New Roscoe Tanner Serve Video.

Finally a clear Roscoe video.

He makes contact at the peak of the toss. 0 inches drop before contact. At full extension. He seems to hit all the checkpoints and gets into a proper trophy position.

Proper shoulder over shoulder motion? Thoracic Extension? We think so.

So minimalist and yet so powerful. Does one need to have super powers (e.g. live arm) to pull this motion off or can average players copy it?

Really don't see why the toss needs to drop more than 12 inches.

There is a new Russian WTA player whose toss drops like 3 feet before contact. Really hate seeing that. Totally unnecessary.


giphy.gif
 
Last edited:
Why? Consistancy. Hitting your flat serve with a dropping toss increases consistancy about 10% compared to hitting at the peak (due to induced topspin).
 
Finally a clear Roscoe video.

He makes contact at the peak of the toss. 0 inches drop before contact. At full extension. He seems to hit all the checkpoints and gets into a proper trophy position.

Proper shoulder over shoulder motion? Thoracic Extension? We think so.

So minimalist and yet so powerful. Does one need to have super powers (e.g. live arm) to pull this motion off or can average players copy it?

Really don't see why the toss needs to drop more than 12 inches.

There is a new Russian WTA player whose toss drops like 3 feet before contact. Really hate seeing that. Totally unnecessary.


giphy.gif
interesting, i play dubs with a few ex-atp guys in their 60's and 70's and they have similar serve motions (lowish toss, similar kind of pinpoint movement, almost as if staying low/in contact with ground).
originally i thought it was just them protecting their back/knees/etc..., but now i wonder if that was the serve motion of the day

also, at first glance, i was thinking, roscoe was serving like this due to the rule of "must keep one foot on the ground"
but that rule was in effect only until 1959, ... roscoe played in the 70's
did this style of serving (appears roscoe is keeping one toe on the ground), due to emulating rosewall (who was playing during the one-foot-on-ground rule).

personally i only really started watching tennis in the 80's, and by then everyone was "jumping" on their serve...

thoughts anyone?
 
i was thinking, roscoe was serving like this due to the rule of "must keep one foot on the ground"
but that rule was in effect only until 1959

It is confusing. To my eye, both of his feet lift an inch off the ground, albeit very briefly. But he is "jumping".
And then Roscoe lands on the back foot, just like Becker did.
That "one foot on the ground" rule looks like a pain in the ass to enforce... The "one foot" rule was lifted in 1959. So is that when jumping was legalized.
:unsure:

Also note that Sharapova's ridiculously high toss was a result of her emulating her hero Graf.
So unfortunately technique often gets propogated for no good reason.
 
Last edited:
Would love to hear from @Serve Doc about pros and cons of this style. Thanks.
Channelling my inner Chas for a moment, take a look at the top 100 (or 50; whatever sample size floats your boat) and categorize toss heights based on low, regulation, and high (you can use any scale you want). For instance, I would say a low toss is anything that drops less than 1', regulation is between a 1' and 2.5' drop, and high is anything dropping over 2.5'. Then check the numbers to see how many top players have a low toss. That's a lot of work but such things have been done before.

Imo hitting at or very near the peak, while delivering an effective serve with such a fast motion, is almost never going to be a viable option at the recreational level. I am not advocating for a high toss either, but it needs to be high enough to allow the body to load correctly in order to execute a powerful serve.
 
Channelling my inner Chas for a moment, take a look at the top 100 (or 50; whatever sample size floats your boat) and categorize toss heights based on low, regulation, and high (you can use any scale you want). For instance, I would say a low toss is anything that drops less than 1', regulation is between a 1' and 2.5' drop, and high is anything dropping over 2.5'. Then check the numbers to see how many top players have a low toss. That's a lot of work but such things have been done before.

Imo hitting at or very near the peak, while delivering an effective serve with such a fast motion, is almost never going to be a viable option at the recreational level. I am not advocating for a high toss either, but it needs to be high enough to allow the body to load correctly in order to execute a powerful serve.
With your usual rhythm it’s hard to serve like that. Hitting arm action starts much earlier with a low toss. Still not easy in my experience. Lower contact also means higher risk.
 
With your usual rhythm it’s hard to serve like that. Hitting arm action starts much earlier with a low toss. Still not easy in my experience. Lower contact also means higher risk.

What is meant by "higher risk"? Risk of faulting?
Roscoe is not making "low contact".
As long as one makes contact at full arm extension, which Roscoe alwaysdoes, there should not be a problem...

Also have heard some say this low-toss style puts more stress on the arm. Which implies that the server does not get fully loaded and
kinetic chain is not properly used and the body is not utilized. There may be some truth to that.
 
Last edited:
With your usual rhythm it’s hard to serve like that. Hitting arm action starts much earlier with a low toss. Still not easy in my experience. Lower contact also means higher risk.
Also, from what I've read (and observed) it is much harder to hit a good topspin (topslice?) or kick second serve with a low toss/fast motion.
 
If you hit a stationary (like at the top of the toss) ball flat, it will travel flat. If you hit a ball that is dropping (falling down the stringbed) flat, you will INDUCE topspin on the serve, increasing it's drop after clearing the net, which increases consistency.
Wow. Never read that anywhere before.
 
Wow. Never read that anywhere before.

Dolgapolov hits a kick at the peak of the toss, proving that ball does not need to be dropping in order to create topspin..
Do not see why a descending ball aids in creating topspin.
On a flat serve , the wrist will be in a more roughly neutral position. On kick, wrist will be in a more radially deviated position prior to contact.
Topspin is created as wrist moves in direction of ulnar deviation thru contact..

Screenshot-2024-02-20-212458.png


giphy.gif

Top
 
Last edited:
Dolgapolov hits a kick at the peak of the toss, proving that ball does not need to be dropping in order to create topspin..
Do not see why a descending ball aids in creating topspin.
On a flat serve , the wrist will be in a more roughly neutral position. On kick, wrist will be in a more radially deviated position prior to contact.
Topspin is created as wrist moves in direction of ulnar deviation thru contact..

Screenshot-2024-02-20-212458.png


giphy.gif

Top
IME
my best kickers are when i'm hitting a falling ball, and making contact "up & through"...
i can kick it "at it's peak" too, but i don't get as much arc/action/etc...
 
Dolgapolov hits a kick at the peak of the toss, proving that ball does not need to be dropping in order to create topspin..
Do not see why a descending ball aids in creating topspin.
On a flat serve , the wrist will be in a more roughly neutral position. On kick, wrist will be in a more radially deviated position prior to contact.
Topspin is created as wrist moves in direction of ulnar deviation thru contact..

Screenshot-2024-02-20-212458.png


giphy.gif

Top
Oh, it's not required, but unlike hitting a kick serve, you can use 100% of your racquethead speed into pace (and zero percent into spin) yet still have the control and consistency associated with a serve with spin.
 
IME
my best kickers are when i'm hitting a falling ball, and making contact "up & through"...
i can kick it "at it's peak" too, but i don't get as much arc/action/etc...

I can't hit kick but think this is true as I have heard many say this. Does a falling ball also produce more slice?
Think I have noticed more slice when I toss higher.

But I can't figure out why swing path and contact "up and through on kick" should be affected by whether the ball is descending or stationary. It should be independent.
All that should matter is the movement of the wrist and arm to impart topspin.

Have not seen the TWU Brody article on this... Does this principle also apply on the forehand. Contacting a descending ball makes it easier to hit topspin.
:unsure:
 
I can't hit kick but think this is true as I have heard many say this. Does a falling ball also produce more slice?
Think I have noticed more slice when I toss higher.

But I can't figure out why swing path and contact "up and through on kick" should be affected by whether the ball is descending or stationary. It should be independent.
All that should matter is the movement of the wrist and arm to impart topspin.

Have not seen the TWU Brody article on this... Does this principle also apply on the forehand. Contacting a descending ball makes it easier to hit topspin.
:unsure:
Do you mean a slice serve? If so, not really, as the slice is about 90 degrees from the axis of the falling toss.

The arm imparts topspin, true. But a falling toss induces more topspin, since the falling ball brushes down the stringbed.

Yes, but the racquet head speed needs to be quite high (think: swinging volley) since the upward brush needs to not only induce topspin but in addition change the trajectory of the falling ball to get the shot over the net.
 
Last edited:
Spin due to ball toss.



Suppose a racquet approaches a ball in a horizontal direction at speed V and the ball is falling vertically at speed v just prior to impact, as shown in Fig. 4(a). In a reference frame where the ball is at rest, the racquet is rising vertically at speed v while simultaneously moving horizontally at speed V , as shown in Fig. 4(b).

The situation is then the same as that shown in Fig. 2 and the spin is given by Eq. (1). If the ball falls say one metre before it is struck then it will be falling at 4.43 m.s−1 when it is struck. In a kick serve, V is typically about 40 m.s−1 . The racquet approaches the ball at an angle A given by tan A = v/V = 0.11 in this case, so A = 6.3 ◦ .

That angle could well be larger than the actual vertical approach angle of the racquet in a typical kick serve, in which case the ball toss would account for more than half of the topspin generated. From Eq. (1), we find that ω = 93 rad.s−1 (889 rpm) due to the ball toss alone.

 
I can't hit kick but think this is true as I have heard many say this. Does a falling ball also produce more slice?
Think I have noticed more slice when I toss higher.

But I can't figure out why swing path and contact "up and through on kick" should be affected by whether the ball is descending or stationary. It should be independent.
All that should matter is the movement of the wrist and arm to impart topspin.

Have not seen the TWU Brody article on this... Does this principle also apply on the forehand. Contacting a descending ball makes it easier to hit topspin.
:unsure:
my "slice", if i'm being precise, is more of a diagonal spin, as i'm swinging up & through contact... so there is an element of topspin, even on a "slice"serve... but obviously will have more sidespin than topspin (or the angle of spin)

if you drop a ball onto a static frame (with a slightly open face for the purposes of the demo), the ball will roll off the strings (voila, topspin).. probably only a few rpms, but it's something...

hitting a dropping ball with a fh is not the same, you're reversing the direction of the spin...
if anything, hitting a ball with underspin with a topspin stroke, should increase the topspin, which maybe explains why i always need to "get under more" when topspinning a ball that is knifed to me
 
Spin due to ball toss.



Suppose a racquet approaches a ball in a horizontal direction at speed V and the ball is falling vertically at speed v just prior to impact, as shown in Fig. 4(a). In a reference frame where the ball is at rest, the racquet is rising vertically at speed v while simultaneously moving horizontally at speed V , as shown in Fig. 4(b).

The situation is then the same as that shown in Fig. 2 and the spin is given by Eq. (1). If the ball falls say one metre before it is struck then it will be falling at 4.43 m.s−1 when it is struck. In a kick serve, V is typically about 40 m.s−1 . The racquet approaches the ball at an angle A given by tan A = v/V = 0.11 in this case, so A = 6.3 ◦ .

That angle could well be larger than the actual vertical approach angle of the racquet in a typical kick serve, in which case the ball toss would account for more than half of the topspin generated. From Eq. (1), we find that ω = 93 rad.s−1 (889 rpm) due to the ball toss alone.

been a while since i had to do force diagrams in physics... but takeaways is that maybe hitting a falling ball is responsible for more than just a few rpm's... :P
but also depends on how high the toss is..
biggest reason folks don't toss high (and i used to do this), is that it increases the risk of not making clean contact (further you need to toss, the more chance something can go wrong)... and most people additionally have a hard timing loading (leg/hip/core, shoulder coil) + accurate toss + accurate swing to toss... so easiert to have a meek toss that's really low... which facilitates a contact from directly behind the ball (frying pan like contact, with horizontal shoulders)
 
been a while since i had to do force diagrams in physics... but takeaways is that maybe hitting a falling ball is responsible for more than just a few rpm's... :p
but also depends on how high the toss is..

Will have to study the article. It says a 1-meter (3 feet) ball drop (commonly seen in WTA) is responsible for more than 1/2 the RPM's.
Not doubting the expert Rod Cross. But was not expecting it to be that high.
More topspin is generated by the ball drop than by a brushing racquet? :unsure:

My guess is that there is a tradeoff between ball drop and velocity. Contact near the peak and you will get more pace but sacrifice spin.
 
Last edited:
Will have to study the article. It says a 1-meter (3 feet) ball drop (commonly seen in WTA) is responsible for more than 1/2 the RPM's.
Not doubting the expert Rod Cross. But was not expecting it to be that high.
More topspin is generated by the ball drop than by a brushing racquet? :unsure:

My guess is that there is a tradeoff between ball drop and velocity. Contact near the peak and you will get more pace but sacrifice pace.
when i hit my "flat" serve, there is definitley some topspin (as my ball seems to dive into the court, and even land short in the box which is impossible for someone 5'4", unless it had some top)
any loss in pace is likely due to just spin/air friction? dunno.
if anything, if RHS is the total units of eneergy i can put into contact... if i rely on ball drop for "topspin", then all my RHS can be put into "pace"... whereas on a static ball, i have to use some portion of RHS to impart spin

side note, i don't toss as high as wta (i think they toss too high - as it becomes risky if you have the yips, wind, sun, etc...). seems like atp tend to toss lower (say 12-18in above contact)... though delbonis comes to mind as having one of the higher/highest? tosses i've seen (atp or wta)
 
Last edited:
Dolgapolov kicks with a static ball.
So if Dolgapolov kicked with a falling toss, he might possibly get more pace AND more spin?
Or would he be sacrificing pace?
:unsure:
the main sacrifice (of a high toss) IMO/IME is that the higher the toss, the harder it is to time... there's always a balancing act between consistency & precision & power & spin
high toss gives more topspin, and more time to load, now you're hitting a ball traveling at X-mph vs. 0-mph
 
reminds me a little of goran.
Yes, Goran has a low toss very much like Tanner. Djoko toss is lower now that Goran has been his coach too. Also, Cilic lowered his toss when Goran coached him and won USOpen. Kevin Curren as a great server with a low toss. Alexander Dolgopovalov was under 6 feet tall and hit the toss on the way up and for his size, he had a strong serve. There have been others too. If you are getting good extension and push up to contact, the toss doesn't have to drop.
 
Dolgapolov kicks with a static ball.
So if Dolgapolov kicked with a falling toss, he might possibly get more pace AND more spin?
Or would he be sacrificing pace?
:unsure:
He would hit with more spin (as noted), he would also have to put a hitch in his rapid flowing motion which would likely see a drop in racquet head speed, plus I think he gets a bit of surprise from the rapid motion. Overall his serve (given it's unusual type) would likely suffer in effectiveness.
 
Yes, Goran has a low toss very much like Tanner. Djoko toss is lower now that Goran has been his coach too. Also, Cilic lowered his toss when Goran coached him and won USOpen. Kevin Curren as a great server with a low toss. Alexander Dolgopovalov was under 6 feet tall and hit the toss on the way up and for his size, he had a strong serve. There have been others too. If you are getting good extension and push up to contact, the toss doesn't have to drop.
yep. low toss club myself. it's the way to go, just feels natural, and better for a nice continuous motion into the ball instead of waiting half an hour for it to drop.
 
Read: Howard Brody's The Physics and Technology of Tennis. You're free to disagree with him.

Vic Braden was a coach from the same era as Brody. Vic advocated hitting at the apex on 1st and 2nd serves. The ball is either stopped or moving very slowly at or just after the apex so the ball hangs in the hitting zone for a long time. A dropping ball is moving faster and requires better timing is Vic's theory. Tanner was the most common example of a Braden server.
 
It's a controversial topic. I hit at or shortly after the apex with the drop typically between 0-6 inches for me. I think it is better to toss under 12 inches vs over 12 inches as you don't have to pause as long to let the ball drop. But, Federer let the ball drop over 12 inches as do many pros. But, many pros toss under 12 inches too. I think shorter is easier to time and to toss accurately. Think of the rhythm as toss into the swing versus toss and then swing if you want to try it.
 
Spin due to ball toss.



Suppose a racquet approaches a ball in a horizontal direction at speed V and the ball is falling vertically at speed v just prior to impact, as shown in Fig. 4(a). In a reference frame where the ball is at rest, the racquet is rising vertically at speed v while simultaneously moving horizontally at speed V , as shown in Fig. 4(b).

The situation is then the same as that shown in Fig. 2 and the spin is given by Eq. (1). If the ball falls say one metre before it is struck then it will be falling at 4.43 m.s−1 when it is struck. In a kick serve, V is typically about 40 m.s−1 . The racquet approaches the ball at an angle A given by tan A = v/V = 0.11 in this case, so A = 6.3 ◦ .

That angle could well be larger than the actual vertical approach angle of the racquet in a typical kick serve, in which case the ball toss would account for more than half of the topspin generated. From Eq. (1), we find that ω = 93 rad.s−1 (889 rpm) due to the ball toss alone.

That's some fancy ciphering but math and playing tennis aren't the same. Tanner, Curren, and a slew of others have had great slice and kick serves hitting at the apex or within a few inches of the apex. If a low toss provides better timing and rhythm for a player, those advantages are huge. Mathematically, players would get more power if they took huge swings at ground strokes but the vast majority of ATP players take compact swings to aid in timing and to allow them to play better on the run.
 
It's a controversial topic. I hit at or shortly after the apex with the drop typically between 0-6 inches for me.

I love low tosses. But have not learned kick yet.

The Rod Cross article suggests it will be much more difficult for club players to kick a serve from 0-6 inches drop versus 12+ inches.
Note that the article uses a 3 foot drop in their analysis.

Google AI Summary:

Three main factors contributing to topspin in a kick serve:

  1. Ball toss: A falling ball can contribute more than half the topspin in some cases.
  2. Racquet tilt: Tilting the racquet head slightly forward at impact helps brush the back of the ball, generating topspin.
  3. Racquet head rotation: A brushing motion where the racquet head moves upwards and forward at an angle is crucial for topspin.

Now, to your question:
Can a player hitting the ball at the peak of the toss generate as much topspin as someone hitting with a drop by adjusting the other two factors?

While a falling ball has the potential to contribute significantly, it's not the sole factor.
Adjusting the other two factors can compensate for a lack of drop to some extent, but it's not always guaranteed to be equally effective.

Here's why:
  • Limited compensation: Adjusting the racquet tilt and rotation requires precise technique and exceptional racquet speed to achieve the same level of topspin as a significant ball drop.

Therefore, while hitting at the peak and adjusting other factors can be effective for some players, it may not always be as efficient or consistent as utilizing the ball drop's natural contribution to topspin.
 
Last edited:
The biggest advantage to a low toss is toss accuracy. It is simply much easier to precisely position the ball when throwing it to the contact point or slightly beyond the contact point. The higher the toss, the more difficult it is to make the contact location the same, which is why players with very high tosses are often chasing the ball around.

A high toss isn't needed for power even if it might aid spin. A lot of the big servers use a fairly low toss: Goran, Groth, Roddick, Kyrgios. I think the toss accuracy aiding serve accuracy is why a lot of the best servers don't toss the ball so high.
 
The biggest advantage to a low toss is toss accuracy

My toss is very accurate with a 10 inch drop. But increasing it to just to 18 inches and it drifts. Just a few inches higher , IME, makes it far more difficult to toss accurately. Granted I never practice tossing higher; but the 10 inch toss has always been accurate for me. Needed very little practice for that.

Sampras looks to have a ~2 foot drop and I do get better result with a 2 foot drop (IF toss is accurate), but the toss drifts.

Sabalenka is #1 with a sky high toss. Has one of the best serves. It is amazing that she can toss accurately with such a high toss.
 
Last edited:
Vic Braden was a coach from the same era as Brody. Vic advocated hitting at the apex on 1st and 2nd serves. The ball is either stopped or moving very slowly at or just after the apex so the ball hangs in the hitting zone for a long time. A dropping ball is moving faster and requires better timing is Vic's theory. Tanner was the most common example of a Braden server.
Of course. Everyone knows it takes better timing to hit a moving object compared to a stationary one. You asked if it was just my opinion if hitting a falling toss imparts topspin. It does.
 
Finally a clear Roscoe video.

He makes contact at the peak of the toss. 0 inches drop before contact. At full extension. He seems to hit all the checkpoints and gets into a proper trophy position.

Proper shoulder over shoulder motion? Thoracic Extension? We think so.

So minimalist and yet so powerful. Does one need to have super powers (e.g. live arm) to pull this motion off or can average players copy it?

Really don't see why the toss needs to drop more than 12 inches.

There is a new Russian WTA player whose toss drops like 3 feet before contact. Really hate seeing that. Totally unnecessary.


giphy.gif
What is really glaring to me is his back shoulder to front shoulder angle at the trophy position is very shallow compared to todays
servers . Compare this angle to Pete Sampras or Federer's. His TOSS arm does not seem to ever go above eye level either yet he hit bullets. His back leg lands into the court instead of the front leg which is another distinction from todays servers.
 
What is really glaring to me is his back shoulder to front shoulder angle at the trophy position is very shallow compared to todays
servers . Compare this angle to Pete Sampras or Federer's. His TOSS arm does not seem to ever go above eye level either yet he hit bullets. His back leg lands into the court instead of the front leg which is another distinction from todays servers.

Coaching consensus is that Shoulder-over-shoulder motion (cartwheeling) is a great source of power.
But he has a very very shallow shoulder tilt. Dont know how he does it.
Have to wonder if steep shoulder-over-shoulder motion is necessary for all.
He certainly gets by without it. This is as steep as he gets.
:unsure:
Screenshot-2024-03-07-215418.png


Screenshot-2024-03-07-222648.png
 
Last edited:
Coaching consensus is that Shoulder-over-shoulder motion (cartwheeling) is a great source of power.
But he has a very very shallow shoulder tilt. Dont know how he does it.
Have to wonder if steep shoulder-over-shoulder motion is necessary for all.
He certainly gets by without it. This is as steep as he gets.
:unsure:
I sometimes feel that some people can just swing their hand very fast. That’s it. It doesn’t mean that mechanical fundamentals can be ignored for the majority of people.
 
Coaching consensus is that Shoulder-over-shoulder motion (cartwheeling) is a great source of power.
But he has a very very shallow shoulder tilt. Dont know how he does it.
Have to wonder if steep shoulder-over-shoulder motion is necessary for all.
He certainly gets by without it. This is as steep as he gets.
:unsure:
Screenshot-2024-03-07-215418.png


Screenshot-2024-03-07-222648.png
Tell coach Nikola from intuitive tennis that Roscoe did none of this stuff!
 
I sometimes feel that some people can just swing their hand very fast. That’s it. It doesn’t mean that mechanical fundamentals can be ignored for the majority of people.
You meant whip their arms LIKE CRACKING A WHIP similar to serving??
 
Tell coach Nikola from intuitive tennis that Roscoe did none of this stuff!

Roscoe is likely an anomaly and low toss arm is not to be copied. How can you achieve shoulder-over-shoulder with such a low toss arm?!

My tossing arm is roughly forward; feels comfortable that way. I experimented with being vertical or even backward.
Feels way less comfortable for me but returners suddenly started to struggle with getting the serves back...
After a few games they started getting a few back... But I think they started to stand a step back.

Screenshot-2024-03-13-193017.png
 
Back
Top