale copy of the court. Draw a line from the Deuce court serving spot to the center of the service box. Do the same for the Ad side serve. Measure the angle between the two lines. The angle is about 21 d.
Probably the server maintains the same body angles relative to the ball's trajectory or aim points. The Add and Deuce trajectories have two different directions, 21 d different for the center spot trajectories. With angles referenced to the court lines, net or fences, the serve body angles would be different for the Deuce and Ad sides. Separation angles are probably the same for Deuce and Ad sides. In addition, the types of serves probably have some effect on the angles.
I recall reading one estimate for separation of 30 d but I can't remember if it was for the serve or forehand.
Recommend that readers don't simplify things down to a few words that are not described well enough.
Serve angles can be compared in high speed videos by using the same camera angle, serve type, Ad or Deuce side only, etc.. Timing should be coordinated somehow.
Even from that non-ideal camera angle i see shoulder/hip separation. look at the serves in the last third of the video clip.
Clearly you can see the hip rotating and shoulder lagging behind in rotation.
The time of the separation angle has to be specified somehow.
or the separation angle at full take back.
or the maximum separation angle during the stroke.
or separation angle vs time for the stroke.
The separation at full take back is one thing. The maximum separation during the stroke is another and may occur in the forward motion when the hips turn first, shoulders 'lag behind'.
Watch carefully to see the mostly obscured right leg move forward and cause the right hip to go forward and rotate the hips and cause separation. When the foot leaves the ground in jumping is not clear. Looks like a flat serve. ?
I have no idea what you're looking at.
At any rate, stop even LOOKING for gotcha counterexamples. I'm not trying to get people to "see something I might have spotted."
I'm joining others in passing along a well known, 100% accepted biomechanical fact that's taught in all high level throwing -- be it baseball, football, or tennis. To activate the big, powerful core rotation in this kind of serve/throw, separation between the angle of the hips and the angle of the shoulders MUST exist.
The angles are trivia. The understanding of the motion is essential.
I am sure we are looking at the same thing but probably seeing it a little differently. I honestly thought there was almost no separation in Kyrgios' serve, I admit camera angle is not ideal. Anyway didn't mean to annoy you.
Anyone else who cannot see much of a separation even here?:
Agree, not much separation in that video.
But those 2009 serves are at a practice court and there is no way to tell if he is just starting to warm up or that serve was representative of his match serve. ?
Also, in 2013 Federer injured his back and was two? years recovering. Federer serve video from 2013-2015 or 2016 may involve modifications for that back injury. He recently discussed his back injury in an interview, on Tennis Channel I believe.
Find a Federer match video from before 2013.
It would be best to have server stats on separation being used by ATP and WTA servers as a standard. Stats- Max serve separation vs % of players in that separation angle range. For example, 12% of ATP servers have 30-40 d. separation for flat serves, etc.
The amount of separation for forehands and serves have probably been measured and published. My searches have not found any yet. 3D motion capture systems are not bothered by the 2D limitation of single cameras that we have been discussing.
If you stand in front of a camera and rotate there will be a signature for each hip and shoulder angle. That signature can be calibrated in degrees. You can get a pretty good estimate better than eye ball. But another problem comes in as the serve is not simply standing straight up and turning but also involves body tilts and trunk bends. If you have 2 cameras you can do 3D measurements.
If you had a high level serving technique, I'd guess you should have a typical separation angle to start that is safe for you and then tweek it.
When you turn slowly to cause separation angle I believe that is called voluntary range of motion and is based on your antagonist muscles providing stretch forces. Stretch may go farther due to other body motions. Risk of injury?
I'd guess that the separation in the take back has to be less than the max separation when the hips turn first.
Often when separation is spoken of I have thought that the separation at full take back was being discussed and not maximum momentary separation in the forward swing. ?
Careful with trunk/spine twisting motions.
Fair enough. At least I don't need an eye check then.
Here is one from 2008 and match play. Again the camera angle is not great, but the separation angle looks like much smaller than 45 degrees to me.
To clarify, I was making no estimate of the the shoulder-hip separation angle of Miryni pic. That 45 seperation was based on previous Sinjin's post stating a 45 degree separation in Miryni pic (relative to baseline)... Just trying to get a general notion of how to spot and make rough estimate of separation. Not interested in any particular angle or "maximum angle"....Also, I had thought we are relating the separation angle to the baseline. I do see the 21 degree angle mentioned but not clear how the path of the ball relates to measuring shoulder-hip separation.
I would think the best way to measure separation would be relative to stance, independent of baseline. If one was facing the net in a fully open stance, toes and chest pointing to net we would get a separation of 0, if one were in a semi-open stance, like the Salazar forehand pic, we could measure separation angle from that semi-open stance. With that method, we would draw a shoulder line and a line between the toes or mid foot.
At 12 seconds, Trophy Position, I see the shoulders back farther than the hips. 30 d?
I used the center of the service box (draw and "X" between the corners)
To measure use a reference in the picture. From above, the baseline is perfect for measuring shoulders line and hips line. I then subtract hip line angle from the shoulder line angle to get separation. In reality, I am eye estimating this because I'm not after the most accurate measurement. From ground level videos, we use our eye and brain to estimate angles as they would be seen from above, angle between the shoulder line and baseline, for example, is not always clear, an inaccurate process. I suggested that you use a needle on the Salazar forehand picture so that you would understand the process and how uncertainty comes into the estimate. When I look at Miryni I can estimate angles but the uncertainty would be large. = pick a better picture to make a ground level angle estimate - camera behind, other....
I pointed out some problems with simply saying "Aim for 45 and 90."
Trajectory of the Ball. To hit the ball is some direction, I believe that the server's body orients itself to the ball's intended trajectory or equivalently, the aim point. You seem to agree. But in the Tennis Serve Nuthouse the trajectory is almost always ignored and angles are commonly discussed with reference to the baseline, fence, etc. This process muddles understanding and leads to endless threads on what is meant. It breaks down when considering the same type serves to the Ad & Deuce courts.
I employ a platform stance,and can clearly physically feel the separation on the take back. I believe this separation occurs on every platform serve (I am not familiar with pinpoint but Sinjin points out this separation occurs after the step forward on pinpoint). I presume other posters are not disputing this separation physically occurs, as in the Federer pics, but only how to recognize it from pics.
For platform stance, I was interpreting "separation angle" as relevant to the start of the motion(around trophy position or racquet drop). I do not agree my starting stance changes based on wide or T serves. The degree of separation may be different at contact based on serve location, but I was considering the separation angle at the start of the forward motion.
Coaches have told me there is only a 1-2 degree of racquet tilt difference between wide and T serves, so I do not think serve location affects the stance.
I don't know about your serve's starting stance. I never assume unknown serve techniques are the same as high level serves. Other techniques are always interesting.
I assume high levels servers mostly orient to the intended ball's trajectory much more than they do to the court lines. But I don't know of publications with measurements.
I don't know what the ATP servers are doing to orient their racket faces. But the widest serve to the Duece court sideline is about 35 d degrees away from the widest serve to the Ad court sideline. For side to side aiming, the servers are using some combination of
1) body orientation,
2) stroke path and
3) azimuthal racket face angle (side-to-side) at impact, largely from ISR angle at impact.
I assume separation is mainly used for the stretch shorten cycle to stretch trunk muscles, as the OP's video says. Racket head speed could drop if less separation is used for the purpose of adjusting aim, but maybe it's done. ?
For the "1-2 degrees of racket tilt" statement, do you mean the elevation angle, how closed the racket face is?
I like the medicine ball toss drill. Great analogous drill to develop kinetic chain during serve.
Is this any better? Shot at Hopman Cup 2017.
Shoulder/hip separation is fairly subtle, unless a player has a very noticeable one only a top-down view can really display it very clearly.
Shoulders appear slightly more rotated than hips, and if I didn't know better I'd say he's practiced the JS elbow the enemy drill!! Some pinpoint servers, particularly on first serves, use more of a seesaw motion and less rotation.
Interesting article comparing the two here:
Not sure I completely agree with everything written there, because it seems to me Kyrgios's shoulder-shoulder-elbow alignment isn't a perfect straight line (although angles can be deceptive), but the jist of it does seem to make sense (if the stats are accurate).
Separation is discussed in these reference books -
Technique Development for Tennis Stroke Production, 2009, B. Elliott, M. Reid. M. Crespo
It says the shoulders turn back typically 20-30 degrees farther than the hips, separation. pg 77. Mentions variety. Mentions how much youth does.
The Biomechanics of Advanced Tennis, 2003, B. Elliott, M. Reid, M. Crespo,
It discusses the purpose of separation on the serve. pg 35.
This important second reference appears to be available free from Researchgate. Link does not work but gets a short article.
I do not agree about NK's reliance on the shoulder tilt (aka as seesaw, cartwheeling, angled spine, lateral flexion/extension, or shoulder over shoulder) for first serves.
There are three relevant motions involved, as shown in the pic below. These motions are often occurring simultaneously but players can rely more on one than another.
I think taller players like NK can get away with much less shoulder tilt. He relies much more on the hip rotation to generate RHS on first serves.
See also Sinjin's comments on NK's hip rotation.
The coach tilted racquet left and right and said just a few degrees is the difference between T and Wide serve and there was no need to orient one's starting stance differently for various placements.
The racquet is also simultaneously closing but do not think this affects placement -- suspect the closing only affects degree of spin. The ball is on the strings for about 3 milliseconds, at the moment of contact the face is roughly vertical to the court, and the slight left/right tilt differences at that moment of contact determines placement.
Another thread (since deleted) with slow motion TP.net footage showed that on topspin serves the dwell time may increase to 5 milliseconds, but the ball does not roll across the strings -- it simply pockets differently. So I conclude that the left/right racquet tilt at moment of contact is what determines placement.
I'm not getting a clear picture of what your coach was thinking or what is meant by racket 'tilt'. ? Up-down, side-to-side or a combination somehow?
It is true that in serve elevation a very small angle of closure determines if the ball is in the net or long and misses the service box. Probably excluding topspin and kick serves. That has been measured and published. Search projection angles serve
I don't know of good information giving racket face azimuth angles for the side-by-side aim. No conclusions.
I feel that if the service box is 10 d wide then something like 10 degrees racket face angle side-to side variation might be associated with that ball placement. I would not conclude anything but that would be a first estimate.
Side to side. Left/Right. Not Up/Down or open/closed.
He said just a few degrees difference between T and Wide.
Dont know the number but it physically feels very small. 5-10 degrees sounds reasonable.
But my main point is that the placement can be manipulated subtly with wrist very near contact. It is not necessary to employ a different starting stance to vary the placement.
I am mostly trying to understand high level strokes and their variations by finding creditable research and analyzing high speed videos. Have hardly scratched the surface of the important motions in a few strokes. Mostly I've studied how stokes develop racket head speed....
Separation is a very interesting topic for racket head speed.
For my interests the best information that I've found so far has been in The Biomechanics of Advanced Tennis, 2003 and the Technique Development for Tennis Stroke Production, 2009. Both by B. Elliott, M. Reid, M Crespo. Best discussions of separation emphasizing its muscle stretching purpose used in the stretch shorten cycle, same purpose as in the OP video.
The question about what is 'necessary' or 'needed' is often very hard to answer. I often don't understand the question - Is XYZ 'necessary' or 'needed'?............ For what? I'm interested for now in what the high level ATP and WTA players are doing, their serves are the best answer I know of.
I don't know what they are doing for side to side racket control on the flat and slice serves. But for the kick serve I believe side to side direction control is why the server's upper body faces more to the side at impact.
IIRC, Federer had a lower back issue.
I don't believe lower back injury would significantly affect lateral flexion (B) or core upper trunk rotation (C) and cause any reduction in resultant hip shoulder separation angle. It would mainly affect back arching/extension (A). I believe a shoulder hip separation angle can be created on a flat serve without employing (A) (or very little of it).
A player may choose to momentarily increase hip-shoulder separation angle on the forward swing, e.g. by leading with the hip. Or, he may choose to move forward
with hips and shoulders in synch. This was discussed in another thread and some posters mentioned they move forward in unison rather than momentarily increasing the seperation angle as they suspected Federer might be doing on the forehand..I think the separation at full take back is the most useful point of discussion since the momentary increase going forward may not be employed in all cases and also will be more difficult to see.
Roger Federer has given a detailed description of the back injuries he suffered last year and the measures he is now taking to avoid more problems.
Federer said it first happened at last year's Indian Wells tournament, in the second round, on a second serve.
This is another very good video
Separate names with a comma.