Stats for 2007 Hamb. F (Federer-Nadal)

By winning the 2007 Hamburg Masters Series, Federer won his 48th career title, his 6th career title on clay, his 4th title at Hamburg, his 13th masters series shield, ended Rafa’s all time clay court streak of 81 matches, but most importantly, it is his only win over Nadal on clay yet or perhaps ever. Rafa was dominant in the 1st set, but Federer came out to hit the best he could on clay to easily win the 2nd and 3rd sets. The match lasted 1 hour and 55 minutes.

Result: R. Federer def. R. Nadal 2-6, 6-2, 6-0

My Stats:
Federer had 23 non-service winners: 14 fh, 2 bh, 4 fhv, 1 bhv, and 2 oh

Nadal had 19 non-service winners: 7 fh, 7 bh, 3 fhv, and 2 bhv.


Non-Service Winners by set:
Federer: 4, 9, 10
Nadal: 6, 10, 3


Federer hit 6 aces, while Nadal hit only 1 (which was a 2nd serve kicker).

Federer’s 1st serve percetage was 51 % (37 of 73), while Nadal’s was 78%
(62 of 80).

Federer hit 12 total unreturned serves with 6 aces while Nadal hit 9 total unreturned serves and 1 was an ace.

The only double fault in the match was hit by Nadal in the first set.

Federer was 5 of 11, or 45%, on breaks, Nadal was 2 of 7, or 29%.

Federer hit 51 errors in the match, which I judged 46 as unforced errors.

Nadal hit 52 errors, which I judged 36 as unforced errors.

There was 153 points in the match, Federer won 81 points and Nadal won 72 points.

Etc. stats:
In Federer’s 1st service game of the 1st set, he only served 5 of 14 1st serves in and ended with a 48% 1st serve percentage for the 1st set. (14 of 29).

In Federer’s first service game of the third set, he only served 2 of 10 1st serves in and ended up with a 29% 1st serve percentage for the third set. (6 of 21). That really surprises me to have a 1st serve percentage of 29 % in the set and win the set w/ a bagel. So, he may have won it so easy because Rafa was fatigued, especially coming from winning against Hewitt in his semi in a 3 set battle the previous day.

thanks for posting. were you able to compare with the stats that flashed at the end of the TC/Masters Series TV broadcast?

According to them Fed had 31 winners, 6 aces. So taking away the 6 aces, that would leave him with 25 winners. And taking away the 23 non service winners you counted, I can assume their statistician gave him 2 service winners.

And they gave Nadal 20 winners, 1 of them ace, so it looks like they didn't give him any service winners(since you gave him 19 non service winners)

Service winners are a gray area, there doesn't seem to be any consistency in how they are counted.

I haven't attempted to count unforced errors before, but you were close to what their statisticians gave Nadal, but gave Fed a bit more than they did(they gave Nadal 34 & Fed 31)

Comparing the unforced errors in this match to the RG final this past Sunday is interesting, Fed had basically the same amount of unforced errors in this win that he had in that loss. I guess that's one way to determine how well he played vs Nadal, comparing the unforced errors in all his matches vs him & see if it was above or below his average.

Moose Malloy said:

thanks for posting. were you able to compare with the stats that flashed at the end of the TC/Masters Series TV broadcast?

According to them Fed had 31 winners, 6 aces. So taking away the 6 aces, that would leave him with 25 winners. And taking away the 23 non service winners you counted, I can assume their statistician gave him 2 service winners.

And they gave Nadal 20 winners, 1 of them ace, so it looks like they didn't give him any service winners(since you gave him 19 non service winners)

Service winners are a gray area, there doesn't seem to be any consistency in how they are counted.

I haven't attempted to count unforced errors before, but you were close to what their statisticians gave Nadal, but gave Fed a bit more than they did(they gave Nadal 34 & Fed 31)

Comparing the unforced errors in this match to the RG final this past Sunday is interesting, Fed had basically the same amount of unforced errors in this win that he had in that loss. I guess that's one way to determine how well he played vs Nadal, comparing the unforced errors in all his matches vs him & see if it was above or below his average.

Click to expand...

I only counted the winners as absolutely winners, no hit at all from the opponent. Otherwise, I remember 1 fhv from Fed, it was basically a winner, but Nadal was soo fast and hit it. Maybe Fed had less unforced errors. But, from my perspective, he should have moved back on some of the high bounces, so its his fault. On some points, I have to re-watch them like 10 times to decide whether they were forced or unforced.

Thanks for posting. Nadal's rate of errors in this match clearly proves that unlike what people and other ATP players think about matches where they play him ( usually everyone thinks it's on their rackets and they should have played differently or attack more ) it is all about Rafael and what kind of day is he having. I imagine its hard to believe that he had 36 unforced errors in a short 3-setter from this perspective but this is the crucial stat for the outcome of the match.

I would ask the idiot who rated the highly valuable thread with 2 stars to come forward and intoruduce himself to us before he commits a mandatory suicide.

Murray_Maniac said:

Federer’s 1st serve percetage was 51 % (37 of 73), while Nadal’s was 78% (62 of 80).

Federer hit 51 errors in the match, which I judged 46 as unforced errors.

Nadal hit 52 errors, which I judged 36 as unforced errors.

There was 153 points in the match, Federer won 81 points and Nadal won 72 points.

Click to expand...

I notice your aces/winners, plus the errors, add up to 152 points, and then 153 if the double-fault is included. The ATP has 152 points altogether, 85 won by Federer, 67 by Nadal.

Do you count double-faults as (unforced) errors? (Moose and I haven't done so, though Leo Levin gave an interview once where he said doubles were ue's by definition).

Total points played/won is a stat I enjoy seeing, and calculating. I'll tell you how I do it and I'm curious to hear your method.

I get the game scores and make columns in Excel for each player, for example: Roger holding at love (game score, 4 to 0), then breaking at fifteen (game score, 4 to 1), then getting broken back in a twelve-point game (game score, 5 to 7).

When I mark down both winners and errors, it's possible to count those individually and get the total points played/won that way. I don't find that method efficient because the way that I mark down winners/errors on my sheet does not allow me easily to get the game scores. And I'd much rather have those scores. If I end up, for example, with a game of 11 points played, then I know I have to look at that game again, since of course the only games that can end in an odd number of points played are games won at 15.

Murray_Maniac said:

Federer hit 12 unreturned serves with 6 service winners while Nadal hit 9 unreturned serves and 1 was a service winner.

Click to expand...

When you use the term "unreturned serve," do you mean only serves touched by the receiver, or are you including aces? I've seen some published stats in which the term includes aces, which is why I prefer to use "return errors."

thanks for your effort with this, esp. for getting the errors as well as winners.

krosero said:

Do you count double-faults as (unforced) errors? (Moose and I haven't done so, though Leo Levin gave an interview once where he said doubles were ue's by definition).

Click to expand...

Actually, I didn't count the doubles as ue's. But I can do it next time.

krosero said:

Total points played/won is a stat I enjoy seeing, and calculating. I'll tell you how I do it and I'm curious to hear your method.

I get the game scores and make columns in Excel for each player, for example: Roger holding at love (game score, 4 to 0), then breaking at fifteen (game score, 4 to 1), then getting broken back in a twelve-point game (game score, 5 to 7).

When I mark down both winners and errors, it's possible to count those individually and get the total points played/won that way. I don't find that method efficient because the way that I mark down winners/errors on my sheet does not allow me easily to get the game scores. And I'd much rather have those scores. If I end up, for example, with a game of 11 points played, then I know I have to look at that game again, since of course the only games that can end in an odd number of points played are games won at 15.

Click to expand...

I dont write down the stats individually for each game, if thats what your doing. I tally how many of each winner, total # of errors, of those errors how many were unforced, etc. for each set. So, instead of writing that Fed broke at 1-1 and 3-5, I just write that Fed broke 2 of 5 chances in the 1st set. Then, at the end of the match, I combine all the set results them and say that Fed broke 4 of 10 chances in the match. (these stats were made up just for example) When adding up how many points Fed won I add up Fed's winners and aces, Nadal's errors and double faults. I found a shortcut to finding how many points there were in the match. Add up the total # of 1st serves in the match. I do that because in order to find their serve percentage, I tally how many total 1st serves each player hit and how many he hit in per set.

krosero said:

When you use the term "unreturned serve," do you mean only serves touched by the receiver, or are you including aces? I've seen some published stats in which the term includes aces, which is why I prefer to use "return errors."

Click to expand...

That includes the aces. The aces are the service winners. So, to find the return errors, just do : Unreturned serves - Service winners = return errors (for the other player of course).

krosero said:

thanks for your effort with this, esp. for getting the errors as well as winners.

Click to expand...

Your welcome It feels too slapdash for me to get lazy and only count the winners.

The reason that Im doing stats is that I feel obliged since not many others do it, plus its fun.

Im going to do passing shots next time (I messed them up on this match, there was a couple by Nadal), Ill fix the mistakes in a couple months to a year (Ive rewatched the matches too much right now to fix these mistakes).

Murray_Maniac said:

That includes the aces. The aces are the service winners. So, to find the return errors, just do : Unreturned serves - Service winners = return errors (for the other player of course).

Click to expand...

So let me see if I can unpack the numbers you gave, because I think your terminology is a little different from what Moose Malloy and I have been using.

Murray_Maniac said:

Federer hit 12 unreturned serves with 6 service winners while Nadal hit 9 unreturned serves and 1 was a service winner.

Click to expand...

So this means that Federer hit 12 serves that did not come back successfully or at all -- and 6 of those serves were aces, which you're referring to as "service winners."

Correct me if I'm wrong.

I read it differently before. I know what you mean about an ace being a "service winner", because technically, that's what it is -- a clean winner on the service stroke. But Moose and I have always understood "service winner" to refer to serves that are not aces but that are returned unsuccessfully. In almost all the published stats we've come across, that's what "service winner" means. There's only one case I know of where "service winners" include aces (Sports Illustrated's stats for the '98 W final), and in that case the term didn't stop at aces but included all return errors.

So your use of "service winner" to refer only to aces is a first for me, though there's a lot of tennis coverage in recent years that I haven't seen.

(Moose have you seen this before?)

When I use the term "unreturned serve" I use it as SI did -- to refer to any serve that lands in, and the server does not have to hit it again. I know when Moose Malloy uses the term, he means the return errors only.

For both of us, we judge some of the return errors as "service winners," meaning that some serves force return errors and are essentially unreturnable. That's the judgment call that Moose was talking about in his post above when he referred to the TTC stats; I thought you were doing the same, that is, judging some of Fed and Nadal's serves as unreturnable and calling them "service winners."

Am I still misunderstanding you?

Murray_Maniac said:

I found a shortcut to finding how many points there were in the match. Add up the total # of 1st serves in the match. I do that because in order to find their serve percentage, I tally how many total 1st serves each player hit and how many he hit in per set.

Click to expand...

That's actually how I get the game scores, through the service percentages. When I first started taking stats I was trying to get the game scores simply by counting the number of points, while I watched, but I found that clumsy.

That's when Moose started taking service percentages, and when I followed suit I found it was a really simple way of getting the total number of points played, and the game scores. I just put the serves in a single row and divide them by game. For example, two service games laid out like this: 12221/2221. After I watch the match, at my leisure, I can count the number of points played/won, game scores, first and second serves, etc.

Tennis boxscores have long carried game scores, for instance this one from 1920. I don't think that they used the service percentages to get the game scores, because I don't see stats for service percentages until the 1970s.

Murray_Maniac said:

The reason that Im doing stats is that I feel obliged since not many others do it, plus its fun.

Click to expand...

Glad to hear this; not everyone enjoys numbers.

krosero said:

So this means that Federer hit 12 serves that did not come back successfully or at all -- and 6 of those serves were aces, which you're referring to as "service winners."

Correct me if I'm wrong.

I read it differently before. I know what you mean about an ace being a "service winner", because technically, that's what it is -- a clean winner on the service stroke. But Moose and I have always understood "service winner" to refer to serves that are not aces but that are returned unsuccessfully. In almost all the published stats we've come across, that's what "service winner" means. There's only one case I know of where "service winners" include aces (Sports Illustrated's stats for the '98 W final), and in that case the term didn't stop at aces but included all return errors.

So your use of "service winner" to refer only to aces is a first for me, though there's a lot of tennis coverage in recent years that I haven't seen.

(Moose have you seen this before?)

When I use the term "unreturned serve" I use it as SI did -- to refer to any serve that lands in, and the server does not have to hit it again. I know when Moose Malloy uses the term, he means the return errors only.

For both of us, we judge some of the return errors as "service winners," meaning that some serves force return errors and are essentially unreturnable. That's the judgment call that Moose was talking about in his post above when he referred to the TTC stats; I thought you were doing the same, that is, judging some of Fed and Nadal's serves as unreturnable and calling them "service winners."

Am I still misunderstanding you?

Click to expand...

Ok, I get what you mean. So, do you just want me to count only return errors as unreturned serves, not aces, and say that the service winners are the forced return errors? I was going to do that until I saw your thread "Stats for 89 matches" thread which only counted winners as cleanly stuck winners. Plus, I thoght "Aces are unreturned serves aren't they?" So, for this match the 12 unreturned serves included aces and return errors and the service winners were aces. But now that you corrected me, Ill start doing unreturned serves based on returned errors and count the service winners as forced errors.

Do you want me to include return errors and df in the error count? For this one, I counted return errors and normal errors, but not df.

krosero said:

That's actually how I get the game scores, through the service percentages. When I first started taking stats I was trying to get the game scores simply by counting the number of points, while I watched, but I found that clumsy.

That's when Moose started taking service percentages, and when I followed suit I found it was a really simple way of getting the total number of points played, and the game scores. I just put the serves in a single row and divide them by game. For example, two service games laid out like this: 12221/2221. After I watch the match, at my leisure, I can count the number of points played/won, game scores, first and second serves, etc.

Tennis boxscores have long carried game scores, for instance this one from 1920. I don't think that they used the service percentages to get the game scores, because I don't see stats for service percentages until the 1970s.

Click to expand...

I don't get game scores at all. Are you counting stats such as winners per game? In that 1920 link, it just looks like a bunch of numbers. Or is it serve percentage? You or Moose will have to explain that to me.

Murray_Maniac said:

Ok, I get what you mean. So, do you just want me to count only return errors as unreturned serves, not aces, and say that the service winners are the forced return errors? I was going to do that until I saw your thread "Stats for 89 matches" thread which only counted winners as cleanly stuck winners. Plus, I thoght "Aces are unreturned serves aren't they?" So, for this match the 12 unreturned serves included aces and return errors and the service winners were aces. But now that you corrected me, Ill start doing unreturned serves based on returned errors and count the service winners as forced errors.

Click to expand...

I’m fine with any way that someone wants to present stats; whatever they count is up to them. The only thing I’ll ask is for the terms to be clear, so I know how to read the stats. This shouldn’t be a problem with most terms, but a few terms like “service winner” are just a problem; there is certainly no universal agreement on what that term means.

And the term “winner” should ideally refer to a ball that the opponent doesn’t touch (with a racquet), but there are all sorts of questions about it. Does it, for example, include service? And if it does, does that include only aces, or does it also include service winners (that is, serves touched by the receiver)?

And then for you, a “service winner” was not a serve that the receiver touches, but a clean winner on the service stroke, that is, an ace. That makes sense linguistically. If “winner” means a clean shot untouched by the opponent, then a “service winner” should just mean an ace; but that is not the meaning that the term carries when used in tennis statistics.

The only possible result of that is confusion.

And you’ve got the French Tennis Federation publishing stats in which the term “Winners (Including Service)” means points won on serve. That is clearly an error, but it may demonstrate the level of confusion there is about tennis stats.

(Last year the term meant something which none of us has yet been able to figure out, though I’ve tried emailing them).

Since you’re going to judge forced return errors as service winners, then I think you’ll be clear if you say something like, Nadal made X return errors, of which Y were forced (or service winners for Fed).

Or anything like that in your own words. Make it clear for the reader, but I think your main concern would just be to choose categories, and a writing style, that best represent what you’re interested in counting.

Murray_Maniac said:

Do you want me to include return errors and df in the error count? For this one, I counted return errors and normal errors, but not df.

Click to expand...

It’s a given that a df is a ue, so again the only thing necessary is to describe what your totals are made of. When I count ue’s, I make some kind of description like, “Apart from return errors and double-faults, so-and-so made X number of unforced errors.” Sometimes I do count service returns in double-faults, and I say so. Anyone can count anything, the reader just needs a description.

For the term “unreturned serve”, I agree with you that linguistically it should include aces, and I’ve seen it used that way in at least one published stat. I use it that way but even then I try to be clear that I’m doing so, because I know Moose uses it differently.

The English word “return” is not exactly precise. For unreturned serves, does that mean serves in which the return was bad? Serves in which no return was made at all (aces)? Both kinds of serves? For a casual or new fan, “return” could even mean a reply in a rally. The word is just not very precise.

Murray_Maniac said:

I don't get game scores at all. Are you counting stats such as winners per game? In that 1920 link, it just looks like a bunch of numbers. Or is it serve percentage? You or Moose will have to explain that to me.

Click to expand...

Game scores are just the point scores in a game. Nothing to do with winners; just the number of points won and lost. If I win a game at love, the point score of that game is 4 to 0 in my favor. If I win at 15, it’s 4 to 1. At 30, then 4 to 2. If it goes to one deuce, then the score is 5 to 3. Etcetera. You always win at least by two, and you can’t have an odd number of points in any tennis game unless it’s a game won at 15.

I wouldn’t bother with it unless it arises naturally out of your own stats. And anyway like you said you already use the other method for getting the total points played/won.

Murray, there is an aspect of match stats that many aren't aware of, that of the 'service winner.'

For example here is the link to the stats on Tsonga-Nadal from this years AO:

http://www.australianopen.com/en_AU/scores/stats/day16/1602ms.html

You can see that Tsonga is credited with 49 winners & 17 aces(and aces are included in winner counts)

Taking away those 17 aces, you are left with 32 winners.

But krosero only counted 29 non service winners:

http://tt.tennis-warehouse.com/showthread.php?t=198319

So where did the 3 remaining winners come from, according to the AO website? Presumably they were service winners. Service winners are a judgment call, there is no set rule as to what one is. These are serves that can be touched on the return that the statistician determines the returner had no play on. They don't include all unreturned serves(or return errors, however you want to count them)

Frankly, I think its a little absurd that they are included in final stats at all, the final winner count in a way is misleading.

krosero & I are just including them for fun, since the definition of what one is is not clear (you can see in the thread below we had a debate on this, I eventually just decided to use his method, in order to have consistency in our stats)

http://tt.tennis-warehouse.com/showthread.php?t=189357

Going back to my post earlier in this thread I was confused by what you said & didn't know when you used the term 'service winner' you just meant aces.

And as you can see from what I posted your stats didn't match up exactly with the Masters Series TV/TC stats(but were very close) because they included 'service winners' in their final stats. And who knows how they determined what one was.

I've noticed that some commentators don't seem to be aware of what a 'service winner' is as well, they just take the 'winner' numbers they get as meaning just clean winners, by some of the discussions I've observed.

Its odd how long this has been part of tennis, without anyone adressing the flaws in having them at all. Since I've started doing this, comparing my stats with 'official' stats has been interesting, since they rarely line up. And 'service winners' are the reason why.

So maybe you shouldn't worry about them at all(& just use the term aces to avoid confusion, it that's what you meant) since we aren't counting them in any of the charts that krosero compiled. Non service winners & aces are the important parts of this project.

Edit: and here's an even more extreme example, krosero gave Djokovic only 39 non service winners/aces vs Federer this year, but the AO website gave him 50 total winners. Those 11 'service winners' can really pad that winner column.

http://www.australianopen.com/en_AU/scores/stats/day17/1601ms.html

Thanks guys for the in depth explanations. Next time, Ill give the # of unreturned serves consisting of forced errors and return errors, but say how many were forced. Ill make aces another catagory.