Katrina Adams in 15-30

[johnyandell]

My take is that the arch is a consequence and not a deliberate movement.

[GeoffWilliams]

It exists whether deliberate or not. MRI scans show people given a knot to knit, that they already know how to do, show a tiny red dot lit up in the primitive motion area of the brain, and when given a new knot, that they don't know how to do, both lobes light up all over, until they learn it, and then back down to a tiny red dot again. The cal coach once told me, "We don't want them thinking too much out there.", and that was a deeper statement than he meant it to be!

We all strive to be tiny red dots. Zone dots/without thought or tactical blunderings.

When I don't do it, my serve does not have any weight or pop. When I do, sometimes it does have pop and weight, but only if I have a fast snap back that causes my hand to increase in weight greatly. How can something that looks so simple be so difficult to replicate? Only Rajeev Ram has done it. They call him Rampras.

You can see in frame 6-7-8, black and white above, his back is arched over sideways, giving him a farther forward contact point, same as becker used to do, and allowing greater kinetic pathway for the frame as it moves farther forward. The whole motion is like a circle, with his back starting curved forwards hunched over forwards, and ending in that sideways crunch with hip over hip rotation.

[DougEng]

My take is that the arch is a consequence and not a deliberate movement.

Agree. It is a matter of kinetic chain order, balance and gravity. The greater the tilt (knee bend) the greater the lumbar extension. But it is normally limited to about 10-15˚. There is also hip extension which makes the back look like it's bending more.

During the upwards leg drive, the kinetic chain works from the bottom first, hence the racquet drops into the slot (cocking) as the legs straighten...almost appearing to move in opposite directions. During this motion, maximal lumbar and hip extension occur, combined perhaps 20-25˚ typically with good servers. May be as higher as 30˚ with hyperflexible servers (Djokovic?).

Doug

[DougEng]

sampras serve ARched: frame 4-5-6-7-8

I would say this is mobility mostly at the hip region where lumbar and hip extension occur together, perhaps, 25-30˚. Most people visually assume hip extension is back extension. Above the lumber region, it might only be 0-5˚ in thoracic (maybe a bit more with Djokovic) and 30˚ in cervical (neck, head tilts back) but those last two won't be significant contributors in the kinetic chain.

Anyhow I can check with the ATP Tour medical staff. But I'm sure Todd would agree.

[DougEng]

It exists whether deliberate or not. MRI scans show people given a knot to knit, that they already know how to do, show a tiny red dot lit up in the primitive motion area of the brain, and when given a new knot, that they don't know how to do, both lobes light up all over, until they learn it, and then back down to a tiny red dot again. The cal coach once told me, "We don't want them thinking too much out there.", and that was a deeper statement than he meant it to be!

We all strive to be tiny red dots. Zone dots/without thought or tactical blunderings.

When I don't do it, my serve does not have any weight or pop. When I do, sometimes it does have pop and weight, but only if I have a fast snap back that causes my hand to increase in weight greatly. How can something that looks so simple be so difficult to replicate? Only Rajeev Ram has done it. They call him Rampras.

You can see in frame 6-7-8, black and white above, his back is arched over sideways, giving him a farther forward contact point, same as becker used to do, and allowing greater kinetic pathway for the frame as it moves farther forward. The whole motion is like a circle, with his back starting curved forwards hunched over forwards, and ending in that sideways crunch with hip over hip rotation.

In frames 6-7-8, I would not call that back arch as traditionally defined. It is a side bending of the thoracic perhaps 15-20˚. Arching is commonly seen as backwards (extension).

Maximal lumbar and hip extension is in frame 4 (ca 25˚).

However, again this is not the arch commonly taught or what Katrina mentions. She refers to right after the toss where it is even less. Perhaps 10˚ of back extension and another 10˚ of hip extension (20˚ at most). In the late cocking stage, side bending and another 5+˚ back extension occur as the upper body is delayed as energy is transferred from the lower to the upper body.

Hope this makes sense.

[stotty]

Hip extension

Interesting this hip extension stuff Doug mentions. I had never heard of it before. I assumed any arching was due solely to back extension.

I hadn't considered side bending of the thorax before either.

Great of Geoff to post this wonderful seqeunce of photos.

[stotty]

Role of the wrist?

Doug,

How does the wrist figure in the kick serve? Is it passive..."along for the ride"...or is it pumping a little more on a kick serve? When I do a kick serve, I feel like the wrist is pumping (actually every part of my action feels like it's working more) a little more compared to flat or slice...but then I don't have a good kick serve...so how should it be? Is it better not to think of the wrist at all and just let it do its job, assuming all the other events in the chain have been correctly met?

Very curious about this one...

Do you have any thoughts on this one also, 10splayer?

[gsheiner]

Here's my question. Whenever I see those overhead views of Sampras's serve, I am always struck by how far he seems to pull his right elbow back as he loads into the trophy position.

I've read stories about how flexible Pete was, but is this degree of flexibility commonly measured and compared between different servers? Is this an area that can be improved with tennis specific exercises?

[GeoffWilliams]

So do pitchers. Internal shoulder rotation. Only by focused practice.

[10splayer]

sampras serve ARched: frame 4-5-6-7-8

I agree that internal shoulder rotation should be a major contributor to RHS. Problem is (at least on the subject of lateral tilt) when a player tilts too much, IRS contribution is reduced.

Take a look at Brian G's article on the serve, specifically "the upward swing" part one. Brian goes to great lengths to describe how the angle of torso, and how it rotates (very complex) has an enormous impact on ideal hitting arm postioins at impact..Which in turn dictates what, and how much each joint contributes to RHS.

The premise being that, a little tilt is good, but too much REDUCES IRS contributions. Fed is a nice model in this regard. Notice the angle of his torso, as evidently he is able to employ MUCHO internal shoulder rotation with these body lines and arm configurations.

That ariticle is absolutely phenomenal.

[DougEng]

Here's my question. Whenever I see those overhead views of Sampras's serve, I am always struck by how far he seems to pull his right elbow back as he loads into the trophy position.

I've read stories about how flexible Pete was, but is this degree of flexibility commonly measured and compared between different servers? Is this an area that can be improved with tennis specific exercises?

Just to pass on from USTA medical experts I get to know and spend time with:

Turn the core, not the elbow. Any medical expert would cringe as rotating the elbow back, particularly when the elbow is high or close to in line with shoulder clavicle. That would place the shoulder joint at high risk for injury.

It looks like about Sampras is only 10˚ behind the shoulder line, so it's not that bad. But it is possible other players will injury the shoulder joint. The joint would be even at higher risk in the late cocking late if the elbow remains in this hyperextended position behind the player with the racquet butt pointing to the sky.

Keep in mind, flexibility is not the sole desired element. It's strength of joints at the end of the range of motion (ROM).

[DougEng]

Doug,

How does the wrist figure in the kick serve? Is it passive..."along for the ride"...or is it pumping a little more on a kick serve? When I do a kick serve, I feel like the wrist is pumping (actually every part of my action feels like it's working more) a little more compared to flat or slice...but then I don't have a good kick serve...so how should it be? Is it better not to think of the wrist at all and just let it do its job, assuming all the other events in the chain have been correctly met?

Very curious about this one...

Do you have any thoughts on this one also, 10splayer?

Hi Stotty,

According to current research, the wrist does contribute. Depending on the study, this could be anywhere from let's say 3-5 mph to probably up to 33-35 mph of racquet head speed (on a 150 mph serve). But please read on...

Bruce Elliott is actually considered the guru on biomechanics of the serve.
Ben Kibler is currently doing a research project that involves about 300-400 WTA and ATP players, mostly on medical implications of biomechanics.
Very much of what I say is actually from them.

Bruce had 2 studies, one about 1990 and the second I think in 1996 or 1997. The latter suggested internal shoulder rotation is the primary factor for racquet head speed. The following URL is based on the latter study.

http://www.coachesinfo.com/index.php?option=com_content&view=article&id=347:tennis-theserve-article&catid=95:tennis-general-articles&Itemid=173

There is a chart in this report.
Note in this last paper, Elliott suggests 30% of racquet head speed is due to hand (wrist) flexion. 10-20% from leg drive and trunk rotation, 10% from upper arm (shoulder to elbow) elevation and flexion. 40% from pronation, forearm extension, upper arm internal rotation. This number changes depending on the study. The original study in 1990, I believe said none with wrist flexion.

Depending on the study, it may be from 0-30% from hand flexion. Keep in mind, most of this racquet head speed is when the racquet is already moving at probably 40-50 mph (for a 120 mph serve). That is most of the kinetic chain has already fired and the wrist flexion is the last link. At this point, the racquet is relatively "weightless" (as momentum already has the racquet moving) and the wrist adds the remaining 0-30 mph (the other parts of the arm are also contributing perhaps 20 mph in the final stage). When the racquet reaches shoulder height, it is moving at maybe at 40-50 mph as the shoulder is partially contributing. So this final acceleration, forward internal shoulder rotation and wrist flexion account for another 40 mph. That is, I believe, typically a 120 mph has racquet head speed around 85-95 mph, it depends on the type of spin, etc.

Much of this is inexact since it does depend on the server. And racquet head speed should not be confused with power. For example, in the serve, probably 50-60% of power is from the leg drive (which contributes to 10-20% of racquet head speed).

It is likely that some serves only rely on 5-10 mph of wrist flexion for racquet head speed. Others may have 25 mph. Perhaps those serves are slower? There are no studies differentiating different servers, biomechanics and results.

Vic Braden used to have a fellow teaching pro sit on the knees and serve maybe 110 mph. That shows that although leg drive is important, it is not the primary source of velocity.

Personally I find it hard to see a serve relying on wrist flexion for racquet head speed. As mentioned it is the end of the kinetic chain. That is like a whip. The tip of the whip moves at perhaps 200 mph. But the hand which initiates the whip might only move at less than 40 mph.

Also keep in mind, that moving the racquet with wrist flexion is different from 0-10 mph than it is from 40-55 mph. Same as pushing a car if you ever had a car broken down. Try pushing the car if it is standing still. It takes considerable more force than pushing it when it is already moving 5 mph. That is because you are overcoming frictional forces and change inertia. Or like being on a swing. The first few swings are used to overcome inertia; when you get moving, you get moving. But not like the snowball effect, that uses gravity which does not apply here.

As the few studies on serve biomechanics do vary (although all say internal shoulder rotation is the primary generator), Elliott came out with a better description since many of the shoulder to arm links occur concurrently and different parts moving different ways. For example, the shoulder has internal rotation but also upwards elevation. The elbow pronates but also extends.
Elliott suggested around 2001 to utilize "long axis rotation" as the better phrase for description as a series of complex coordination body parts are moving in synergy. Unfortunately, this wording has not gained much popularity perhaps because it is ambiguous. It's sort of like saying I'm watching a movie about a fish. Rather than saying that, we'd rather quote "You're gonna need a bigger boat." Then everyone knows what you are talking about. Perhaps that is why we say internal shoulder rotation OR pronation OR wrist snap. One movement only rather than the whole thing. Give a quote to summarize the movie. But biomechanically, it is preferred and acknowledges that the arm segments cannot be truly isolated in their effects. So the whole script better describes the synergy.

Here is another recent study (summary) by Elliott. Stotty, it also mentioned Duane Knudson whose book I recommended to you (basic but quite readable) and Rafael Bahamonde (great biomechanist in Puerto Rico, actually one of Brian Gordon's former advisors).

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2577481/

[GeoffWilliams]

I'm going to need a bigger boat.

[stotty]

Percentages

Hi Stotty,

According to current research, the wrist does contribute. Depending on the study, this could be anywhere from let's say 3-5 mph to probably up to 33-35 mph of racquet head speed (on a 150 mph serve). But please read on...

Bruce Elliott is actually considered the guru on biomechanics of the serve.
Ben Kibler is currently doing a research project that involves about 300-400 WTA and ATP players, mostly on medical implications of biomechanics.
Very much of what I say is actually from them.

Bruce had 2 studies, one about 1990 and the second I think in 1996 or 1997. The latter suggested internal shoulder rotation is the primary factor for racquet head speed. The following URL is based on the latter study.

http://www.coachesinfo.com/index.php?option=com_content&view=article&id=347:tennis-theserve-article&catid=95:tennis-general-articles&Itemid=173

There is a chart in this report.
Note in this last paper, Elliott suggests 30% of racquet head speed is due to hand (wrist) flexion. 10-20% from leg drive and trunk rotation, 10% from upper arm (shoulder to elbow) elevation and flexion. 40% from pronation, forearm extension, upper arm internal rotation. This number changes depending on the study. The original study in 1990, I believe said none with wrist flexion.

Depending on the study, it may be from 0-30% from hand flexion. Keep in mind, most of this racquet head speed is when the racquet is already moving at probably 40-50 mph (for a 120 mph serve). That is most of the kinetic chain has already fired and the wrist flexion is the last link. At this point, the racquet is relatively "weightless" (as momentum already has the racquet moving) and the wrist adds the remaining 0-30 mph (the other parts of the arm are also contributing perhaps 20 mph in the final stage). When the racquet reaches shoulder height, it is moving at maybe at 40-50 mph as the shoulder is partially contributing. So this final acceleration, forward internal shoulder rotation and wrist flexion account for another 40 mph. That is, I believe, typically a 120 mph has racquet head speed around 85-95 mph, it depends on the type of spin, etc.

Much of this is inexact since it does depend on the server. And racquet head speed should not be confused with power. For example, in the serve, probably 50-60% of power is from the leg drive (which contributes to 10-20% of racquet head speed).

It is likely that some serves only rely on 5-10 mph of wrist flexion for racquet head speed. Others may have 25 mph. Perhaps those serves are slower? There are no studies differentiating different servers, biomechanics and results.

Vic Braden used to have a fellow teaching pro sit on the knees and serve maybe 110 mph. That shows that although leg drive is important, it is not the primary source of velocity.

Personally I find it hard to see a serve relying on wrist flexion for racquet head speed. As mentioned it is the end of the kinetic chain. That is like a whip. The tip of the whip moves at perhaps 200 mph. But the hand which initiates the whip might only move at less than 40 mph.

Also keep in mind, that moving the racquet with wrist flexion is different from 0-10 mph than it is from 40-55 mph. Same as pushing a car if you ever had a car broken down. Try pushing the car if it is standing still. It takes considerable more force than pushing it when it is already moving 5 mph. That is because you are overcoming frictional forces and change inertia. Or like being on a swing. The first few swings are used to overcome inertia; when you get moving, you get moving. But not like the snowball effect, that uses gravity which does not apply here.

As the few studies on serve biomechanics do vary (although all say internal shoulder rotation is the primary generator), Elliott came out with a better description since many of the shoulder to arm links occur concurrently and different parts moving different ways. For example, the shoulder has internal rotation but also upwards elevation. The elbow pronates but also extends.
Elliott suggested around 2001 to utilize "long axis rotation" as the better phrase for description as a series of complex coordination body parts are moving in synergy. Unfortunately, this wording has not gained much popularity perhaps because it is ambiguous. It's sort of like saying I'm watching a movie about a fish. Rather than saying that, we'd rather quote "You're gonna need a bigger boat." Then everyone knows what you are talking about. Perhaps that is why we say internal shoulder rotation OR pronation OR wrist snap. One movement only rather than the whole thing. Give a quote to summarize the movie. But biomechanically, it is preferred and acknowledges that the arm segments cannot be truly isolated in their effects. So the whole script better describes the synergy.

Here is another recent study (summary) by Elliott. Stotty, it also mentioned Duane Knudson whose book I recommended to you (basic but quite readable) and Rafael Bahamonde (great biomechanist in Puerto Rico, actually one of Brian Gordon's former advisors).

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2577481/

Thanks for the comprehensive follow up to my post. I haven't read the Bruce Elliot and Duane Knudson articles yet, but will avidly.

Speaking from personal experience, I feel the bulk of my power comes from the upper arm and shoulder...virtually all of it. There is a link to me serving somewhere in the forum but I couldn't find it. Nevertheless, as you mentioned, all servers are different. I use very little leg drive so it would make sense the feeling of power comes mostly from the upper arm and shoulder.

Pro serves of course are a whole lot different. Every part of the chain is being used to generate power, and it's interesting to see how and where the power is coming from. The wrist is the biggest mystery in the serve, for me...and the only way to judge it is from personal experience...but it's so hard to quantify it even in your own serve. The wrist feels like it is doing very little when I serve...except perhaps in a kick serve, where it feels it's doing a little more. But I only serve at 85-90mph and have no conception of what it feels like to deliver one at 130mph...let alone the 150mph barrier which some pro's can cross.

Pro's don't always get all parts of the motion right either, though. Below is a clip of Ryan Sweeting, which I took a few years ago at Wimbledon - where he was routed by Nadal. His serve is a mess...the elbow is so low and he struggles to compensate for this error during the second part of the action. It's scary how he could have passed through so many coaches' hands and not have such a basic fault rectified.

http://www.youtube.com/watch?v=UsT2mqnQrQ0

This is a great thread. I've learned a good deal...things I didn't know. Interesting how the input from the wrist may be higher or lower depending on the server....with pro servers, one would think it would be roughly the same.

I just wonder how such quantitive measurements are made? How do biomechanists come up with these percentages? It would seem impossible without sophisticated devices of some kind?

[tennis_chiro]

Need a smaller boat with a longer, lighter paddle!!!

I am constantly amazed and impressed by the depth and breathe of knowledge that Doug Eng brings to the forum. I consider myself pretty good and try to educate myself about what is being learned about the game, but I always come away feeling woefully inadequate when I see the information that Doug puts up here. I'm so glad that he is participating a little more regularly now. I try to go through some of these scientific presentations, but it's really not that easy. I really appreciate some of the simplifications and clarifications that Doug brings to us here. I hope you keep it up, Doug.

I will try to get through these journal articles and see what I can cram into my thick skull (seems to be awful leaky these days), but I do have a few questions that some of these power contribution breakdowns raise for me. I may not be able to handle a bigger boat, but I could certainly use a longer, lighter paddle to help me navigate this boat along this blankety-blank creek of knowledge!

First of all, about the leg drive.
There is no question that the legs make a significant contribution to the power of the serve. But I would argue the 10 to 20% of power that comes from leg thrust and trunk rotation comes primarily from trunk rotation. I used to be able to serve pretty well standing on one leg (as a demonstration); but while I stayed balanced and my knees bent hardly at all, I would never have thought of restricting the rotation of my hips or of my shoulders from sideways to facing the net as I did that. I see a tremendous problem with kids trying to leave the ground or at least use as much upward leg thrust as possible when they have not yet truly mastered trunk rotation leading them up to a solid contact with the ball.

In fact I'm one of those coaches who tries to get my students to serve with no step and simple balance

http://www.tennisplayer.net/bulletin/showthread.php?p=20687#post20687

before I let them move on to leaving the ground. I may just be too wedded to the old school serve and volley principles, but even though that rear foot kicks back up into the air, its first contact with the ground should be well in front of the other foot. This is true even for a committed groundstoker like Agassi. My first movement in the direction of the kickback is a simple rotation on the rear toe so the right knee (for righties) lets the right hip get forward.

Sure baseball is limited by the rule that the toe has to stay on the "rubber" until the pitcher releases the ball, but quarterbacks have no such limitation. And we don't see them trying to leave the ground to get more power into their throws. Better yet, take any of your good players and stand next to them and feed them some high lobs for bounce overheads a few feet behind the service line. I have yet to find a player who hits the ball better in that situation by leaving the ground. On the contrary, I try to use the feeling they develop hitting that ball to get the sense of using their legs for power in their serve with stability and balance that leads to consistency and accuracy as well as power. You can use the ground to drive the trunk rotation that provides the real leg power to that overhead.

I'm getting carried away as usual, but my point is there needs to be a better understanding that the "10 to 20%" from leg drive and trunk rotation is primarily trunk rotation (I have no scientific study or biomechanics data to corroborate what I am saying here so feel free to prove me wrong) and most of the kids I see are sacrificing 10 to 20% of their overall power because of the compromise to their kinetic efficiency created by overemphasizing knee bend before they have true command of the basic upward strike. Oh yeah, they might be getting 3-5% more power because of their excellent knee bend; on top of that they can't get the ball in the box because the motion is too complicated for them to repeat consistently. NO, I don't think 60 to 65% is a good first serve percentage; especially for anyone over 6' 2" tall, much less 6' 4".

Second, the idea about it being easier to go faster at 50mph than 10mph. Recognize the difference between static and sliding friction. It's much easier to keep something moving once you get it started moving, but from a physics point of view, it is not going to be easier to accelerate something from 40 to 50 mph than it is to accelerate it from 10 to 20 mph. Percentage wise, it may be a smaller jump (25% vs 100%), but the amount of work is actually more. In terms of the forces on the racket, perhaps the coefficient of sliding friction for the racket through the air molecules my be slightly less (I don't know), but I do know that the wind resistance is proportional to the square of the velocity of the racket. That's one reason your gas mileage goes down as you get your velocity goes up. So no, you don't want the racket head to come to a dead stop, but it doesn't make sense to me that it is any easier to accelerate the racket head from 50 to 80 mph than it is to accelerate it from 20 to 50 mph.

Finally, all this emphasis is always on how to generate more power. But at what cost. The stroke has to be produced in a manner that enables consistency and accuracy under sometimes very difficult conditions. Big, thick, heavily muscled bodies might be able to generate a little more power but those players would then have to schlep those muscles with them wherever they went. Physical power has to be balanced and matched by flexibility, speed and endurance. (What a wonderful game!!) In a different way, powerful strokes (or service motions including the toss) have to include the ability to generate tremendous levels of consistency and accuracy. Where does the "flip" or the Type III vs the Type II or Type I forehand backswing fit with regard to that perspective? Could an old timer without much of a flip hold up with his greater consistency against the heavy ball of a modern player with a "flip" and the concomitant increased spin and velocity. I doubt it, but you might get a different answer from some to the NorCal players trying to handle Karsten Popp.

http://www.tennisplayer.net/members/classiclessons/scott_murphy/traditional_forehand/

Looking forward to getting ripped to shreds here. Always good to learn. Maybe it will make it easier for me to accept that the kids don't take everything I say as gospel. I keep thinking if I had only had someone with my knowledge to tell me what to do, it would have been so much easier and I could have been so much better. Probably would have still been too slow, but what fun it would have been to have a reliable forehand, much less a weapon instead of a gross liability. After all, you know the reason Hop named Rod Laver "the rocket". He was the slowest in the team wind sprints! At least that was the story I remember.

don