Aces and DFs

I was talking about this in another thread, but I had not put all the numbers together carefully.

I was calling aces free points, and thinking of DFs as giving away free points to the opponent.

Several people objected to this as a valuable metric, saying that DFs are also tactical (when is it wise to go for the 2nd serve), and that not counting unreturnable serves and weak returns is too simplistic.

My points:

From aces, we should also get an idea of the relationship between them and unreturnable serves, and weak returns. This should be, roughly, somewhat predictable.

Aces and DFs are recorded well, at least since 1991, so it's something we can look at for all players. Good luck getting data for unretunables and weak returns.

But for the moment, I'm going to ask people to go along with what I'm doing, to see where it gets us.

One statistic we can instantly track is this: How many DFs per game? Because this tends to show how "clean" the service game is. No one likes to give a way points.

I tried to include in my list of players all players who are known for having very effective serves. Not the most aces, but the guys who top the % of service games won. This list includes the guys who are servebots, the guys who smoke the ball and get a ton of free points, but it also includes players who have a relatively low ace count but who win a lot of games.

One of the first players I checked was Dolgopolov because of his quick release serve and reputation for being a very good server, for someone his size. But it turns out that he fails as a very good server for two reasons:

1. Only .35 aces per game, meaning he has to play about 3 service games to ace once. This by itself is not bad at all, for a shorter guy. He also aces nearly three times for every DF, which is not bad at all, but when you subtract DFs from aces, the net is 1 clean free point every 4 games. This is about the same as Djokovic, so again, not bad at all. But still fairly far down on the list.

2. Of all the guys I checked, his % of service games is second worst, with only Chang down a bit more. Neither won 80% of their games, career. But numbers were lower, across the board, in the 90s. No poly. So Chang was no doubt more effective in winning service games, and very close to the Dog in aces per game.

The second lowest guy in my list, DFs to games, is one of the best players in tennis, in this era.

Question:

Who do you think that player is?

You may find someone else even better, career, all surfaces, but I'll be shocked if it is anyone who any of us think of as a great player.

Guesses?

Next question, and this one should be obvious:

Who has the record for aces per game since 1991, and most likely in the history of tennis? Because it is VERY unlikely that this stat was higher, by anyone, before 1991?

If you had to guess the top five players, what would you guess?

I have the answers. Some would be obvious to me. Some are a surprise. I got my list from the ace list, simply going right down the list, then dividing by the number of games they played. Hint: Federer and Sampras are WAY down the list, so don't go by total aces. Federer, for instance, has played a massive amount of games.

A second question, related to the first: Of the great players, who do you think is lowest on this list? Who gets the lowest number of aces per game?

Now, one more "puzzler".

Which of the players we associate with being servebots, the guys with immense numbers of aces, also have the highest number of aces to DFs?

This is easy to compute, but I have not seen it done this way.

Hint: there is no good player who DFs as often as he aces. That's pretty obvious.

But there is one player who only aces 1.91 times for every DF. Out of all the top players, and go down as far as perhaps top 20, who do you think gets this very low ratio?

The answer might surprise you.

Second question:

One guy aces between 7 and 8 times for every DF. This is the best ratio since these stats have been recorded. Who do you think that is? And if you had to pick the top five, just guessing, who would you pick?

I would say Ivo, Isner, Arthurs, Johannson, Raonic? For most aces per game?

Isner definitely has the best ace to DF ratio. Maybe Goran for most DFs to aces?... Or Mark?

Gary Duane said:

Next question, and this one should be obvious:

Who has the record for aces per game since 1991, and most likely in the history of tennis? Because it is VERY unlikely that this stat was higher, by anyone, before 1991?

If you had to guess the top five players, what would you guess?

I have the answers. Some would be obvious to me. Some are a surprise. I got my list from the ace list, simply going right down the list, then dividing by the number of games they played. Hint: Federer and Sampras are WAY down the list, so don't go by total aces. Federer, for instance, has played a massive amount of games.

A second question, related to the first: Of the great players, who do you think is lowest on this list? Who gets the lowest number of aces per game?

Click to expand...

1. Karlovic, Isner, Ivanisevic, Arthurs would be in the top 4 in some order. Karlovic #1. 5th spot would be between Roddick, Raonic and Krajicek, I'm guessing

2. Among the ATGs, (6 slams or more) has to be between Connors and Wilander, no ?

Gary Duane said:

Second question:

One guy aces between 7 and 8 times for every DF. This is the best ratio since these stats have been recorded. Who do you think that is? And if you had to pick the top five, just guessing, who would you pick?

Click to expand...

top 4 would probably be Karlovic, Isner, Roddick, Federer in some order ?

abmk said:

top 4 would probably be Karlovic, Isner, Roddick, Federer in some order ?

Click to expand...

In this order:

Isner
Karlovic
Roddick
Ljubicic
Raonic
Tsonga
Federer

Those are the only guys who ace 4 times or more for every DF. Tsonga and Fed are so close that one error would reverse them, tied to 2 decimal points. So put them even.

The stat you don't see is that of all of them, Fed is least likely to DF in any game, only 15% of the time. Isner and Roddick are close, DFing 16% of the time (per game) but with a higher ace count.

abmk said:

1. Karlovic, Isner, Ivanisevic, Arthurs would be in the top 4 in some order. Karlovic #1. 5th spot would be between Roddick, Raonic and Krajicek, I'm guessing

2. Among the ATGs, (6 slams or more) has to be between Connors and Wilander, no ?

Click to expand...

Here are the guys pretty much in a class by themselves:

Karlovic
Isner
Arthurs
Raonic
Ivanisevic
Kyrgios
Krajicek
Ljubicic
Roddick

Where guys like Connors and Wilander fit in, no idea. I have no data. But we know that service% keeps going up, and return% keeps going down. Fed is at 61%, Sampras at 83%.

We can't weight for era/equipment, and that's a shame. Bigger shame: lack of data before 1991.

Old Becker is incredibly impressive. He double-faulted in about 1 out of 3 games (36%), but I'd wager he would be fully competitive in this era, full career and playing with poly.

You won't find Kyrgios on the ATP list because the ATP does not yet list him - not enough games. But since that number is very likely to go down, unless his career is cut short with injuries, he is destined to move up.

Roddick 93%, Karlovic a whopping 143%.

My take-away, so far, is that ATGs don't get to the top of the list, probably in any era. They get there for other reasons.

When it comes to winning games, by the way, Fed and Sampras are 5th and 6th on the whole list, and barely separated. They are easily most impressive in their height range (no surprise), and if anything Sampras is even more impressive because he's the only guy in the top of the list from the 90s. So I would say that if we could weight for era, he goes very close to the top, and I think if we had stats from the 50s, the same thing would probably be true for Gonzalez.

In fact, I'm thinking that Kyrgios is the closest thing we've seen to Pancho, in decades, with the single and huge problem that Gonzalez was hungry, very healthy and hated losing more than anyone else in tennis history. So Kyrgios may never win a major, but the talent is there - Big Cat.

The stat that I really want to see is '1st and/or 2nd shot wins the rally' in a service game setting. This would be, for me, the leading stat for service greatness.

Russeljones said:

The stat that I really want to see is '1st and/or 2nd shot wins the rally' in a service game setting. This would be, for me, the leading stat for service greatness.

Click to expand...

Russel, this gets to the crux of service game vs the serve alone.

Aces tell us about players who get lots of points with close to zero effort. I don't think any of us watch tennis to enjoy seeing balls go flying by, while the opposing players can do nothing.

I would argue that the balance of power has been shifting more and more to the server for decades, which is reflected in the balance of service games and return games.

In other words, when we go back to the time people like Pancho Gonzalez, whom other players thought was so dominant on serve that it was unfair, we think we are talking about another guy like Kyrgios. The reality is quite different. He had to step into the court, no jumping. Those old rackets could not produce the monstrous serves, again and again, with so much spin.

Even since the 90s we see that these stats are going up and up and up. Sampras was a freak in the 90s, and you can see many years that he won more games by percent than any other player. You can say that the players were shorter - perhaps a valid point - but it's more than that.

To find the players you are interested in, you can use your eyes, but stats will back those eyes up.

An example: Nadal is incredibly stingy with DFs. His number of DFs per game are as low as we see from any top players, and stats bear that out. Last year I checked, and he DFs 11% per game on clay, meaning that you have to see him serve a LOT of games before he gives a way points. We also know that on that surface he is wicked for a 1-2 punch, tricky serve with a weak return, than BAM.

On fast surfaces, I would say that Sampras and Fed, statistically, are right at the top of the guys who do it with the one two punch, same reason. Fed may have to do it more often in three shots, but it's the same idea.

You look for the guys who are NOT the biggest acers, then look for guys who are close to the top in games won, on serve. Those are the guys I want to watch, not the bots.

Gary Duane said:

Russel, this gets to the crux of service game vs the serve alone.

You look for the guys who are NOT the biggest acers, then look for guys who are close to the top in games won, on serve.

Click to expand...

How does someone like Edberg stack up? He didn't usually have many aces in a given match compared to servebots (relatively slow, spinny 1st serve), yet he usually held serve due to his GOAT volley skills. Would be interesting to see the stats for his incredible 1991 and 1992 US Open tournaments.

Gary Duane said:

You look for the guys who are NOT the biggest acers, then look for guys who are close to the top in games won, on serve. Those are the guys I want to watch, not the bots.

Click to expand...

The players you describe are bots. They're just not servebots. They're grindbots. And they're the dullest of the dull.

Give me a Karlovician serve artiste over a Djokovician grindbot every time. At least there's something to be awed by.

IowaGuy said:

How does someone like Edberg stack up? He didn't usually have many aces in a given match compared to servebots (relatively slow, spinny 1st serve), yet he usually held serve due to his GOAT volley skills. Would be interesting to see the stats for his incredible 1991 and 1992 US Open tournaments.

Click to expand...

Here's the problem. One site copies data from another site. Look at this:

http://www.tennisabstract.com/cgi-bin/player.cgi?p=StefanEdberg&f=ACareerqq

Immediately you see this:

796-266 (75%)

That winning% in matches does not quite get to the elite level that the very top ATGs get to. You can check that in a flash by comparing the Big Four.

Murray, with only three majors:

719-206 (78%)

The others are higher.

I think you'll find that the best of the OE are around 80%, all matches. So the first thing I would do is look at return/serve.

The first thing I see is that over 30% of games, career, is very good. I believe this figure used to be higher, and the % for games won on serve was lower, but when you put it all together, career, for the very best players you want to be around a minimum of 57%.

Edberg is a bit low, total of 56.45%.

Here's the problem. How LONG is that for? It says career.

So I go here:

http://www.atpworldtour.com/en/players/stefan-edberg/e004/player-stats

And there are the same stats.

WRONG.

I go to the main stats page, see this for serves:

82.56%

So no where on the ATP site or the TA site does it warn that their "career" figures are coming from post 1991.

Since Edberg's first title was in 84, we know we are looking at total nonsense. We are missing at least 8 years.

So the only thing I can tell you is that if we go by the ATP, which starts in 1991, and where the data in those first years was not so good, who knows how accurate their data is.

The only way you can get even accurate info on games is to scour all the data, for each player, then compute it yourself. I have not yet done that for Edberg, but I can tell you that this looks very wrong to me:

Aces/DFs:

1405 1630

According to this Edberg double faulted more often than he aced.

Back to here:

http://www.atpworldtour.com/en/players/stefan-edberg/e004/player-stats

Check out aces and DFs. Even if we accept the idea that this data is from 1991 on, he was hardly a crap player then. So those stats absolutely do not make any sense at all.

If you are a big Edberg fan, what do you think? No other top player I've checked has more DFs than aces.

It just makes no sense.

SinjinCooper said:

The players you describe are bots. They're just not servebots. They're grindbots. And they're the dullest of the dull.

Give me a Karlovician serve artiste over a Djokovician grindbot every time. At least there's something to be awed by.

Click to expand...

I doubt you even read what I wrote:

"On fast surfaces, I would say that Sampras and Fed, statistically, are right at the top of the guys who do it with the one two punch, same reason. Fed may have to do it more often in three shots, but it's the same idea.

You look for the guys who are NOT the biggest acers, then look for guys who are close to the top in games won, on serve. Those are the guys I want to watch, not the bots."

here is a thread with serve stats for the 1982 season, if you're interested.
http://tt.tennis-warehouse.com/index.php?threads/1982-service-statistics.357452/#post-5257628

Moose Malloy said:

here is a thread with serve stats for the 1982 season, if you're interested.
http://tt.tennis-warehouse.com/index.php?threads/1982-service-statistics.357452/#post-5257628

Click to expand...

I am MORE than interested.

What is useful about this is that we have comparisons of the greatest, from that era.

Let's throw out Isner, who aces 7.66 as many times as he DFs for his career. He is the obvious outlier, and there were no players like him decades ago.

Karlovic and Roddick are almost tied, about 5.75 aces per DF.

The lowest person I have in this era is Sock: 1.52

No great player in the last 25 years is that low, except for players with incomplete data.

Nadal, who is very low, is 1.91.

Comparing the list you gave me:

Only one player, Noah, got close to 4 aces to 1 DF ration:

Noah:
3.88

Lendl is next at 3.15

The average for the whole list is 1.57. In other words, in those days if you won two aces for every DF, you were doing well.

In contrast, the list I have compiled is for careers, so we can expect the figures to be lower, than peak years, but my list of players is off the ace list, so that should be higher. Even so, I believe that ratio is much higher now.

For example, Chang was
3.02 in '96, and he's about the weakest server on the list.

Fed hit 9 in 2008.

This kind of ratio was simply not possible decades ago. It's poly and the modern rackets, all the spin.

Gary Duane said:

If you are a big Edberg fan

Click to expand...

He's one of my favorites! 1991 he was on fire at the US Open - beat Chang, Lendl, and Courier all in straight sets!!!

Gary Duane said:

According to this Edberg double faulted more often than he aced.

Click to expand...

I have watched a lot of his matches. He didn't have a 1st serve that tended towards aces (fairly spinny and slow compared to, say, Ivanisevic or Roddick). Many times, his 2nd doesn't look all that different from his 1st. And, on most surfaces, he was often coming in behind that 2nd serve so maybe wanted to get a little extra on it compared to someone like Murray or Nadal who always stay back on 2nd serves.

So, I could see that he would be weak in this particular metric of aces to DF.

Would be interesting to see if other big S&V with spinnier/slower 1st serves (e.g. Rafter) who regularly came in behind 2nd serves also had lots of DF in comparison to aces? Due to trying to come in behind a strong 2nd serve so they don't get passed, and not having lots of aces that a bigger S&V server such as Sampras or Ivanisevic would have.

Would also be interesting to know if opponents of these agressive net players also tended to double-fault more when playing them, as the pressure was so intense knowing that Edberg, Rafter, Sampras was going to chip & charge your 2nd serve if you floated it.

Many of those guys also played their S&V on quite a bit of clay, which might further skew the stats for % matches won and % service games won.

Just some thoughts...

Gary Duane said:

In this order:

Isner
Karlovic
Roddick
Ljubicic
Raonic
Tsonga
Federer

Those are the only guys who ace 4 times or more for every DF. Tsonga and Fed are so close that one error would reverse them, tied to 2 decimal points. So put them even.

The stat you don't see is that of all of them, Fed is least likely to DF in any game, only 15% of the time. Isner and Roddick are close, DFing 16% of the time (per game) but with a higher ace count.

Click to expand...

Ljubicic is an interesting entry. Excellent server, but didn't expect him to be that high as far as ace to DF ratio goes.

Gary Duane said:

Here are the guys pretty much in a class by themselves:

Karlovic
Isner
Arthurs
Raonic
Ivanisevic
Kyrgios
Krajicek
Ljubicic
Roddick

Where guys like Connors and Wilander fit in, no idea. I have no data. But we know that service% keeps going up, and return% keeps going down. Fed is at 61%, Sampras at 83%.

We can't weight for era/equipment, and that's a shame. Bigger shame: lack of data before 1991.

Old Becker is incredibly impressive. He double-faulted in about 1 out of 3 games (36%), but I'd wager he would be fully competitive in this era, full career and playing with poly.

You won't find Kyrgios on the ATP list because the ATP does not yet list him - not enough games. But since that number is very likely to go down, unless his career is cut short with injuries, he is destined to move up.

Roddick 93%, Karlovic a whopping 143%.

My take-away, so far, is that ATGs don't get to the top of the list, probably in any era. They get there for other reasons.

When it comes to winning games, by the way, Fed and Sampras are 5th and 6th on the whole list, and barely separated. They are easily most impressive in their height range (no surprise), and if anything Sampras is even more impressive because he's the only guy in the top of the list from the 90s. So I would say that if we could weight for era, he goes very close to the top, and I think if we had stats from the 50s, the same thing would probably be true for Gonzalez.

In fact, I'm thinking that Kyrgios is the closest thing we've seen to Pancho, in decades, with the single and huge problem that Gonzalez was hungry, very healthy and hated losing more than anyone else in tennis history. So Kyrgios may never win a major, but the talent is there - Big Cat.

Click to expand...

Raonic above Goran is interesting. Not much of a difference, but I guess the difference is in poly helping Raonic to get more first serves in, leading to a slightly higher ace%.

Gary Duane said:

The second lowest guy in my list, DFs to games, is one of the best players in tennis, in this era.

Question:

Who do you think that player is?

Click to expand...

Gary Duane said:

But there is one player who only aces 1.91 times for every DF. Out of all the top players, and go down as far as perhaps top 20, who do you think gets this very low ratio?

Click to expand...

Gary Duane said:

A second question, related to the first: Of the great players, who do you think is lowest on this list? Who gets the lowest number of aces per game?

Click to expand...

You answered them yourself. Nadal.

Charleneriva said:

You answered them yourself. Nadal.

Click to expand...

I guess I did. The only player I've found, so far, who DFs less per game is Dolgopolov, so there may be others who are lower ranked. But no one comes close to Nadal of all the top players, and he hits something like .11 aces per game on clay, meaning that on average he serves around 9 games before he serves the next DF, on average. Since surface has nothing to do with DFs, it's tactics.

abmk said:

Raonic above Goran is interesting. Not much of a difference, but I guess the difference is in poly helping Raonic to get more first serves in, leading to a slightly higher ace%.

Click to expand...

Between Ivanisevic and Raonic it's 1.10 to 1.12 aces per games. And Roanic may go up or down now, depending on how his career plays out. Usually it keeps going up, because the career average includes the first few years on the tour, and the serve is the last thing to go.

Ivanisevic was 2.86 times as likely to ace. Roanic, 4.79. Other people don't agree with me, but I think aces-DFs is a better metric.

This widens the gap, .89 to .72, and I think that number is very important in any era, because when that number is lower, a lot lower, it at least opens up the possibility of a server taking too many chances and blowing a lead.

Without going into too much detail, Sampras clearly outaced Fed.

Sampras .83, Fed .61. Looking only at that stat you have to conclude that Sampras probably won more games, which he actually did only by the tiniest hair.

Sampras DFed more, subtracting gives a different view. The gap now narrows to .55 against .46. Sampras is still better, but it's gets closer when you look at DFs.

For yet another view, because Sampras DFed more often, we look at DFs per game. Fed DFs 14.9% of games, meaning that he'll play close to 7 games before he DFs the next time. Sampras DFed 28.2% of the time, so he would give away a point close to 1 time in every 3 games. So no matter how tactical he was about hitting DFs, he also took more changes and had to get burned more often. But his higher ace rate still indicates that overall he tended to win more points with exactly one shot.

Using aces-DFs, here is the list from the top through to Fed:

Karlovic
Isner
Raonic
Arthurs
Roddick
Kyrgios
Ljubicic
Ivanisevic
Krajicek
Anderson
Querry
Muller
Tsonga
Sampras
Stich
F. Lopez
Rusedski
Becker
Safin
Cilic
Kuerten
Philippoussis
Almagro
Johnson
Federer

In other words, there are a lot of guys who get more free points than Fed, if we do not also count unreturnables weak returns.

If, on the other hand, we go strictly by % of games won on serve, the list changes radically:

Karlovic
Isner
Raonic
Roddick
Sampras
Federer

I had this list years before the ATP finally started showing decimals:

http://www.atpworldtour.com/en/stats/service-games-won

Intuitively I would say that Sampras should be even higher on the list, because players from the 90s have lower stats, as a group, analyzing by height.

But what is most interesting to me is that by looking at aces-DFs, then comparing to game%, we can see that these "free points" get balanced against all the other skills ATGs have to back up the serve.

Nadal is at least close to dead last, but he's #15 in games on all surfaces, and #6 on clay. On that metric Fed and Nadal are only a hair apart, and that shows us that all Fed's amazing skills beside the serve do not work nearly so well on clay, which is more than obvious.

abmk said:

Ljubicic is an interesting entry. Excellent server, but didn't expect him to be that high as far as ace to DF ratio goes.

Click to expand...

Totally shocked me.

I'm embarrassed to admit that I know next to nothing about him. From those stats alone I would guess a really big serve but lacking skills to back it up. Aces/game, .002 difference between Roanic and Ljubicic, which means pretty much that one wrong entry would change it. So for all purposes, tied. He doubled faulted a bit more often, but still only about 1 time every 5 games, so a very clean server. But 85.24% of games, which is more like what we usually see in the 90s.

So something had to be missing. Net skills, or groundstrokes, movement. Usually when you see players with such great ace/DFs stats but not amazing game%, they are low in return games and thus in return skills and stats.

I should check:

There it is: only 18% of games won on return. That may not even get him on the all time list of return game%, or if so he's going to be close to the end.

That's why Karlovic is only at 92% of games on serve, career. As high as that is, with top defensive skills that number would be more like 95-96%.

Gary Duane said:

Totally shocked me.

I'm embarrassed to admit that I know next to nothing about him. From those stats alone I would guess a really big serve but lacking skills to back it up. Aces/game, .002 difference between Roanic and Ljubicic, which means pretty much that one wrong entry would change it. So for all purposes, tied. He doubled faulted a bit more often, but still only about 1 time every 5 games, so a very clean server. But 85.24% of games, which is more like what we usually see in the 90s.

So something had to be missing. Net skills, or groundstrokes, movement. Usually when you see players with such great ace/DFs stats but not amazing game%, they are low in return games and thus in return skills and stats.

I should check:

There it is: only 18% of games won on return. That may not even get him on the all time list of return game%, or if so he's going to be close to the end.

That's why Karlovic is only at 92% of games on serve, career. As high as that is, with top defensive skills that number would be more like 95-96%.

Click to expand...

Ljubicic had a great serve, a pretty good BH and was pretty good at the net.
His problems were the FH and movement.

abmk said:

Ljubicic had a great serve, a pretty good BH and was pretty good at the net.
His problems were the FH and movement.

Click to expand...

That would make sense. Statistically weakness in "groundstrokes" means an exploitable weakness on either wing.

The thing that makes all of the Big Three so successful on serve is that they all defend their serves so well. It's pretty obvious that Fed wins most effortlessly serving, but without his great return skills he'd be more like 85% of service games instead of close to 90%, also true of Sampras.

I think we do not consider the whole psychological battle enough regarding neutral rallies. Fed and Sampras always seem/seemed ridiculously confident about winning neutral rallies while serving, not so much on return games. Whereas Nadal thinks he should win every point that way. You can see the mindset of guys like Nadal, Djokovic, Agassi. They expend more energy defending.

Pete was probably sloppiest about trying really high on defense, until later rounds in majors, when he was an absolute beast in comparison to early rounds. This coasting in early rounds and in sets up only one break seems to be typical of great servers.

Gary Duane said:

That would make sense. Statistically weakness in "groundstrokes" means an exploitable weakness on either wing.

The thing that makes all of the Big Three so successful on serve is that they all defend their serves so well. It's pretty obvious that Fed wins most effortlessly serving, but without his great return skills he'd be more like 85% of service games instead of close to 90%, also true of Sampras.

I think we do not consider the whole psychological battle enough regarding neutral rallies. Fed and Sampras always seem/seemed ridiculously confident about winning neutral rallies while serving, not so much on return games. Whereas Nadal thinks he should win every point that way. You can see the mindset of guys like Nadal, Djokovic, Agassi. They expend more energy defending.

Pete was probably sloppiest about trying really high on defense, until later rounds in majors, when he was an absolute beast in comparison to early rounds. This coasting in early rounds and in sets up only one break seems to be typical of great servers.

Click to expand...

I don't agree with regard to Fed. He used to suffocate players with his return game at his peak. He used to win a lot of the neutral rallies.
Big servers used to dread playing him.

On grass, Federer:

Career wise : RPW = 38.4%, Break % = 25.2%
Best 5 years ( 2003-07) : RPW = 40.9%, Break% = 29.2%

http://www.tennisabstract.com/cgi-bin/player.cgi?p=RogerFederer&f=ACareerqqB2
http://www.tennisabstract.com/cgi-b...derer&f=A2007qq,2006qq,2005qq,2004qq,2003qqB2

Nadal :

Career wise : RPW = 36.9%, Break% = 23.8%
Best 5 years (2006-08,10-11) : RPW = 37.2%, Break = 22.9%

http://www.tennisabstract.com/cgi-bin/player.cgi?p=RafaelNadal&f=ACareerqqB2
http://www.tennisabstract.com/cgi-b...Nadal&f=A2011qq,2010qq,2008qq,2007qq,2006qqB2

Its not even close in the best 5 years.
Federer of course has had more years past his prime on grass and is still better by a margin of about 1.5% on each


On HC,
Federer :

Career wise : RPW = 40.1%, Break% = 27.7%
Best 6 years (2004-09) : RPW = 40.6%, Break% = 28.6%

http://www.tennisabstract.com/cgi-bin/player.cgi?p=RogerFederer&f=ACareerqqB0

Nadal :

Career-wise : RPW = 40.4%, Break% = 29.4%
Best 6 years (2008-13) : RPW = 40.9%, Break% = 29.9%

http://www.tennisabstract.com/cgi-bin/player.cgi?p=RafaelNadal&f=ACareerqqB0


Overall career-wise, Nadal's RPW is ahead by only 0.3%, while break % is about 1.7%. This discrepancy is of course thanks to Fed's propensity for squandering BPs.
Best 6 years-wise, the trend is similar.


But now, lets have a look vs top 20 opponents :

Federer:
Career-wise : RPW = 38.7% , Break% = 24.7%
Best 6 years (2004-09) : RPW = 39.5%, Break% = 26.5%

http://www.tennisabstract.com/cgi-bin/player.cgi?p=RogerFederer&f=ACareerqqB0ITop_20qq
http://www.tennisabstract.com/cgi-b...2008qq,2007qq,2006qq,2005qq,2004qqB0ITop_20qq


Nadal :
Career-wise : RPW = 37.4%, Break% = 22.5%
Best 6 years (2008-13) : RPW = 37.6%, Break% = 22.9%


http://www.tennisabstract.com/cgi-bin/player.cgi?p=RafaelNadal&f=ACareerqqB0ITop_20qq
http://www.tennisabstract.com/cgi-b...2012qq,2011qq,2010qq,2009qq,2008qqB0ITop_20qq

federer is ahead on all counts by a significant margin

vs top 10 :

Federer :
career-wise : RPW = 38.4% , Break% = 24.2%
best 6 years (2004-09) : RPW = 39.3%, Break% = 26.2%

http://www.tennisabstract.com/cgi-bin/player.cgi?p=RogerFederer&f=ACareerqqB0ITop_10qq
http://www.tennisabstract.com/cgi-b...2008qq,2007qq,2006qq,2005qq,2004qqB0ITop_10qq

vs top 10 :

Nadal :
career-wise : RPW = 36.2%, Break% = 20.4%
best 6 years (2008-13) : RPW = 36.7%, Break% = 21.7%

http://www.tennisabstract.com/cgi-bin/player.cgi?p=RafaelNadal&f=ACareerqqB0ITop_10qq
http://www.tennisabstract.com/cgi-b...2012qq,2011qq,2010qq,2009qq,2008qqB0ITop_10qq

shifts even more in favour of Federer.


Point being vs players ranked below 20, Nadal plunders them on the return.
Federer doesn't go all out as much. Its not that he always settles for a break a set, he just doesn't try to be as ruthless as Nadal.

But when it comes to the toughest, the biggest players, fed's RPW% and break% are considerably ahead of Nadal's on HC (both in their best years and overall)

Hence, Fed over Nadal returning-wise on HC. (both return and return game)

---

I think the point about Sampras and one break per set gets overblown at times, but I'll expand that in a later post.

Point by point:

abmk said:

I don't agree with regard to Fed. He used to suffocate players with his return game at his peak. He used to win a lot of the neutral rallies.
Big servers used to dread playing him.

On grass, Federer:

Career wise : RPW = 38.4%, Break % = 25.2%
Best 5 years ( 2003-07) : RPW = 40.9%, Break% = 29.2%

Click to expand...

First point: at the moment I don't trust TA because it screwed up really badly in 2017 on some stats, and has not corrected them. Many are high, and that's going to skew up career numbers.

ATP is showing 24.28%. TA 25.2%. That's a big difference.

Sampras is the same on both sites. TA is screwing up on current info.

I'll go with your 29.2% figure for his peak play. I can't check it - no data - but that's about as high as it goes.

At no time have I ever said anything but good things about Sampras on grass. It's his best surface.

It would be nice to check his peak period against 2017, but unless this changes by the time you click on it, it's a mess:

http://www.atpworldtour.com/en/stat...page=1&sortByField=percentage&ascending=False

Right now it won't sort.

It shows Fed last year at: 27.39. TA shows 40.4. That's where the problem is coming from.

On grass, on his favorite surface, Fed was as stubborn about giving up return games during his peak as anyone I've seen, but to be fair you have to chart that against other players, at their peak, and comparing eras is problematical. Edberg has higher career numbers, but in his day players won more return games and fewer service games.

I would absolutely not use Nadal as a comparison, nor Djokovic. The fact is that no one is a fair comparison in this era for Fed on grass.

On clay Nadal beats down everyone returning except, perhaps, peak Coria. He has no rivals in this area.

But I think it's a LOT more complicated on HCs, and I'll get to that later.

Gary Duane said:

The only way you can get even accurate info on games is to scour all the data, for each player, then compute it yourself. I have not yet done that for Edberg, but I can tell you that this looks very wrong to me:

Aces/DFs:

1405 1630

According to this Edberg double faulted more often than he aced.

Back to here:

http://www.atpworldtour.com/en/players/stefan-edberg/e004/player-stats

Check out aces and DFs. Even if we accept the idea that this data is from 1991 on, he was hardly a crap player then. So those stats absolutely do not make any sense at all.

If you are a big Edberg fan, what do you think? No other top player I've checked has more DFs than aces.

It just makes no sense.

Click to expand...

That's entirely believable, actually. I've watched a couple of Edberg matches recently - he really opted for spin and placement over power, generating a lot of forced returned errors and weak volleyable returns, but few direct aces. Given that these stats are only since 1991, and Edberg's serve was the main culprit of his decline since 1993 (even the best volley can only help so much when the serve is getting weaker and weaker), leading to more "double fault days", I believe the stat is right. That said, I think Edberg still should have more aces than DFs for the entire career if we had the stats.

AnOctorokForDinner said:

That's entirely believable, actually. I've watched a couple of Edberg matches recently - he really opted for spin and placement over power, generating a lot of forced returned errors and weak volleyable returns, but few direct aces. Given that these stats are only since 1991, and Edberg's serve was the main culprit of his decline since 1993 (even the best volley can only help so much when the serve is getting weaker and weaker), leading to more "double fault days", I believe the stat is right. That said, I think Edberg still should have more aces than DFs for the entire career if we had the stats.

Click to expand...

When things look very odd, I start checking things:

Total games played on serve, all surfaces, career: 5,452

Breakdown of HC, carpet, grass, clay:

1376+1046+669+869=3960

What surface were the other 1492 service games played on?

When you add up all service games, all different surfaces, the total for Fed does not differ by one.

NEVER TRUST THE ATP WITHOUT CHECKING.

AnOctorokForDinner said:

That's entirely believable, actually. I've watched a couple of Edberg matches recently - he really opted for spin and placement over power, generating a lot of forced returned errors and weak volleyable returns, but few direct aces. Given that these stats are only since 1991, and Edberg's serve was the main culprit of his decline since 1993 (even the best volley can only help so much when the serve is getting weaker and weaker), leading to more "double fault days", I believe the stat is right. That said, I think Edberg still should have more aces than DFs for the entire career if we had the stats.

Click to expand...

When things look very odd, I start checking things:

Total games played on serve, all surfaces, career: 5,452

Breakdown of HC, carpet, grass, clay:

1376+1046+669+869=3960

What surface were the other 1492 service games played on?

When you add up all service games, all different surfaces, the total for Fed does not differ by one.

NEVER TRUST THE ATP WITHOUT CHECKING.

Gary Duane said:

Point by point:

First point: at the moment I don't trust TA because it screwed up really badly in 2017 on some stats, and has not corrected them. Many are high, and that's going to skew up career numbers.

ATP is showing 24.28%. TA 25.2%. That's a big difference.

Sampras is the same on both sites. TA is screwing up on current info.

I'll go with your 29.2% figure for his peak play. I can't check it - no data - but that's about as high as it goes.

At no time have I ever said anything but good things about Sampras on grass. It's his best surface.

It would be nice to check his peak period against 2017, but unless this changes by the time you click on it, it's a mess:

http://www.atpworldtour.com/en/stat...page=1&sortByField=percentage&ascending=False

Right now it won't sort.

It shows Fed last year at: 27.39. TA shows 40.4. That's where the problem is coming from.

On grass, on his favorite surface, Fed was as stubborn about giving up return games during his peak as anyone I've seen, but to be fair you have to chart that against other players, at their peak, and comparing eras is problematical. Edberg has higher career numbers, but in his day players won more return games and fewer service games.

I would absolutely not use Nadal as a comparison, nor Djokovic. The fact is that no one is a fair comparison in this era for Fed on grass.

On clay Nadal beats down everyone returning except, perhaps, peak Coria. He has no rivals in this area.

But I think it's a LOT more complicated on HCs, and I'll get to that later.

Click to expand...

you are right. TA stats are screwed up for 2017.
ATP one is right.
Career break% for Fed on grass is 24.28%
But Nadal's is down as well , 22.37% as per ATP, not 23.8% as per TA.

So the difference is actually a tad more career-wise.

Its not about the comparision of Federer vs Nadal/Djokovic/Murray overall on grass. We know Fed is clearly ahead.
My point is at his peak (say best 5 years), he's ahead of even Djokovic/Murray/Nadal on the return stats as well. The service stats - difference is even more, obviously.

Regarding HC:

best 4 years : 2004-07 for federer (stats really dipped in 2008 and 2009 for him on HC) ..

RPW = 41.5%
Break% = 30.1%

http://www.tennisabstract.com/cgi-bin/player.cgi?p=RogerFederer&f=A2007qq,2006qq,2005qq,2004qqB0

abmk said:

you are right. TA stats are screwed up for 2017.
ATP one is right.
Career break% for Fed on grass is 24.28%
But Nadal's is down as well , 22.37% as per ATP, not 23.8% as per TA.

So the difference is actually a tad more career-wise.

Its not about the comparision of Federer vs Nadal/Djokovic/Murray overall on grass. We know Fed is clearly ahead.
My point is at his peak (say best 5 years), he's ahead of even Djokovic/Murray/Nadal on the return stats as well. The service stats - difference is even more, obviously.

Click to expand...

I never got to the other part.

I would never compare Nadal to other HC greats, like Agassi, Djokovic, others. He's essentially a clay player to this very moment.

I don't know what is going on with the ATP:

http://www.atpworldtour.com/en/stats/return-games-won/2011/hard/all/

It's all gone totally wrong since they changed the format to decimal points, which is beyond absurd.

But let's start with this:

Murray
Djokovic
Chang
Agassi
Edberg
Hewitt
Ferrer
Coria
McEnroe
Lendl
Nadal
Korda
Davydenko
Nalbandian
Federer

Murray on top, career, HCs, around 33% of return games on HCs.

Federer around 27%.

That's a big difference.

If you are picking peaks, peak years:

Chang
Djokovic
Agassi
Chang
Murray
Agassi
Djokovic
Djokovic
Agassi
Agassi
Murray
Djokovic
Agassi
Hewitt
Murray
Agassi
Chang
Chang
Djokovic
Djokovic
Edberg
Hewitt
Sampras
Murray
Hewitt
Murray
Murray
Agassi
Agassi
Murray
Nadal
Korda
Hewitt
Murray
Nadal
Nadal
Djokovic
Agassi
Federer

Chang at 41.5%, Djokovic at 41%.

Federer, 2006, 31.69%

My takeaway: either a lot of guys are better returners at their peaks, or Fed (and Sampras even more so) coasted.

Gary Duane said:

My takeaway: either a lot of guys are better returners at their peaks, or Fed (and Sampras even more so) coasted.

Click to expand...

The coasting hyopthesis seems like a real possibility for a dominant server like Sampras. Though, interesting that Edberg is so high in the returner stats! Given his "weak" FH and all

Sampras/Fed coasting hypothesis:
Not up a break: try hard to get a break
Up a break: don't need to try too hard for break, but if good opportunity (find yourself up 30-0 or 40-15), then go for 2nd break
Up 2 breaks: coast as much as possible on opponent's serve

Would be fun to see % return games won by Sampras/Fed when: not up a break, up 1 break, or up 2 breaks in a set.

IowaGuy said:

The coasting hyopthesis seems like a real possibility for a dominant server like Sampras. Though, interesting that Edberg is so high in the returner stats! Given his "weak" FH and all

Sampras/Fed coasting hypothesis:
Not up a break: try hard to get a break
Up a break: don't need to try too hard for break, but if good opportunity (find yourself up 30-0 or 40-15), then go for 2nd break
Up 2 breaks: coast as much as possible on opponent's serve

Would be fun to see % return games won by Sampras/Fed when: not up a break, up 1 break, or up 2 breaks in a set.

Click to expand...

I keep stressing this: the best players in the world get to 60% of games or close, for careers. I have stats for all the ATGs in the OE, and that holds true.

Double that number, to make it easy.

We are talking about:

95/25
90/30
85/35

And so on.

There is probably about a 4-5% spread between fastest surfaces and slowest, so figure 58% on grass is great, but maybe 62-63% on clay. HCs somewhere in the middle.

Here are the highest examples I have:

Nadal 68% on clay, 2012, with 6 years above 65%

Djokovic, 2011: HC, 63.51%

Djokvic and Agassi top the list, with Fed next in 2006.

Grass is a problem, because grass for a year may reflect only Wimbledon, so Wimbledon winners are very high on the list for years. So peak figures for grass are high, but there are fewer.

I have 16 examples of more than 60%, grass.

26 on HCs.

Probably two screens of years on clay at 60% or higher.

No one has a career of 60% of games on grass. Sampras is a bit over 59%, Fed rounds to 60%. Those two are highest.

So now the big question is this: Why do we not have one player who comes really close to topping the lists of both serving and returning? Why do the very best servers who are also ATGs lag behind a bit on return stats?

My guess is that they do what they need to do to win and put more energy into winning serve, then go up to another gear on return when they get into trouble.

The great returners all have/had weaker serves - Agassi, Murray, even Djokovic, and Nadal most definitely. So they never drop their level much. They try to win every return game as insurance against breaks.

That's my theory. Otherwise we have to conclude that Sampras was a relatively weaker returner, even Fed, and I just don't think that is the whole picture.

Sampras won more games on grass in the last three rounds than anyone else in OE history. Part of that had to reflect ATG returning skills.