how to best measure serve speed, and compare to pros...

N

nyta2

Legend
...which i presume is measuring fastest speed off contact
i've been working on my serve in the last 6mos or so... and have been getting comments that i'm serving much bigger than before... to me, i actually feel like i'm putting in less effort so have no idea quantitatively if it has improved...
ways i've "measured" my serve:
  1. sony sensor - which attaches to bottom of racquet, and guestimates serve speed based on swing... my old serve topped 95mph'ish... the sensor has since died (rechargeable battery is no longer charging)
  2. speedChek radar - a little box that you put behind the net... "cheap" (compared to a real speed gun) $150 unit... fastest i've clocked was 85, avg 75mph, which i presume is just the fastest speed "around the net"... supposedly the ball speed decays to 50% by the time it reaches the farServiceLine... so can only presume my max speed was higher, but by how much?
  3. swing vision - iphone app... which according to documentation, records average speed from service line to service line... avg was 80mph (with faster ones typically hitting the back fence on a single bounce), but i'll get readings as high as 113mph, when ball hits the tape
  4. "real" speed guns - like the baseball players use, but the cheapest seems to be the pocket radar, for $400 - but don't really want to spend that much for a one time use
  5. playsight - the clubs that used to have one around me, seem to have gotten rid of it...
any way to extrapolate "fastest speed at contact" using methods #2, #3?

this link exists: https://tt.tennis-warehouse.com/ind...ion-ai-line-calls-and-analysis.653818/page-16
but didn't want to wade through 16 pages of comments...
 
Instead of answering your question ( :laughing: ) I'm going to say - time and effort spent caring about your serve speed would be much better spent on improving anything about your tennis - (placement, control, depth, consistency), technique, decision making, game plan, anything.

Is your serve earning you unreturned serves, or easy approach shots? Knowing your actual serve speed is almost exclusively about feeding ego.
 
nyta2 said:
...which i presume is measuring fastest speed off contact
i've been working on my serve in the last 6mos or so... and have been getting comments that i'm serving much bigger than before... to me, i actually feel like i'm putting in less effort so have no idea quantitatively if it has improved...
ways i've "measured" my serve:
  1. sony sensor - which attaches to bottom of racquet, and guestimates serve speed based on swing... my old serve topped 95mph'ish... the sensor has since died (rechargeable battery is no longer charging)
  2. speedChek radar - a little box that you put behind the net... "cheap" (compared to a real speed gun) $150 unit... fastest i've clocked was 85, avg 75mph, which i presume is just the fastest speed "around the net"... supposedly the ball speed decays to 50% by the time it reaches the farServiceLine... so can only presume my max speed was higher, but by how much?
  3. swing vision - iphone app... which according to documentation, records average speed from service line to service line... avg was 80mph (with faster ones typically hitting the back fence on a single bounce), but i'll get readings as high as 113mph, when ball hits the tape
  4. "real" speed guns - like the baseball players use, but the cheapest seems to be the pocket radar, for $400 - but don't really want to spend that much for a one time use
  5. playsight - the clubs that used to have one around me, seem to have gotten rid of it...
any way to extrapolate "fastest speed at contact" using methods #2, #3?

this link exists: https://tt.tennis-warehouse.com/ind...ion-ai-line-calls-and-analysis.653818/page-16
but didn't want to wade through 16 pages of comments...
Click to expand...


I bought a Bushnell radar gun years ago for around $80 and looks like they are still around for like $120 - https://amzn.to/3jL6BRt . Used it with the juniors and high school kids I worked with for fun and actual tracking. Does a decent job, probably like +/- 5mph.
 
Dakota C said:
Instead of answering your question ( :laughing: ) I'm going to say - time and effort spent caring about your serve speed would be much better spent on improving anything about your tennis - (placement, control, depth, consistency), technique, decision making, game plan, anything.

Is your serve earning you unreturned serves, or easy approach shots? Knowing your actual serve speed is almost exclusively about feeding ego.
Click to expand...
well, i'd say it's about 25% about feeding ego...
the other 75%, being a tech guy, is using quantitative data to show proof of improvement, and use that data to compare how far off i am (speed wise) from someone say 2 levels higher than me
 
ChaelAZ said:
I bought a Bushnell radar gun years ago for around $80 and looks like they are still around for like $120 - https://amzn.to/3jL6BRt . Used it with the juniors and high school kids I worked with for fun and actual tracking. Does a decent job, probably like +/- 5mph.
Click to expand...
saw that one, but feedback seems to be all over the place (which implies a quality control issue with the product)
that said, +/- 5mph is a biggish error range

for the sake of tracking my own performance, i can continue to use my crappy speedChek, and compare over time... but won't be able to compare that data to others (eg. pro's are gunned at contact)
 
nyta2 said:
saw that one, but feedback seems to be all over the place (which implies a quality control issue with the product)
that said, +/- 5mph is a biggish error range

for the sake of tracking my own performance, i can continue to use my crappy speedChek, and compare over time... but won't be able to compare that data to others (eg. pro's are gunned at contact)
Click to expand...

I mean, it states +/- 1mph, but I have seen it clearly misread. However, if you are tracking 50 serves in a session, you will certainly average it out and be able to compare to other groups of serves. My consistency in hitting a particular MPH is way more variable than the margin for error the gun measures it at. Been well worth it for me without dropping big $$$.
 
When they measure a pro serve at 110 or 120mph, where is the ball location at that speed? Do these devices you mention above and Playsight courts measure at the same ball location? Basically, is it a apples-apples comparison?
 
socallefty said:
When they measure a pro serve at 110 or 120mph, where is the ball location at that speed? Do these devices you mention above and Playsight courts measure at the same ball location? Basically, is it a apples-apples comparison?
Click to expand...
my understanding is that pros, playsight, stalker, bushbell, etc… record a swath of speeds, then display the fastest (typically ball off contact)

my speedchek does the same, but not as powerful a range to emit the doppler signal

arguably i could place it direct behind/above me, but that’s a hassle.

so net best safe place to place it is behind the net…
 
socallefty said:
When they measure a pro serve at 110 or 120mph, where is the ball location at that speed? Do these devices you mention above and Playsight courts measure at the same ball location? Basically, is it a apples-apples comparison?
Click to expand...
apples to apples comparison is exactly what i’m looking for… in the cheap
 
nyta2 said:
saw that one, but feedback seems to be all over the place (which implies a quality control issue with the product)
that said, +/- 5mph is a biggish error range

for the sake of tracking my own performance, i can continue to use my crappy speedChek, and compare over time... but won't be able to compare that data to others (eg. pro's are gunned at contact)
Click to expand...

It's not so much quality control but you have to be directly behind and in line with the ball for the reading to be accurate otherwise you get into a geometry problem.

My son bought one for a high school project and in my opinion it was a futile effort. The best results we achieved was at a club with an open area that was above and behind the server where we could stand with the gun.
 
LOBALOT said:
It's not so much quality control but you have to be directly behind and in line with the ball for the reading to be accurate otherwise you get into a geometry problem.

My son bought one for a high school project and in my opinion it was a futile effort. The best results we achieved was at a club with an open area that was above and behind the server where we could stand with the gun.
Click to expand...
good to know that maybe even with a good($$$) radar gun, the positioning (above & behind) is key... (which is where playsight and guns at pro events, position the radar)
was guessing this might be the case, which i was another reason i was reluctant to buy a more expensive radar gun.
 
nyta2 said:
good to know that maybe even with a good($$$) radar gun, the positioning (above & behind) is key... (which is where playsight and guns at pro events, position the radar)
was guessing this might be the case, which i was another reason i was reluctant to buy a more expensive radar gun.
Click to expand...

You would never know I have a MS in EE and I am sure someone is going to point out my error but I am doing this brain exercise while on a Teams call but you end up having to guesstimate the angle and if the gun is faced toward the horizontal side of the triangle and the serve is directed downward as the hypotenuse you divide the reading by the cosine of your estimated angle.

In the end it is just a cobbled together value and one might as well just tell people you have a bigger serve than Roscoe Tanner and have them prove otherwise.
 
found my old playsight account, and data... last used in 2014.... top serve speed 98mph, avg (1st) 90mph... my strokes & movement look really ugly :P
that said, can't find the playsight smartcourt locator... maybe a dying biz?
 
Mount your phone on the back fence. Make sure the camera angle can show contact point and service box where ball bounces.

Record at 240 fps or more. Count the frames.

Someone made an online calculator to compute the speed. I tried comparing it with a speedchek I had access to and the speeds were about the same. 103-107 mph
 
onehandbh said:
Mount your phone on the back fence. Make sure the camera angle can show contact point and service box where ball bounces.

Record at 240 fps or more. Count the frames.

Someone made an online calculator to compute the speed. I tried comparing it with a speedchek I had access to and the speeds were about the same. 103-107 mph
Click to expand...

To me this seems way more accurate than a radar gun.
 
LOBALOT said:
...you have to be directly behind and in line with the ball for the reading to be accurate otherwise you get into a geometry problem.
...above and behind the server where we could stand with the gun.
Click to expand...

We found at about head height and just off to the side where the radar could track the ball moving gave the most consistent. Can't post the old kids stuff, but found this example from 2109. Now my buddy wasn't as high as he should be and a bit wider than I would do, but is a good example of how we used to measure.



PS - funny to watch this. I had muscle pull/strain issue I had been working through and developed what we called the Captain Morgan serve, with the leg coming up like that lol.

91f44fa603b054f0872a4c21e5da7e18.jpg



PPS - And we also would pull the trigger right as the ball peaked and the person started their push up/swing. If you pull too early or too late it wouldn't read. We assumed that was just how it tracked the ball.
 
ChaelAZ said:
We found at about head height and just off to the side where the radar could track the ball moving gave the most consistent. Can't post the old kids stuff, but found this example from 2109. Now my buddy wasn't as high as he should be and a bit wider than I would do, but is a good example of how we used to measure.



PS - funny to watch this. I had muscle pull/strain issue I had been working through and developed what we called the Captain Morgan serve, with the leg coming up like that lol.

91f44fa603b054f0872a4c21e5da7e18.jpg



PPS - And we also would pull the trigger right as the ball peaked and the person started their push up/swing. If you pull too early or too late it wouldn't read. We assumed that was just how it tracked the ball.
Click to expand...

Thanks!

I have similar arse condition right now as that other post I saw from earlier this week so certainly understand the Captain's Morgan serve. I get a very sharp pain when lifting my leg in my rear end above and to the right of my left hip (Upper rear end area). I did it scrambling to my backhand side to chase down a lob over my partners head. I have had it for 3 weeks.

Maybe I got me some "Jimmy Leg" from too much Jim Beam.
 
onehandbh said:
Mount your phone on the back fence. Make sure the camera angle can show contact point and service box where ball bounces.

Record at 240 fps or more. Count the frames.

Someone made an online calculator to compute the speed. I tried comparing it with a speedchek I had access to and the speeds were about the same. 103-107 mph
Click to expand...
i read about that (probably on ttw), the gist of my understanding is that counting frames will give you the avg velocity across the distance traveled but not the max speed (eg. baseline to farServiceLine, the ball supposedly loses ~50% velocity just before the bounce). so i'm guessing if the speedcheck was positioned behind the net, perhaps the avgSpeed is approximately the same as the speedAsTheBallCrossesTheNet...
any flaws in my understanding?
 
onehandbh said:
Mount your phone on the back fence. Make sure the camera angle can show contact point and service box where ball bounces.

Record at 240 fps or more. Count the frames.

Someone made an online calculator to compute the speed. I tried comparing it with a speedchek I had access to and the speeds were about the same. 103-107 mph
Click to expand...
Seems like that resource has been dead at my end for quite a while:

The Wayback machine still works for me.

So I made a local copy, which anyone can do.
 
nyta2 said:
i read about that (probably on ttw), the gist of my understanding is that counting frames will give you the avg velocity across the distance traveled but not the max speed (eg. baseline to farServiceLine, the ball supposedly loses ~50% velocity just before the bounce). so i'm guessing if the speedcheck was positioned behind the net, perhaps the avgSpeed is approximately the same as the speedAsTheBallCrossesTheNet...
any flaws in my understanding?
Click to expand...
I haven't looked at his formula in detail but I think he was estimating initial speed (peak).
 
nyta2 said:
i read about that (probably on ttw), the gist of my understanding is that counting frames will give you the avg velocity across the distance traveled but not the max speed (eg. baseline to farServiceLine, the ball supposedly loses ~50% velocity just before the bounce). so i'm guessing if the speedcheck was positioned behind the net, perhaps the avgSpeed is approximately the same as the speedAsTheBallCrossesTheNet...
any flaws in my understanding?
Click to expand...

A served tennis ball will lose one MPH for every 2-3 feet of travel due to aerodynamics and offset by gravity, so if it travels 60 feet to the contact point, assume it loses 25 MPH. The bounce will reduce the speed by like 30%, and then it will travel another 25 feet or so to your racquet, losing another 10 MPH.
 
onehandbh said:
Mount your phone on the back fence. Make sure the camera angle can show contact point and service box where ball bounces.

Record at 240 fps or more. Count the frames.

Someone made an online calculator to compute the speed. I tried comparing it with a speedchek I had access to and the speeds were about the same. 103-107 mph
Click to expand...

Based on the practice serves you hit the other day, 103-107 seems about right and especially considering that you haven't been serving much for quite some time. On that court, hitting the back wall at waist level like your serves were doing is the sign of a well hit serve and not many guys will do that.
 
If you don't need immediate results, framecount method is best and easiest you can use. Serve is very restricted, particularly fast serve, in terms of what path it can go to get over the net and land in. In such circumstances time from impact to bounce is very reliable thing. Of course there are different aerial conditions, temperatures and ball fuzziness, but after all who cares if you hit it faster off the racquet but then it slows down a bit more - if you use actual matchplay video to calculate your serve speed range, it's what you need.

You can always find pro videos with radar readings to ensure that method is working rather precise.
 
nyta2 said:
i've been working on my serve in the last 6mos or so... and have been getting comments that i'm serving much bigger than before... to me, i actually feel like i'm putting in less effort so have no idea quantitatively if it has improved...
Click to expand...
That's exactly what you should expect when you learn better technique. Less effort, more power. That's why women and children with better serve technique are able to serve bigger than much taller, stronger men with bad technique.
 
Radar gun. "Pocket Radar" had some good comments here in some posts..

Video camera with known frame rate. 1/30 sec. 1/60 sec

The simplest way to measure speed is to measure the distance traveled in a known time.

D = v x t

D- distance traveled in one frame time
v - serve speed
t - time between frames, 1/30 sec, 1/60 sec, etc.

I posted this video measurement method several times some years ago

An easy way to get the scale is to use the known ball diameter.

If you attempt it, find my old posts and let me know and I'll help you.
 
Last edited:
You must log in or register to reply here.

Similar threads

ace_pace
Serving woes and questions
Replies
30
Views
11K
boramiNYC
B
Steady Eddy
How fast to most rec. players serve?
Replies
26
Views
2K
Tenthinker
T
Z
Clearance and racquet speed on kick serve
Replies
6
Views
1K
kicker75
K
jiangrm
How did you increase serve speed from 90mph to 110mph?
2
Replies
53
Views
2K
jiangrm
jiangrm
W
Learning to pronate
Replies
17
Views
1K
5263
5263
Back
Top