Peak level of play (Federer, Nadal, Djokovic & Co.)

corners

Legend
This is the rally: http://www.youtube.com/watch?feature=player_detailpage&v=4T4EqWC1WkI#t=5218s

I wouldn't call that a simple unforced error; I could see it being scored either as forced or unforced.

Djokovic is changing the direction of the ball by going DTL, not a percentage play. He's very good at that particular shot and I'm sure you've seen him do damage to Federer and other players with it. But it's not percentage; and he's trying to make something happen when he uses it. I mean more than throwing in a slice to change the pace a little: I mean trying to make his opponent run after an unexpected shot.

Federer lost his balance there a little and made a wild shot -- possibly because he saw Djoker in that corner ready to receive any DTL shot Federer might give him. So maybe he tried to go too close to the line, or changed his mind at the last minute. Who knows -- but I think Djokovic outmaneuvered Federer on that point. He covered the space where his opponent wanted to put the ball -- kind of like a net player covering the sideline and forcing a defender's passing shot to go wide.

Anyway he did make Federer run, and that's one of the factors usually mentioned when statisticians talk about how to score UE's. I would also consider things like a player changing the direction of the rally to go DTL (not a percentage play).

This to me is a simple UE: http://www.youtube.com/watch?feature=player_detailpage&v=4T4EqWC1WkI#t=6073s

Any thoughts from others on this?

I would have called it an unforced error. Yes, Djoker had him moving. But Fed went for the line and missed. He didn't need to, he could have played it elsewhere. But he went for the line and missed. To me, if a player goes for a winner and hits it out it should be an unforced error. (Is this assumption incorrect?)

Also, although he was moving to the ball his footwork was not what we would see on a very wide ball that he could barely get to. Technically, it was not hit "on the run." Just after hitting that shot he was already back on balance and would have recovered position, ready if Djokovic managed to return the down the line shot.

Also, I would call this a fatigue error. I can't find the reference right now, but accuracy is severely impacted due to motor control impairment once a person gets winded. I think Fed was getting fatigued, got a good look at the line and had the choice of staying in the rally and doing some more running or taking a shot at the line. He took it and missed. Unforced error in my book, as a result of fatigue, with deference to the experts.

On the other hand, if the fatigue error hypothesis is correct, one could say that Djokovic made him run in the rally, fatiguing him, and thereby forced the error. But Djokovic was also being made to run, but didn't miss. In my opinion this is one of the reasons Djokovic would prefer to get Fed into long rallies and why Fed would have liked to pull the trigger prior to this.

Haha. Then, to the contrary, one could say that Djoker forced the error by pulling Fed into a long rally where he knew his own fatique would handicap his style less than it would handicap Federer's.

(This may be an example of that player bias we were talking about before. I've seen Fed make that shot before, and miss it before. I expect him to make it off the ball Djokovic gave him in this instance. Maybe that expectation is unrealistic. Maybe this would be a shot where Federer would say, "Man, you don't understand how difficult that is! Yeah, I make it sometimes, but that's because I'm Federer!")

Anyway, my two cents. I'll consider any responses as part of my eduction as a novice box score man. I had no idea errors were this complicated.
 
Last edited:

krosero

Legend
I would have called it an unforced error. Yes, Djoker had him moving. But Fed went for the line and missed. He didn't need to, he could have played it elsewhere. But he went for the line and missed. To me, if a player goes for a winner and hits it out it should be an unforced error. (Is this assumption incorrect?)

Also, although he was moving to the ball his footwork was not what we would see on a very wide ball that he could barely get to. Technically, it was not hit "on the run." Just after hitting that shot he was already back on balance and would have recovered position, ready if Djokovic managed to return the down the line shot.

Also, I would call this a fatigue error. I can't find the reference right now, but accuracy is severely impacted due to motor control impairment once a person gets winded. I think Fed was getting fatigued, got a good look at the line and had the choice of staying in the rally and doing some more running or taking a shot at the line. He took it and missed. Unforced error in my book, as a result of fatigue, with deference to the experts.

On the other hand, if the fatigue error hypothesis is correct, one could say that Djokovic made him run in the rally, fatiguing him, and thereby forced the error. But Djokovic was also being made to run, but didn't miss. In my opinion this is one of the reasons Djokovic would prefer to get Fed into long rallies and why Fed would have liked to pull the trigger prior to this.

Haha. Then, to the contrary, one could say that Djoker forced the error by pulling Fed into a long rally where he knew his own fatique would handicap his style less than it would handicap Federer's.

(This may be an example of that player bias we were talking about before. I've seen Fed make that shot before, and miss it before. I expect him to make it off the ball Djokovic gave him in this instance. Maybe that expectation is unrealistic. Maybe this would be a shot where Federer would say, "Man, you don't understand how difficult that is! Yeah, I make it sometimes, but that's because I'm Federer!")

Anyway, my two cents. I'll consider any responses as part of my eduction as a novice box score man. I had no idea errors were this complicated.
All good points. I think this rally is a good one to talk about because it's one of those baseline rallies that are so characteristic of today's game: a long, fatiguing rally, not like Borg-Vilas, but a rally in which nearly every shot is aggressive and forcing. So you've got this long point where most of the strokes were aggressive, and taking their cumulative effect. Those aggressive blows have to have an effect on the final shot.

I wouldn't say that fatigue, by itself, should make us score an error as forced. But in this rally you've got fatigue plus a rapid-fire series of events. I agree with the way you put it, that technically Federer was not on the run; not on the sprint. But he was rushed; the ball arrived quickly and in a place where Federer did not expect it, since they had been exchanging crosscourt shots.

Of course on another day he makes that shot; but you made a great point about player bias.

On the assumption you mentioned: in a net point, a defender going for a passing shot winner is always scored as making a forced error. So is that similar in any way to a baseline rally? Well, if a player goes for a winner in a wide open space, in a baseline rally, and he misses -- that's a clear UE. But if he's trying to put the ball into a spot that the opponent has made into a tight space, I think it's ambiguous. In such a case, I'd say UE if the player was not rushed in any way, or it was not a long point. On this Fed-Djok point, I'm honestly going both ways on it :)
 

abmk

Bionic Poster
This is the rally: http://www.youtube.com/watch?feature=player_detailpage&v=4T4EqWC1WkI#t=5218s

I wouldn't call that a simple unforced error; I could see it being scored either as forced or unforced.

Djokovic is changing the direction of the ball by going DTL, not a percentage play. He's very good at that particular shot and I'm sure you've seen him do damage to Federer and other players with it. But it's not percentage; and he's trying to make something happen when he uses it. I mean more than throwing in a slice to change the pace a little: I mean trying to make his opponent run after an unexpected shot.

Federer lost his balance there a little and made a wild shot -- possibly because he saw Djoker in that corner ready to receive any DTL shot Federer might give him. So maybe he tried to go too close to the line, or changed his mind at the last minute. Who knows -- but I think Djokovic outmaneuvered Federer on that point. He covered the space where his opponent wanted to put the ball -- kind of like a net player covering the sideline and forcing a defender's passing shot to go wide.

Anyway he did make Federer run, and that's one of the factors usually mentioned when statisticians talk about how to score UE's. I would also consider things like a player changing the direction of the rally to go DTL (not a percentage play).

This to me is a simple UE: http://www.youtube.com/watch?feature=player_detailpage&v=4T4EqWC1WkI#t=6073s

Any thoughts from others on this?

yeah, the thing is in that shot, the pace/depth combo of the BH DTL from djokovic wasn't enough to call it forced .... djokovic had him moving, but it was not a shot on the run, he wasn't off-balance mainly due to djoker's shot, it was sloppy footwork from federer and wrong execution of the FH DTL ...

the 2nd link you gave of course is a simple UE ...
 

abmk

Bionic Poster
I just put your stats together here for comparing the two matches.

On service returns: by your count Murray made 4 less UE's than Djokovic. That would mean that Murray's superiority over Djokovic as a returner actually cut down the total UE's in the final and raised the AM's of the match.

Overall: using your total count of UE's, the two matches were of similar quality, the semifinal only slightly higher.

I'm glad you took the UE's because it shows how much discrepancy there can be between any two statisticians.

I don't think the problem is the concept of the UE as such. The concept is fine; but there is no standardization of stats in tennis. No universal conception of how UE's should be counted.

As you know, I've seen endless problems even with Winners, which you would think are basically straightforward. I've seen the term used to refer to clean aces and winners the way most of us use it. But I've also seen boxscores, in this era, that use the term to mean the number of groundstroke winners, no volleys included, no service included (as they did at the Olympics in '08 ). RG once used "Winners" to mean all points won on serve. And though most statisticians include service when reporting the total winners, there is still discrepancy because there are multiple definitions of what a service winner is.

Lack of standardization is a major problem in tennis stats.

I don't know what those 'return winners' at Wimbledon.com refer to.

One guess: maybe they refer to all returns that went past the server cleanly (what you counted, ABMK) or forced an error from the server. I would call all those returns "winning returns." It's a little like the concept of the service winner: you're counting all winning serves, whether they go by cleanly or they force the receiver into an error.

yeah, even for winners, there can be discrepancies , let alone UEs ...

regarding the return winners,

IMO the service winner analogy isn't appropriate, in a service winner, you normally have the returner barely getting the racquet on the ball ....

forcing an error from the server with the return would be comparable to a forced return error ...

maybe a bit off topic, but aces+forced errors off the return would be a more useful parameter than just aces ....

now coming back to the semi and final matches of Wimbledon 2012, having watched them both again, I think I may have under-estimated the quality of the semi and bit and over-estimated that of the final a bit ... Still think the final was the better quality one, but not by that much ...

on the return unforced errors, yes, murray's better returning did cut down on them, but in total , the semi and final had the same no of UEs on the return - 11 ( due to federer going for more on the murray 2nd serve ) ... murray getting back plenty more 1st serves than djoker did had more of an impact ....
 
Last edited:

krosero

Legend
yeah, the thing is in that shot, the pace/depth combo of the BH DTL from djokovic wasn't enough to call it forced .... djokovic had him moving, but it was not a shot on the run, he wasn't off-balance mainly due to djoker's shot, it was sloppy footwork from federer and wrong execution of the FH DTL ...
I can see it being called an UE. But something else makes it ambiguous for me: Federer is not in control when he makes the shot. It's not like when he's running around his BH and going for the kill, and misses. Here he's reacting.

That's something else I'm likely to ask when counting UE's... is the player in control, or is he reacting?

regarding the return winners,

IMO the service winner analogy isn't appropriate, in a service winner, you normally have the returner barely getting the racquet on the ball ....

forcing an error from the server with the return would be comparable to a forced return error ...
I should have specified what I meant by 'service winner' in this instance. I was just referring broadly to all unreturned serves (including aces). I was not using the more narrow definition of serves barely touched.

Anyway in this case my emphasis is not on the analogy. The "Return Winners" just reminded me of the broad category of "Service Winners". My guess about the "Return Winners" at Wimbledon.com is just that they are a count of all clean return winners + returns that forced an error. Generally speaking, that would be the reverse of 'service winners': instead of counting serves that are 'winners', you're counting returns that are 'winners.'

Just a guess, I could be way off.
 
Last edited:

krosero

Legend
Nadal and Djokovic on three different hardcourts.

(As always, winner listed above the loser, regardless of AM).

2010 USO
Nadal - 22.8%
Djokovic - 16.1%

2011 USO
Djokovic - 21.6%
Nadal - 12.7%

2012 AO
Djokovic - 14.4%
Nadal - 9.8%

2011 Miami
Djokovic - 16.1%
Nadal - 9.8%
 
Last edited:

krosero

Legend
Nadal and Djokovic on various claycourts

2007 RG
Nadal - 28.7%
Djokovic - 11.6%

2008 RG
Nadal - 32.3%
Djokovic - 23.2%

2012 RG
Nadal - 17.8%
Djokovic - 14.1%

2012 Monte Carlo
Nadal - 22.1%
Djokovic - (3.4%) (negative)

2012 Rome
Nadal - 9.8%
Djokovic - 3.5%

2011 Rome
Djokovic - 11.5%
Nadal - 5.4%

2011 Madrid
Djokovic - 14.3%
Nadal - 4.5%

2009 Madrid
Nadal - 11.2%
Djokovic - 13.1% (Novak leading in total pts. won 125-120)
 
Last edited:

abmk

Bionic Poster
I can see it being called an UE. But something else makes it ambiguous for me: Federer is not in control when he makes the shot. It's not like when he's running around his BH and going for the kill, and misses. Here he's reacting.

That's something else I'm likely to ask when counting UE's... is the player in control, or is he reacting?

yeah, reacting , but only a bit, djoker didn't really hit with that much pace/depth that it would force an error without mistake on the part of fed ...

the error happened in major part because of sloppy footwork and going for too much on the FH ...

if we started counting UEs only when the player was fully in control, then we'd have very few UEs ...

I might post a couple of examples where I think it is ambiguous whether it is an unforced error or a forced error in some time ...

the way I see it :

1) shot that makes you think - that was an unforced error ..

2) shot that make you think - that was an unforced error , looks like ... 'maybe' should think about it, watch it once more ..

3 ) shot that make you think, ah, can't decide whether that was unforced or forced ... not easy to say .. might have to watch it more than once ..

that shot in that rally IMO falls in category 2, still clearly an unforced error ... the link to the other missed FH you gave IMO falls in category 1 ...


I should have specified what I meant by 'service winner' in this instance. I was just referring broadly to all unreturned serves (including aces). I was not using the more narrow definition of serves barely touched.

Anyway in this case my emphasis is not on the analogy. The "Return Winners" just reminded me of the broad category of "Service Winners". My guess about the "Return Winners" at Wimbledon.com is just that they are a count of all clean return winners + returns that forced an error. Generally speaking, that would be the reverse of 'service winners': instead of counting serves that are 'winners', you're counting returns that are 'winners.'

Just a guess, I could be way off.

could be, but I didn't really count that ...
 
Last edited:

corners

Legend
2013 AO AMs for top players, through 3rd round

2094k6g.jpg
 

krosero

Legend
I've never been able to get UE for Kuerten-Federer at RG but I've found them, and a few other stats, at http://tennis.matchstat.com/index.p...O+W&Name_Player2=Gustavo+Kuerten&Id_Player2=1

2004 RG
Kuerten - 24.7%
Federer - 18.3%

It was a well played match, though there's something funny about the winner/error differentials. Federer's (+8 ) is higher than Kuerten's (+4): so if Kuerten did not get ahead by winners and UE's alone, he must have gotten a large margin over Federer in the category of forced errors.
 

krosero

Legend
Some others for Kuerten.

1997 RG final
Kuerten - 12.3%
Bruguera - (-1.5) (negative)

2000 RG final
Kuerten - 0.0%
Norman - (-0.6%) (negative)

2001 RG final
Kuerten - 5.7%
Corretja - (-5.7) (negative)

I wonder if the clay at RG was slower in those years than it was in 2004 or in recent years?
 

krosero

Legend
that's mathematically impossible given nadal won 95 points and djoker won 79 points in that match .... (ATP site ) !
I typed out the wrong year. I meant the 2012 MC final which Nadal won 6-3, 6-1.

Another good catch, ABMK. Keep letting me know if something looks off -- I have all these AM's but I've never had the chance to really proof them.
 

abmk

Bionic Poster
I typed out the wrong year. I meant the 2012 MC final which Nadal won 6-3, 6-1.

Another good catch, ABMK. Keep letting me know if something looks off -- I have all these AM's but I've never had the chance to really proof them.

yeah, that's what I thought ...
 

abmk

Bionic Poster
I've never been able to get UE for Kuerten-Federer at RG but I've found them, and a few other stats, at http://tennis.matchstat.com/index.p...O+W&Name_Player2=Gustavo+Kuerten&Id_Player2=1

2004 RG
Kuerten - 24.7%
Federer - 18.3%

It was a well played match, though there's something funny about the winner/error differentials. Federer's (+8 ) is higher than Kuerten's (+4): so if Kuerten did not get ahead by winners and UE's alone, he must have gotten a large margin over Federer in the category of forced errors.

heh, thanks for that mentioning about that site. I'd seen that site, but I hadn't noticed that the graph icon was a hyperlink to the match stats.

I am not that surprised at federer's W-UE being better . Its happened on other occasions as far as fed is concerned ,

ex : in the RG 2010 QF,

fed was +13 and soderling was +7

in the wimbledon 2008 final,

fed was +37 and nadal was +33
 

krosero

Legend
Two night session matches late in Agassi's career, both five-setters.

2005 USO
Agassi - 18.6%
Blake - 16.7%

2006 USO
Agassi - 14.6%
Baghdatis - 10.9%

Nothing too surprising here. Agassi's number is lower in '06 as you would expect.

But the winner/error differentials are strange: in each match Agassi's was much poorer than the differential of the loser.

2006 W/E diff
Agassi (–13)
Baghdatis (–3)

2005 W/E diff
Agassi (–1)
Blake +14
 

abmk

Bionic Poster
From tennis.matchstat.com, I got the stat for one match I was searching for for quite some time :

federer ancic wimbledon 2006 QF:

federer : 35 winners, 17 UEs, 92 points won
ancic : 24 winners, 13 UEs, 75 points won

federer's AM ~ 37.1%
ancic's AM ~ 27%

This was a brilliant, highly under-rated match ... the scoreline of 6-4,6-4,6-4 does no justice to how well Ancic played . Federer was just too good and had all the answers ...
 

abmk

Bionic Poster
Another match from the same tourney ,

federer-berdych wimbledon 2006 4R :

federer : 47 W, 15 UEs, 93 points won
berdych : 19 W, 20 UEs, 64 points won

federer's AM ~ 37%
berdych's AM ~ 18.5%
 

krosero

Legend
2006 Wimbledon
Nadal - 41.2%
Agassi - 28.9%

This is the highest AM we have for Nadal apart from the '08 Wimb QF over Murray. And Agassi's AM is pretty good, considering it's '06.

I knew Nadal played better grasscourt tennis in '06 than he is often given credit for, though I was not expecting to find an AM over 40%.
 

corners

Legend
2006 Wimbledon
Nadal - 41.2%
Agassi - 28.9%

This is the highest AM we have for Nadal apart from the '08 Wimb QF over Murray. And Agassi's AM is pretty good, considering it's '06.

I knew Nadal played better grasscourt tennis in '06 than he is often given credit for, though I was not expecting to find an AM over 40%.

Yeah, that's pretty surprising.
 

piece

Professional
2006 Wimbledon
Nadal - 41.2%
Agassi - 28.9%

This is the highest AM we have for Nadal apart from the '08 Wimb QF over Murray. And Agassi's AM is pretty good, considering it's '06.

I knew Nadal played better grasscourt tennis in '06 than he is often given credit for, though I was not expecting to find an AM over 40%.

Funny, when abmk initially asked what match it might be that Nadal had an AM over 40 for, I wondered whether it would be a match from 2006 Wimbledon. I remember, for instance, that Nadal hit over 20 winners to 3 unforced errors (or thereabouts) in one of the sets he won against Federer in the final.
 

krosero

Legend
2006 Wimb SF
Nadal - 35.0%
Baghdatis - 24.7%

Again this is very high; on grass I have only two AM's higher for Nadal, both in '08, against Youzhny and Murray.

I don't think this is an accident or a quirk of the stats. Unforced errors, of course, will always be subjective, but Nadal also has objective stats that are off the charts at this Wimbledon.

I'm referring to his streak of 80 consecutive service holds, broken in the final. The record is 118 straight by Sampras. Since Pete's time there have been only a few streaks at Wimbledon better than Nadal's: Federer had 105 straight holds stretching across the 2003-04 Wimbledons; Karlovic got 80 straight in '09; and Isner and Mahut each got 85 straight in their match.

Nadal, with his serve, is the unlikeliest candidate to post such a streak. But these AM's show that he was playing at a very high level -- maybe playing "out of his mind," or zoning.

His stats so far support the argument that he reached his grasscourt peak in '08 and that in '06 he was already closer to that peak than is generally assumed. In the '06 final which he lost in 4 sets, he produced an AM of only 22.4%, but that was partly due to how nervous he was at the start, when he got bageled.
 

corners

Legend
2006 Wimb SF
Nadal - 35.0%
Baghdatis - 24.7%

Again this is very high; on grass I have only two AM's higher for Nadal, both in '08, against Youzhny and Murray.

I don't think this is an accident or a quirk of the stats. Unforced errors, of course, will always be subjective, but Nadal also has objective stats that are off the charts at this Wimbledon.

I'm referring to his streak of 80 consecutive service holds, broken in the final. The record is 118 straight by Sampras. Since Pete's time there have been only a few streaks at Wimbledon better than Nadal's: Federer had 105 straight holds stretching across the 2003-04 Wimbledons; Karlovic got 80 straight in '09; and Isner and Mahut each got 85 straight in their match.

Nadal, with his serve, is the unlikeliest candidate to post such a streak. But these AM's show that he was playing at a very high level -- maybe playing "out of his mind," or zoning.

His stats so far support the argument that he reached his grasscourt peak in '08 and that in '06 he was already closer to that peak than is generally assumed. In the '06 final which he lost in 4 sets, he produced an AM of only 22.4%, but that was partly due to how nervous he was at the start, when he got bageled.

If I remember right that first set featured lots of Federer sliced returns that landed right at the service mark, which really seemed to throw off Nadal, like he didn't know what to do with that ball. TV cut to Roche several times and I remember thinking, "Clever old fox, you." That was probably the last time Fed received dividends for hitting slices in large doses to Nadal.
 

abmk

Bionic Poster
If I remember right that first set featured lots of Federer sliced returns that landed right at the service mark, which really seemed to throw off Nadal, like he didn't know what to do with that ball. TV cut to Roche several times and I remember thinking, "Clever old fox, you." That was probably the last time Fed received dividends for hitting slices in large doses to Nadal.

that was a feature of that whole match, not just the first set ..

nadal committed quite a few errors of his FH wing because of that ... his BH however was on fire and he hit loads of winners with it ..
 

corners

Legend
that was a feature of that whole match, not just the first set ..

nadal committed quite a few errors of his FH wing because of that ... his BH however was on fire and he hit loads of winners with it ..

Thanks. I'll have to go back and watch that one. Haven't seen it in over six years :)

Do you think it was immediately after that match that Toni and Rafa resolved to attack the slice mercilessly? Nowadays, Rafa seems utterly committed to attacking the slice with the forehand whenever he sees it, and with the spin he can produce he can rip the short slice up over the net and into a corner like no one else. He's nearly eliminated that shot from Roger's arsenal when they play, further torturing Fed's backhand side. I also remember reading a piece by Steve Tignor in which he relates walking by a practice court at Indian Wells and seeing Rafa ripping backhands off short slices too (this was maybe 3-4 years ago). The domination of your backhand is now complete, no?
 

krosero

Legend
What's fun about this is when the finalists emerge, we can see if the one with the higher AM's wins.

I have not heard of the AM method being used that way (to predict the winner of a tournament). And if we do it by adding up the percentages and getting an average for each player, I'm not certain how exact that can be. But let's see what we get.
 

krosero

Legend
2007 Hamburg
Federer - 13.2%
Nadal - 1.3%

2008 Hamburg
Nadal - 20.1%
Federer - 16.9%

Nadal's level of play was quite low in '07 when he lost to Federer (getting bageled in the final set). It's the only match I know of in which Nadal's unforced errors were higher than Federer's (34-31).

Federer himself arguably played better in '08 than he did the previous year. That is apparent in the AM's, and supported by the fact that he made only 52% of his first serves in '07.

But Nadal won the rematch because he was back up at a level at which he's probably never lost on clay.

(Rafa was at 15% when he lost to Soderling, and at 5% in each of his losses to Novak at Madrid and Rome).
 
Last edited:

krosero

Legend
2009 Madrid
Federer - 18.2%
Nadal - 14.1%

2010 Madrid
Nadal - 14.8%
Federer - 14.2%

This mirrors the Hamburg matches a bit. The numbers show Nadal playing at roughly the same level both years, but Federer won in '09 because his level was so high.
 

corners

Legend
What's fun about this is when the finalists emerge, we can see if the one with the higher AM's wins.

I have not heard of the AM method being used that way (to predict the winner of a tournament). And if we do it by adding up the percentages and getting an average for each player, I'm not certain how exact that can be. But let's see what we get.

Yeah, for all we know AMs are a staple of bookies.

As the first two sets of Ferrer-Almagro concluded last night I thought to myself, "This, given Almagro's ~9% average edge over Ferrer in earlier matches, could have been predicted." The AM's were 15.5% for Almagro and 6.4% for Ferrer through two sets. Then the crisis hit Almagro, throwing all numbers out the window. Sad match to watch. Full match was 12.4% for Ferrer, so he finished close to the average of his last two matches, while Almagro dropped well below his average for the tournament, to 7.6%.

It will be interesting to see if Djokovic's AMs stay so high as he starts facing tougher opposition and has to play more defense.

Going by the AMs, Murray would look to be in worse form than the others in the lower half of the draw, and that Chardy might have a chance against him, if his form holds. But my guess is that Chardy's level will drop in the quarters or that Murray will drop it for him, whatever that means.

One of the interesting things for me with this was watching Tomic's AM's leading up to his clash with Federer. I checked out some of the banter in the current pro sub-forum on that match and a lot of the debate was about whether Tomic really was playing at a level that would threaten the top guys. The AM's showed that he clearly was, in fact his average AM through two rounds was top-5 level. I felt like I had something objective to go by in assessing his form and level, which was kind of cool especially as I haven't watched him enough to have a good sense of how good he really is. And even in defeat to Federer he posted a 16.2% AM, second only to DelPotro's 16.9% among match losers (that I've charted). If hed oesn't drop off a cliff as he did last year after Melbourne, this small sample of successful aggression suggests he is definitely top-10 material.
 
Last edited:

krosero

Legend
Yeah, for all we know AMs are a staple of bookies.

As the first two sets of Ferrer-Almagro concluded last night I thought to myself, "This, given Almagro's ~9% average edge over Ferrer in earlier matches, could have been predicted." The AM's were 15.5% for Almagro and 6.4% for Ferrer through two sets. Then the crisis hit Almagro, throwing all numbers out the window. Sad match to watch. Full match was 12.4% for Ferrer, so he finished close to the average of his last two matches, while Almagro dropped well below his average for the tournament, to 7.6%.

It will be interesting to see if Djokovic's AMs stay so high as he starts facing tougher opposition and has to play more defense.

Going by the AMs, Murray would look to be in worse form than the others in the lower half of the draw, and that Chardy might have a chance against him, if his form holds. But my guess is that Chardy's level will drop in the quarters or that Murray will drop it for him, whatever that means.

One of the interesting things for me with this was watching Tomic's AM's leading up to his clash with Federer. I checked out some of the banter in the current pro sub-forum on that match and a lot of the debate was about whether Tomic really was playing at a level that would threaten the top guys. The AM's showed that he clearly was, in fact his average AM through two rounds was top-5 level. I felt like I had something objective to go by in assessing his form and level, which was kind of cool especially as I haven't watched him enough to have a good sense of how good he really is. And even in defeat to Federer he posted a 16.2% AM, second only to DelPotro's 16.9% among match losers (that I've charted). If hed oesn't drop off a cliff as he did last year after Melbourne, this small sample of successful aggression suggests he is definitely top-10 material.
The numbers at the end will be interesting, though we know of many examples where one finalist has seemed to be in clearly better form than the other, but then an upset occurs (Ashe-Connors is a great example). The AM's, in that case, should reflect how well the upset finalist had been playing in the first six rounds; they should point to a very high level.

Who knows, but the exact numbers will be interesting, and may differ from our own impressions. We'll see.

Tomic had an AM of 33.6% in the first round. On your chart, outside of Djokovic, Murray and Del Potro, nobody in the first four rounds was able to get anything higher. So yeah, great potential there.

Federer's AM's against Tomic and Raonic were nearly identical. I thought the first match was more interesting and better played, just in the sense that Tomic was a better opponent than Raonic (who honestly bored me whenever he wasn't in his service motion); and that's reflected in the AM's.
 

corners

Legend
2013 AO AMs, through Quarters

f58wuf.jpg


Four men left, the top four seeds.

A breakdown of average AMs for these four players over the past five rounds, alongside the average AMs of their opponents:

Djokovic: 25.6%
Opponents: 10.26%

Federer: 24.0%
Opponents: 11.7%

Murray: 23.64%
Opponents: 3.84%

Ferrer: 16.7%
Opponents: 4.0%

These numbers are interesting but I think they could be interpreted in a variety of ways. Clearly, the Big 3 are playing more aggressively and with more success than Ferrer. Djokovic's average is substantially above Ferrer's, which aligns with the general expectation that Djokovic will win their semifinal with relative ease.

On the bottom of the draw, Federer and Murray's AM averages are nearly identical. But Federer's opponents have posted much better AMs than Murray's opponents. Is this a consequence of Murray's weak draw and weak opposition? Or does it reflect differences in playing style - Murray more willing to engage in long baseline rallies where his opponent will tend to make an error, while Federer looking to shorten points?

It's interesting that the average AMs of both Murray and Ferrer's opponents are very similar, 3.84% to 4.0%. Again, is this a reflection of poorly-playing opposition or the counterpunching/grinding style of the 3rd and 4th seeds? Given that the 2005 whole-tournament average was around 10%, it seems unlikely that Murray and Ferrer faced opposition through five rounds that were playing at a level significantly below that average. However, a look on a match-by-match basis shows that Murray's opponents, from the 1st round to the Quarters, all posted low numbers. Conversely, Ferrer's 1st, 2nd and Quarterfinal opponents posted respectable AMs while his 3rd and 4th round opponents, both generally highly regarded on tour, posted negative AMs. What does this mean?

By contrast, the average of AMs of Federer and Djokovic's opponents were both right around the 2005 tournament average. Both men also posted AMs of their own in the mid-24s, similar to Murray. Again, are the high average AMs of their opponents a reflection of the playing style of the victor and/or the opponent? Or, more reflective of a higher level of play by those opponents?

4 out of 5 of Federer's opponents I would characterize as aggressive players in the context of playing styles today, so the high AM's posted by these players could be a reflection of that style. If this were the case, Federer's much higher AM average would suggest that he dominated these matches, which, excepting the Quarterfinal with Tsonga, was indeed the case. If this were true I think we could form similar conclusions about Djokovic's numbers, as at least three of his opponents I would characterize as aggressive.

So what does it all mean for the semifinal matchups? :-?
 

krosero

Legend
The Hamburg match that Nadal lost to Federer (top of page) is the only match I can find in which Nadal made more unforced errors than Federer.

I have the official UE's for all their matches except:

2004 Miami
2006 Dubai
2006 Masters Cup
2011 Madrid
 

krosero

Legend
2008 Wimb
Murray - 30.4%
Gasquet - 28.0%

2011 Wimb
Murray - 42.4%
Gasquet - 32.8%

This is by far the highest AM we have for Murray now (higher than anything we have for Djokovic).

It may reflect how Murray has tended to underperform on the biggest stages, because when he lost to Nadal in the '11 semis his AM was only 24.1%.

I don't know whether Murray's AM against Gasquet at 2011 Wimby was the highest achieved by anyone at that tournament, but at 42.4% it certainly could be. Djokovic, who won the tournament, was in the high 30s in his last two matches. Nadal reached the final, and in his last two matches he was in the low 30s. Federer was at 40.1% in his loss to Tsonga (who posted 36.5% in that same match).
 

krosero

Legend
2013 AO
Federer - 18.3%
Tsonga - 16.8%

2010 AO
Federer - 32.4%
Tsonga - 10.1%

(Yesterday's match went five sets, the 2010 match was a blowout).
 

abmk

Bionic Poster
so fed-tsonga was better match quality wise when compared to djoker-stan going by the AMs .....not by a small margin ... question is how much better was it and how much of the difference was caused by djoker's insane defense ?
 

krosero

Legend
Four men left, the top four seeds.

A breakdown of average AMs for these four players over the past five rounds, alongside the average AMs of their opponents:

Djokovic: 25.6%
Opponents: 10.26%

Federer: 24.0%
Opponents: 11.7%

Murray: 23.64%
Opponents: 3.84%

Ferrer: 16.7%
Opponents: 4.0%
The AM's illustrated nicely how close the Big 3 are, and how far below Ferrer is. I did not expect the AM's to illustrate that situation so ideally.

Too bad Nadal's not around, I wonder if he would have posted a number right in there in the mid-20s.

These numbers are interesting but I think they could be interpreted in a variety of ways. Clearly, the Big 3 are playing more aggressively and with more success than Ferrer. Djokovic's average is substantially above Ferrer's, which aligns with the general expectation that Djokovic will win their semifinal with relative ease.

On the bottom of the draw, Federer and Murray's AM averages are nearly identical. But Federer's opponents have posted much better AMs than Murray's opponents. Is this a consequence of Murray's weak draw and weak opposition? Or does it reflect differences in playing style - Murray more willing to engage in long baseline rallies where his opponent will tend to make an error, while Federer looking to shorten points?

It's interesting that the average AMs of both Murray and Ferrer's opponents are very similar, 3.84% to 4.0%. Again, is this a reflection of poorly-playing opposition or the counterpunching/grinding style of the 3rd and 4th seeds? Given that the 2005 whole-tournament average was around 10%, it seems unlikely that Murray and Ferrer faced opposition through five rounds that were playing at a level significantly below that average. However, a look on a match-by-match basis shows that Murray's opponents, from the 1st round to the Quarters, all posted low numbers. Conversely, Ferrer's 1st, 2nd and Quarterfinal opponents posted respectable AMs while his 3rd and 4th round opponents, both generally highly regarded on tour, posted negative AMs. What does this mean?

By contrast, the average of AMs of Federer and Djokovic's opponents were both right around the 2005 tournament average. Both men also posted AMs of their own in the mid-24s, similar to Murray. Again, are the high average AMs of their opponents a reflection of the playing style of the victor and/or the opponent? Or, more reflective of a higher level of play by those opponents?

4 out of 5 of Federer's opponents I would characterize as aggressive players in the context of playing styles today, so the high AM's posted by these players could be a reflection of that style. If this were the case, Federer's much higher AM average would suggest that he dominated these matches, which, excepting the Quarterfinal with Tsonga, was indeed the case. If this were true I think we could form similar conclusions about Djokovic's numbers, as at least three of his opponents I would characterize as aggressive.

So what does it all mean for the semifinal matchups? :-?
All good questions, no easy answers. But a few points.

- The AO courts may be playing slower than they were in 2005. Plexicushion is said to be slower than Rebound Ace; and there's some evidence (brought up by ABMK) that even the Plexicushion itself has been slowed down in the last few years.

- Federer's opponents have high AM's but that may simply be due to the fact that they're good players, rather than due to their styles. Federer's draw was generally regarded as full of dangerous players.

The styles issue is not an easy one but I'm skeptical that high AM's are due to aggressive styles. In this thread we've seen countless AM's that are high, but produced by "grinders" or consistent baseliners who kept their UE's down.

If merely having an aggressive style should lead more readily to having a high AM, we would expect Federer's AM in this tournament to be significantly ahead of Djokovic's and Murray's, since Fed has the most aggressive style of the three men. Djokovic, too, is more aggressive than Murray. But all three players are bound up closely together in their AM's.
 

World Beater

Hall of Fame
The AM's illustrated nicely how close the Big 3 are, and how far below Ferrer is. I did not expect the AM's to illustrate that situation so ideally.

Too bad Nadal's not around, I wonder if he would have posted a number right in there in the mid-20s.

All good questions, no easy answers. But a few points.

- The AO courts may be playing slower than they were in 2005. Plexicushion is said to be slower than Rebound Ace; and there's some evidence (brought up by ABMK) that even the Plexicushion itself has been slowed down in the last few years.

- Federer's opponents have high AM's but that may simply be due to the fact that they're good players, rather than due to their styles. Federer's draw was generally regarded as full of dangerous players.

The styles issue is not an easy one but I'm skeptical that high AM's are due to aggressive styles. In this thread we've seen countless AM's that are high, but produced by "grinders" or consistent baseliners who kept their UE's down.

If merely having an aggressive style should lead more readily to having a high AM, we would expect Federer's AM in this tournament to be significantly ahead of Djokovic's and Murray's, since Fed has the most aggressive style of the three men. Djokovic, too, is more aggressive than Murray. But all three players are bound up closely together in their AM's.

Perhaps i dont understand AMs well enough, but isnt the AM of a player dependent on the opponent i.e. it is relative only to the opponent to the other side.

I suppose what one can deduce is whether players are performing "well" as compared to how they usually play. The AM can be used as a benchmark in matchups and whether they had a good day relatively speaking to past matches.

But to compare djokovic to federer for example..there is a small difference in AM. Is this because federer's opponents were better? They certainly performed "better" based on AMs, but this is relative to federer of course. As they can only play federer's shots not djokovic's.

Perhaps what we need is a Delta AM - that is the differene between the player and the opponent. That way we can potentially cancel out stylistic differences that would lead aggressive players to have a higher absolute AM than defense oriented players...

This way we have a better feel for the form of the player as compared to the competition.

Ferrer's form was clearly not that great, and he got smashed as predicted by the AMs. His opponents also subjectively were not necessarily tougher than novak's opponents, and they also performed worse that he did as per opponent's AM.
 

krosero

Legend
2007 RG
Federer - 13.3%
Davydenko - 10.4%

2007 USO
Federer - 25.2%
Davydenko - 14.8%

2006 USO
Federer - 29.5%
Davydenko - 13.9%

2006 AO
Federer - 15.6%
Davydenko - 13.8%

2010 AO
Federer - 14.0%
Davydenko - 4.8%

2013 AO
Federer - 15.2%
Davydenko - 2.5%
 

krosero

Legend
2006 Wimb
Federer - 44.7%
Henman - 17.4%

2006 USO
Federer - 28.9%
Henman - 14.4%

2004 USO
Federer - 32.3%
Henman - 20.9%
 

krosero

Legend
From tennis.matchstat.com, I got the stat for one match I was searching for for quite some time :

federer ancic wimbledon 2006 QF:

federer : 35 winners, 17 UEs, 92 points won
ancic : 24 winners, 13 UEs, 75 points won

federer's AM ~ 37.1%
ancic's AM ~ 27%

This was a brilliant, highly under-rated match ... the scoreline of 6-4,6-4,6-4 does no justice to how well Ancic played . Federer was just too good and had all the answers ...

Another match from the same tourney ,

federer-berdych wimbledon 2006 4R :

federer : 47 W, 15 UEs, 93 points won
berdych : 19 W, 20 UEs, 64 points won

federer's AM ~ 37%
berdych's AM ~ 18.5%
A few more from 2006 Wimbledon.

R128
Federer - 44.6%
Gasquet - 20.9%

R32
Federer - 39.2%
Mahut - 27.0%

SF
Federer - 39.8%
Bjorkman - 9.8%
 

krosero

Legend
2003 USO
Nalbandian - 16.5%
Federer - 9.8%

2005 USO
Federer - 29.5%
Nalbandian - 7.4%

2004 AO
Federer - 16.4%
Nalbandian - 13.9%

2011 Wimb
Federer - 40.8%
Nalbandian - 28.2%
 

krosero

Legend
2011 RG
Nadal - 26.1%
Soderling - 15.2%

2010 RG
Nadal - 21.6%
Soderling - 11.1%

2009 RG
Soderling - 20.7%
Nadal - 15.1%
 
Top