Fed stats on Court Speed

Nadal_Freak

Banned
Stats on Court Speed

I already posted it but I decided to make it a thread as well to keep track of. I did some stat looking. Fed's stats in Monte Carlo, Hamburg, Rome, and Roland Garros based on First serve points won and opponents first serve points won
Monte Carlo 2007-08 Fed 70.7% Opp. 60%, Hamburg 2007 Fed 74% Opp 57.6%, Rome 2007-08 Fed 74.3% Opp. 70.5%, and Roland Garros 2007 Fed 75.6% Opp 60.1%.
So added up and average speeds would put Monte Carlo the slowest at 65.35% First Serve Points won. After that Hamburg 65.8%, RG at 67.85%, and Rome at 72.4%. So see Rome is clearly the fastest clay. At least for Fed it is.
 
Last edited:

iamke55

Professional
Court speed doesn't make any difference if it wasn't obvious enough to figure out without statistics.
 

PROTENNIS63

Hall of Fame
I already posted it but I decided to make it a thread as well to keep track of. I did some stat looking. Fed's stats in Monte Carlo, Hamburg, Rome, and Roland Garros based on First serve points won and opponents first serve points won
Monte Carlo 2007-08 Fed 70.7% Opp. 60%, Hamburg 2007 Fed 74% Opp 57.6%, Rome 2007-08 Fed 74.3% Opp. 70.5%, and Roland Garros 2007 Fed 75.6% Opp 60.1%.
So added up and average speeds would put Monte Carlo the slowest at 65.35% First Serve Points won. After that Hamburg 65.8%, RG at 67.85%, and Rome at 72.4%. So see Rome is clearly the fastest clay. At least for Fed it is.

Your analysis is dead wrong. Fed can be playing differently each tournament.. good day/bad day etc. Plus perhaps his opponent(s) is returning good.
 

Nadal_Freak

Banned
Fed's stats are pretty close in Monte Carlo last year to this year. The stats don't differentiate that much. First serve is much harder to return on a faster surface. Better stat than aces are. I disagree with all you on this. Also his opponents stats help keep it more consistent as well. The lower bouncing surface of Hamburg allows Fed to win more points on his opponents serve and allows him more points on his serve than Monte Carlo. Rome is fast and high bouncing so Fed has some trouble with the return of serve.
 

Vision84

Hall of Fame
I agree. This thread fails.
Did someone say Fail?
041_fail.jpg
 
Last edited:

Benhur

Hall of Fame
I suppose the best way to measure court speed would be to compare the average number of breaks per set over the course of many matches. No such stats are kept as far as I know.

There was a thread around here based on a related idea, measuring number of games per set. This was based on the idea that there will be fewer games in slower courts, since it is easier to break. Although this seems less accurate than the other method of just countng the number of breaks, the results seemed to conform to known reality, with more games per set showing up on faster surfaces. Still, since breaking is easier for both players on slower surfaces, you could conceivably get an equally high number of games per set as on a slow court as on a faster court. So the first method seems safer.
 

Nadal_Freak

Banned
I suppose the best way to measure court speed would be to compare the average number of breaks per set over the course of many matches. No such stats are kept as far as I know.

There was a thread around here based on a related idea, measuring number of games per set. This was based on the idea that there will be fewer games in slower courts, since it is easier to break. Although this seems less accurate than the other method of just countng the number of breaks, the results seemed to conform to known reality, with more games per set showing up on faster surfaces. Still, since breaking is easier for both players on slower surfaces, you could conceivably get an equally high number of games per set as on a slow court as on a faster court. So the first method seems safer.
Alright I'll do that. I'll come back with better stats though I thought First serve points won is a good indicator of the surface speed. I'll do it the other way as well.
 

krosero

Legend
I suppose the best way to measure court speed would be to compare the average number of breaks per set over the course of many matches. No such stats are kept as far as I know.

There was a thread around here based on a related idea, measuring number of games per set. This was based on the idea that there will be fewer games in slower courts, since it is easier to break. Although this seems less accurate than the other method of just countng the number of breaks, the results seemed to conform to known reality, with more games per set showing up on faster surfaces. Still, since breaking is easier for both players on slower surfaces, you could conceivably get an equally high number of games per set as on a slow court as on a faster court. So the first method seems safer.
True that breaking would be easier for both players. Someone might break at 4-all and get broken right back. But if you got broken and your opponent already had won 5 games, the set would be over. So statistically there might be something to the idea of fewer games on slower surfaces. Certainly when you get two big servers on good days and fast surfaces, it's easy to get a lot of tiebreaks.

I agree, just counting the breaks rather than the games would be simplest. But even so there might be a monkey wrench: what if you counted fewer breaks, because there were fewer games?

That really means you have to know both numbers: the number of games per set and the number of breaks.

What about simply getting a stat for longest average rally?
 

Nadal_Freak

Banned
True that breaking would be easier for both players. Someone might break at 4-all and get broken right back. But if you got broken and your opponent already had won 5 games, the set would be over. So statistically there might be something to the idea of fewer games on slower surfaces. Certainly when you get two big servers on good days and fast surfaces, it's easy to get a lot of tiebreaks.

I agree, just counting the breaks rather than the games would be simplest. But even so there might be a monkey wrench: what if you counted fewer breaks, because there were fewer games?

That really means you have to know both numbers: the number of games per set and the number of breaks.

What about simply getting a stat for longest average rally?
Yeah I'm going to count the breaks per game rate rather than just the amount of breaks per match. That should make the court speed official but kick serves can also throw a wrench as it can be more effective on a higher bouncing surface. Therefore make Hamburg look like the slowest rather than Monte Carlo which has a higher bounce.
 

drakulie

Talk Tennis Guru
Freak, the only way you could measure court speed is by measuring the speed of the ball when it comes off the court.

So, you would need a player for this experiment. For example, the player would hit a FH from the baseline at 80 mph?? and you would then need to measure the speed when the ball leaves the court on each surface. Additionally, the player would have to hit the same shot (spin, depth, etc) on each respective surface.

a) grass court
b) clay court
c) hard court
d) etc.
 

fgzhu88

Semi-Pro
not sure if serve percentages mean anything. The best way I think is to hear from player interviews. Why don't the interviewers ever ask Nadal what his opinion on the clay speeds/ bounces are?

I'm actually quite interested in knowing.
 

Nadal_Freak

Banned
Freak, the only way you could measure court speed is by measuring the speed of the ball when it comes off the court.

So, you would need a player for this experiment. For example, the player would hit a FH from the baseline at 80 mph?? and you would then need to measure the speed when the ball leaves the court on each surface. Additionally, the player would have to hit the same shot (spin, depth, etc) on each respective surface.

a) grass court
b) clay court
c) hard court
d) etc.
*Updated Monte Carlo is slower than Hamburg as I thought would happen. Not that Hamburg isn't slow as well.
Round 1 Breaking Percentage (24 matches) 27.97%
Round 2 Breaking Percentage (16 matches) 26.37%
Round 3 Breaking Percentage (8 matches) 31.95%
Rounds 4-6 Breaking Percentage (7 matches) 36.92%
Hamburg 2007
Round 1 Breaking Percentage (24 matches) 27.61%
Round 2 Breaking Percentage (16 matches) 25.39%
Round 3 Breaking Percentage (8 matches) 27.47%
Rounds 4-6 Breaking Percenatage (7 matches) 26.48%
Rome 2008
Day 1 Breaking Percentage (14 matches) 23.51%
Day 2 Breaking Percentage (12 matches) 24.45%
Day 3 Breaking Percentage (14 matches) 23.1%
Days 4-5 Breaking Percentage (11 matches) 22.22%
to be continue...
Roland Garros 2007
Round 1 Breaking Percentage (64 matches) 23.33%
Round 2 Breaking Percentage (32 matches) 24.86%
Round 3 Breaking Percentage (16 matches) 24.20%
Round 4 Breaking Percentage (8 matches) 28.11%
Rounds 5-7 Breaking Percentage (7 matches) 23.43%
Wimbledon 2007
Round 1 Breaking Percentage (64 matches) 17.41%
Round 2 Breaking Percentage (32 matches) 17.07%
Round 3 Breaking Percentage (16 matches) 17.744%
Round 4 Breaking Percentage (8 matches) 19.28%
Rounds 5-7 Breaking Percentage (7 matches) 14.737%
US Open 2007
Round 1 Breaking Percentage (64 matches) 21.49%
Round 2 Breaking Percentage (32 matches) 21.59%
Round 3 Breaking Percentage (16 matches) 21.908%
Round 4 Breaking Percentage (8 matches) 23.23%
Rounds 5-7 Breaking Percentage (7 matches) 24.995%
Aussie Open 2008
Round 1 Breaking Percentage (64 matches) 23.84%
Round 2 Breaking Percentage (32 matches) 23.60%
Round 3 Breaking Percentage (16 matches) 20.69%
Round 4 Breaking Percentage (8 matches) 24.21%
Rounds 5-7 Breaking Percentage (7 matches) 19.72%
Miami 2008
Round 1 Breaking Percentage (32 matches) 24.699%
Round 2 Breaking Percentage (32 matches) 20.954%
Round 3 Breaking Percentage (16 matches) 17.85%
Round 4 Breaking Percentage (8 matches) 22.91%
Rounds 5-7 Breaking Percentage (7 matches) 18.792%
 
Last edited:

Nadal_Freak

Banned
Alright I updated my list. Monte Carlo is the slowest followed closely by Hamburg. Rome is much faster and I will do Roland Garros of 2007 tomorrow to see where that lies.
 

Benhur

Hall of Fame
True that breaking would be easier for both players. Someone might break at 4-all and get broken right back. But if you got broken and your opponent already had won 5 games, the set would be over. So statistically there might be something to the idea of fewer games on slower surfaces. Certainly when you get two big servers on good days and fast surfaces, it's easy to get a lot of tiebreaks.

I agree, just counting the breaks rather than the games would be simplest. But even so there might be a monkey wrench: what if you counted fewer breaks, because there were fewer games?

That really means you have to know both numbers: the number of games per set and the number of breaks.

What about simply getting a stat for longest average rally?

I see your point. Breaks per set should work pretty well as a rough indicator, but for more accuracy what you need is ratio or percentage of breaks to total games played in a match -- and get an average for all matches in each tournament each year. This ratio can go from 0 (no breaks the entire match) all the way to 1 or 100% in the case of a match where all the games played were breaks. It seems this should be fairly easy to compute. It's just a matter of recording how many breaks there were in a match and dividing it by total number of games (excluding tiebreaks).

Average number of times the ball is struck per point would be nice to have but probably much more difficult to obtain. Though of course playing styles by period would pose a problem here. The exact same surface 40 years ago would probably yield shorter rallies on average.
 

Benhur

Hall of Fame
*Updated Monte Carlo is slower than Hamburg as I thought would happen. Not that Hamburg isn't slow as well.
Round 1 Breaking Percentage (24 matches) 27.97%
Round 2 Breaking Percentage (16 matches) 26.37%
Round 3 Breaking Percentage (8 matches) 31.95%
Rounds 4-6 Breaking Percentage (7 matches) 36.92%
Hamburg 2007
Round 1 Breaking Percentage (24 matches) 27.61%
Round 2 Breaking Percentage (16 matches) 25.39%
Round 3 Breaking Percentage (8 matches) 27.47%
Rounds 4-6 Breaking Percenatage (7 matches) 26.48%
Rome 2008
Day 1 Breaking Percentage (14 matches) 23.51%
Day 2 Breaking Percentage (12 matches) 24.45%
Day 3 Breaking Percentage (14 matches) 23.1%
Days 4-5 Breaking Percentage (11 matches) 22.22%
to be continue...

This is interesting, but where do you get the number of breaks per match? Is this statistic kept somewhere?

Also, why do you break it by round (day 1, day 2 etc)? Why not just average it for the entire tournament (total breaks in the tournament / total games in the tournament)? That should indeed give a more clear clue.
 

Nadal_Freak

Banned
This is interesting, but where do you get the number of breaks per match? Is this statistic kept somewhere?

Also, why do you break it by round (day 1, day 2 etc)? Why not just average it for the entire tournament (total breaks in the tournament / total games in the tournament)? That should indeed give a more clear clue.
The link to find these statistics is http://www.atptennis.com/5/en/vault/
Monte Carlo Breaking Percentage (Total) 29.22%
Hamburg Breaking Percentage (Total) 26.8%
Rome Breaking Percentage (Total) 23.34%
Roland Garros Breaking Percentage (Total) 24.132%
So there you you go. All the stats to the 4 major clay tournaments.
Wimbledon Breaking Percentage (Total) 17.337%
US Open Breaking Percentage (Total) 21.87%
 
Last edited:

Nadal_Freak

Banned
Those stats would be considered playing stats in percentage and not his actual speed...




Need help with tennis or strokes: Click Here
Huh? I give up. The stats consistently show that Monte Carlo, Hamburg, RG, and Rome are in that order from slowest to fastest. I think it shows quite well the speed of the court. It also matches Fed's stats in my first post close.
 

Benhur

Hall of Fame
The link to find these statistics is http://www.atptennis.com/5/en/vault/
Monte Carlo Breaking Percentage (Total) 29.22%
Hamburg Breaking Percentage (Total) 26.8%
Rome Breaking Percentage (Total) 23.34%
Roland Garros Breaking Percentage (Total) 24.132%
So there you you go. All the stats to the 4 major clay tournaments.

I can't find any statistics like that from the link you gave. I don't know if am doing something wrong. But I'd be interested to see how those percentages compare with those of tournaments on other types of surfaces. From the percentages above it does seem that Montecarlo is the slowest and Rome the fastest of the main clay tournaments. What is the percentage for tournaments like Indian Wells, Miami, AO, USO, Wimbledon and so on?

I do agree this is probably a good proxy measure of the relative "quickness" of courts.
 

illkhiboy

Hall of Fame
Sometimes stats tell the story, and sometimes they don't.

Listen, courts play differently depending on the weather. So using from just two years is not enough. Consider this statement of Agassi:

"ANDRE AGASSI: Well, I mean, I think Monte Carlo is very early in the season to really consider it good preparation. I think it is a completely different environment than Paris. I think it is very heavy and cold and the conditions it is a lot thicker there. I think Paris plays quicker than that. I think Rome is a close match to Paris and now that that they have the roof at Hamburg, I look forward to seeing that. So-- but they are also where the weeks fall and how my training is going too. So, there is a lot to consider."

http://www.asapsports.com/show_interview.php?id=304
 

fps

Legend
none of the statistics posted in this thread provide evidence one way or another in relation to court speed.
 

Nadal_Freak

Banned
I can't find any statistics like that from the link you gave. I don't know if am doing something wrong. But I'd be interested to see how those percentages compare with those of tournaments on other types of surfaces. From the percentages above it does seem that Montecarlo is the slowest and Rome the fastest of the main clay tournaments. What is the percentage for tournaments like Indian Wells, Miami, AO, USO, Wimbledon and so on?

I do agree this is probably a good proxy measure of the relative "quickness" of courts.
Yeah it is a lot of work to get these stats for all the matches.
Wimbledon 2007 http://www.atptennis.com/5/en/vault/draws.asp?TournamentID=540&TournamentYear=2007
US Open 2007 http://www.atptennis.com/5/en/vault/draws.asp?TournamentID=560&TournamentYear=2007
Aussie Open 2008 http://www.atptennis.com/3/en/vault/draws.asp?TournamentID=580&TournamentYear=2008
Indian Wells 2008 http://www.atptennis.com/3/en/vault/draws.asp?TournamentID=404&TournamentYear=2008
Miami 2008 http://www.atptennis.com/3/en/vault/draws.asp?TournamentID=403&TournamentYear=2008
none of the statistics posted in this thread provide evidence one way or another in relation to court speed.
Yeah I guess breaking serve doesn't have anything to do with court speed. :rolleyes:
Sometimes stats tell the story, and sometimes they don't.

Listen, courts play differently depending on the weather. So using from just two years is not enough. Consider this statement of Agassi:

"ANDRE AGASSI: Well, I mean, I think Monte Carlo is very early in the season to really consider it good preparation. I think it is a completely different environment than Paris. I think it is very heavy and cold and the conditions it is a lot thicker there. I think Paris plays quicker than that. I think Rome is a close match to Paris and now that that they have the roof at Hamburg, I look forward to seeing that. So-- but they are also where the weeks fall and how my training is going too. So, there is a lot to consider."

http://www.asapsports.com/show_interview.php?id=304
I disagree. The surface is more important than the weather. Though the weather can effect, not as much as the surface does. I went day-to-day and everyday Monte Carlo played slower than Hamburg, Rome, or France. It's a slower surface. Not all clay is equal.
 
Last edited:

Nadal_Freak

Banned
Show me some evidence that it does have something to do with court speed and maybe i'll believe you!
The slower the surface, the easier it is to return a serve. More serves returned equals more breaks. It's not that hard to figure out.
 

edmondsm

Legend
Huh? I give up. The stats consistently show that Monte Carlo, Hamburg, RG, and Rome are in that order from slowest to fastest. I think it shows quite well the speed of the court. It also matches Fed's stats in my first post close.

For the record, I think this is a really good thread. In you OP, Roland Garros was actually where Federer had his best serving %. I suppose this is because the field is twice as large and so his stats are inflated do to the inferiority of the opponents.
 

Nadal_Freak

Banned
that's not evidence. that's a theory.
A theory that is usually correct. Show me some evidence where it would not be correct. The more data, the more evidence to support the theory. So I guess you can't say hard courts are faster than clay either since there is no evidence. It's not exact evidence but a correlation to the evidence. With enough data you can get really close to the exact.
 

drakulie

Talk Tennis Guru
You need evidence of actual ball speed when it leaves the surface. This is the only way to determine which court is slowest to fastest.
 

fps

Legend
A theory that is usually correct. Show me some evidence where it would not be correct. The more data, the more evidence to support the theory. So I guess you can't say hard courts are faster than clay either since there is no evidence. It's not exact evidence but a correlation to the evidence. With enough data you can get really close to the exact.

you can only tell me the theory is correct if you can confirm that the courts really are, from fastest to slowest, in the order you say they are. but you're using your own theory to prove your theory. you're saying your return stats prove which courts are fastest, then say that the fastest courts are resulting in more service breaks. that's a circular argument.
 

edmondsm

Legend
You need evidence of actual ball speed when it leaves the surface. This is the only way to determine which court is slowest to fastest.


Didn't you already say this? In the absence of very scientific evidence, like the kind you suggest, Nadal_Freak's service stats do shed some light on the issue of clay court speed. It at least provides some interesting banter.
 

fps

Legend
Didn't you already say this? In the absence of very scientific evidence, like the kind you suggest, Nadal_Freak's service stats do shed some light on the issue of clay court speed. It at least provides some interesting banter.

how do they shed light on the issue? unless there's a third source that has already confirmed categorically which courts were fastest, they shed light on nothing except break point %s at the different tournaments. this they do very well.
 

Nadal_Freak

Banned
you can only tell me the theory is correct if you can confirm that the courts really are, from fastest to slowest, in the order you say they are. but you're using your own theory to prove your theory. you're saying your return stats prove which courts are fastest, then say that the fastest courts are resulting in more service breaks. that's a circular argument.
The higher percentage of breaks relates to the slower the surface is my theory. Yeah it's not exact like Drakulie's idea but it gives you a good idea on the speed of the court. Those that know tennis should agree that they would rather serves slow down on the surface to give them more time to return. Only a few maybe are hurt from the lack of pace from the serve and rather get a fast and low bouncing serve.
 

fps

Legend
The higher percentage of breaks relates to the slower the surface is my theory. Yeah it's not exact like Drakulie's idea but it gives you a good idea on the speed of the court. Those that know tennis should agree that they would rather serves slow down on the surface to give them more time to return. Only a few maybe are hurt from the lack of pace from the serve and rather get a fast and low bouncing serve.

it's an interesting one- big servers do tend not to do well on clay and nadal's serve ain't exactly ferocious. could well be something in it.
 

drakulie

Talk Tennis Guru
Nadal_Freak's service stats do shed some light on the issue of clay court speed. It at least provides some interesting banter.

Yes, it provides for some interesting discusion, however, it in no way, shape, or form provides any evidence of actual court speed.

For his theory to have any type of merit, he would have to do comparative stats on the same tournament over a period of several years. If the numbers stay relatively the same, at the same tournament, then perhaps that could show some proof his theory is correct. He would have to do this for each tournament, with the results being the same for each tournament.

I think it would be interesting to see stats, like the one the Freak is providing, related to Wimbledon the last few years compared to 10 years ago.

Perhaps Freak could dig up some stats on the web. With the Hawkeye system now in almost every tournament, they may have some stats that show ball speed when it leaves the surface of a respective court.
 

Benhur

Hall of Fame
how do they shed light on the issue? unless there's a third source that has already confirmed categorically which courts were fastest, they shed light on nothing except break point %s at the different tournaments. this they do very well.

Your point is well taken, but the theory does seem to hold prettty well in accordance with the perceived quickness of the courts as reported by players. There seems to be little controversy that hard courts and grass courts are faster than clay. And within them players distinguish between relative quickness. If break perecentage analysis corresponds to those impressions, that should be good enough to prove the theory.

The theory that slower courts allow more breaks is not controversial. It is simply something that is supported by commons sense. Pick a tennis pro and tell him: you have to play one game of tennis against Roddick. You are playing for your life, and you will be receiving serve. You can pick the surface.
What surface do you think most would pick?

Regarding the actual measurement of the speed after the bounce, it is a kind of tricky buisiness. As far as I can tell, the perceived quickness of a court refers only to horizontal ball speed after the bounce, not to actual linear ball speed along the balls trajectory after the bounce. That's why a slick grass court with a low bounce is said to play "faster" than a cement court, even though in physical terms a hard court will absorb less of the ball's energy (i.e. more of that energy will be reflected back to the ball on the bounce) and therefore if you measure the linear speed of the ball along its trajectory after the bounce, it will be faster on a hard court. But the horizontal component of that movement is larger on a slick grass court because of the lower angle, and it is the horizontal component that determines the perceived quickness.

It would be a very interesting exercise for a tennis magazine article to compile a meticulous analysis of the ratio of total breaks to total games for different tournaments across several years. Doing it by hand from the ATP vault records would be extremely tedious work. The only way I see is first adding the total number of games played in the tournament. Then get the total number of breaks by adding the "break points converted" by both players in each match. Then divide by the total games... kind of discouraging work. But I would be willing to bet that, overall, there should be a pretty neat correlation between the results and the *perceived* relative quickness of the court. As far as I have seen, the results posted by Freak for those clay tournaments already correlate pretty well with what was known. I had always heard that the Foro Italico was the quickest of the main clay courts, with Hamburg and Monte Carlo being the slowest.
 
Top