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

[krosero]

Originally Posted by abmk

sampras vs rafter/phillippoussis in Davis Cup

We have a new high for Sampras.

1997 GS Cup final
Sampras - 47.8%
Rafter - 28.6%

1997 Davis Cup
Sampras - 40.0%
Rafter - 21.8%

The Davis Cup match was the highest I had for him till now.

But I don't have AM's for any of Sampras' Wimbledon finals. Not one.

I can't get reliable UE counts for any of them. In the 1993 final I'm missing UE counts for both Sampras and Courier. In all of the rest of Pete's finals I have his UE count but no UE count for his opponents, without which I can't calculate AM's.

[pc1]

Originally Posted by krosero

It's here: http://tt.tennis-warehouse.com/showthread.php?t=171490

Thanks. 11 characters.

[krosero]

Originally Posted by abmk

well, IMO, the roof being closed did play a part in those stats.

at wimbledon :

I think federer served exceptionally brilliantly in the semis, but he did return better in the finals ..

murray in the finals also returned quite a bit better than djoker did in the semis ...

the matches were closer in quality than the respective AMs seem to suggest ...

differences being the roof and the better quality of returning resulting in more rallies and as a result more UEs in the final

I will admit, I am a bit surprised that they have federer at only 10 UEs in the semi. I thought he made more than that ...

I agree with a lot of this, though I would question whether the lower AM's in the final were due to better returning.

This gets to the issue we've been debating in recent days in this thread, about whether it's possible for two players to lift their play, from one match to the next, but for their AM's to stay the same or even to go lower.

I doubt that better returning can result in lower AM's. Let's say I make 20 unforced errors on the return, in one match. I cut that down to 10 in the rematch. The quality of play has gone up, and my AM will be higher, because I've cut down on my UEs.

You said that the better returning produced a greater number of rallies, and that long rallies tend to end in UEs. I agree that long rallies often end in UEs. But not all of them. Some of those points -- in fact, a decent number, even among insanely good defenders like Nadal and Djokovic -- will end in winners, or in forcing shots.

Let's do the math. On the return, I reduced my UE's by 10. Yesterday on those 10 points I didn't put the ball back in play; today I made rallies out of those 10 points. Let's say I make UE's on 6 of those rallies. I still have an overall improvement of 4 UE's.

If I reduced my return UE's by 10, and ALL TEN rallies ended in UE's, then my AM would stay the same. But the chances of that happening are small. If even one of those points ends with me making a winner or a forcing shot, then I've come out ahead overall, and my AM will rise from the previous day.

As for the real matches at Wimbledon, I'm not sure that the returning was better in the final than in the semi.

Unfortunately I only saw the middle of Fed-Djokovic. (I saw all of the final). I did notice Djokovic having trouble with Federer's serve, and I wondered whether Djokovic was just having an off day, or whether his trouble was due to Federer's serving. I really thought Djokovic looked handcuffed by Federer's serve, for the most part.

Wimbledon.org had Djok making only 8 UE on the return, which is pretty good returning in a four-set match. They had Federer making only 5 UE on the return -- so Federer really could not have returned much better, if at all, in the final.

Fed/Djok had the roof closed throughout. Federer's serve traveled faster by a few mph whenever they closed the roof. In the final, when they closed the roof one of the announcers said that his average 1st serve went up by 5 mph.

But the Fed/Djok semi was played entirely under the roof, and if you look at the Wimbledon boxscores Federer did have more mph on his serves in the semifinal, as compared to the final. His average 1st serve was 2 mph higher in the semis. His average 2nd serve was 4 mph higher, which is a significant margin -- and that partly explains how Federer managed to win 72% of his 2nd serve points in that match. That's just phenomenal; I fully agree with you that he served extremely well against Djok.

But I think that means that Djokovic looked like he was returning worse than Murray, because Djokovic had to deal with better serves.

The same seems to be true on the other side of the coin. Did Federer improve his returns, going from the semis to the finals? Or did his returning just look worse against Djokovic, because Novak was serving better than Andy?

I think Novak was serving better. Perhaps the indoor conditions helped the mph on his serve. Murray barely had a faster 1st serve, on average, than Djokovic (121 mph vs 120 mph). But on second serve Djokovic was clearly superior: 95 mph vs 88 mph.

So Federer's returning may have looked better in the final because he had more of a puffball to deal with, on second serves.

7 mph is a significant difference. And yet Federer, dealing with these faster second serves from Djokovic, made only 5 ue's errors on the return.

Federer cannot have cut that number down by much, if at all, in the final.

So I think if the AM's are lower in the final than in the semifinal, we may need to look for another cause.

The mph figures, and the return UE's counts, actually suggest that the semifinal had better serving and returning than the final did.

And if so, that's reflected in the AM's.

[Bursztyn]

I wonder what was the AM for Sampras in this match;

http://www.youtube.com/watch?v=LmbervegakA

He won pretty convincingly.

[abmk]

I watched the sampras stich wimbledon 1992 QF match again :

I had earlier done the stats for it , but those seem to be off a tiny bit ..

http://tt.tennis-warehouse.com/showthread.php?t=327265

sampras won 84 points , stich 61

sampras had 2 FH UEs, 3 BH UEs and 2 DFs - 7 UEs overall
stich had 2 FH UEs, 0 BH UEs and 3 DFs - 5 UEs overall

so a total of 12 UEs in the match ..

sampras' AM = (84-12)/(84+61) = 49.66%
stich's AM = (61-12)/(84+61) = 33.8%

again, these are my own stats and not official ones ...

[krosero]

Originally Posted by Flash O'Groove

After thinking about it for a few days, I think their is some issue with this reasoning. The AM is the measure of the ability of a player to execute successfully an aggressive pattern of winning points. From our spectator's point of view, we will often equate an high AM with high quality of play, because winners are more spectacular . Thus, we will consider a match where one player as high AM as a pleasant match, and more so if both players have an high AM. Thus, quality means more, is this case, spectacular than "high competitive level of play".

Actually the AM's cannot tell you which matches were most entertaining -- if we think of entertaining matches as having a lot of winners.

There are two ways to get a high AM. You could increase your winners; that would be a more entertaining match. Or you could get a high AM by decreasing your unforced errors. In the latter case, maybe the number of winners you made was not unusually high and what was really great about your level of play was how few errors you made. That sort of match would not qualify as entertaining -- not if we're looking for winner-heavy matches.

Maybe you read my post above to Urban, about the 1987-88 USO finals between Wilander and Lendl. Most people, myself included, would consider the '88 final to be more exciting. There is more net play and Lendl is hitting a lot of passes and other aggressive winners. The rate of winners -- meaning how frequently they occurred in the match -- was faster than in the '87 match. But the rate of errors was also faster than it was in '87.

The '87 final featured endless rallies, and some people said it was like watching paint dry. I actually think the match was better quality than the '88 final, because Lendl's UE's were so low. (He went from 58ue in '87 to 85ueE in '88.) But that made the match less exciting.

And you can guess what I'm going to say next: the less exciting match had the higher AM's.

Originally Posted by Flash O'Groove

For exemple, I disagree with your statement that the level of play of Murray in the Olympic final is as high as Fed's level during the Wimbledon final. We know that he was able to execute an aggressive pattern of winnings points as successfully as Roger did, and we know than from a spectator standpoints, his performance was as enjoyable. But that doesn't necessarily means that he played better, as the AM is related to the level of his opponent. In this case, the AM of Federer for the Olympic final suggest that he played worse: he couldn't execute successfully an aggressive pattern of play, and he probably didn't defend as well, thus enabling Murray to considerably rise his own AM, even if himself played worse too! The inverse is possible too: Murray was so good in the Olympic final that Federer's AM dropped. I know for seeing these matches that Federer's level did drop off (he was clearly mentally tired after is SF against Del Potro), but the statistics themselves don't really tell the competitive level of play of both players. They say that the Wimbledon final was a better match from a spectator points of view.

The part I bolded is where I have disagreements.

Federer did not play well in the Olympic final, agreed. He made a lot more UE's than in the Wimbledon final, yes, so far so good.

But if your opponent's UE's rise, your own AM will fall. You can work it out in Excel or with a calculator: punch in higher numbers for Federer's UE. Murray's AM will fall.

Or just think about it this way: when your opponent's UEs increase, you're getting more of the points for free. More of the points in the match are being handed to you on a platter (I'm exaggerating to make the point) -- that is, you did not have to make any aggressive plays on those points. The number of points that you have to win through your own aggressive plays to win the match has suddenly decreased.

Let's take the Olympic final and increase Federer's UEs.

Federer won 68 points and made 31 UE's - AM of 12.1%
Murray won 97 points and made 17 UE's - AM of 29.7%

Those were the real numbers. Murray won 97 points in the match, making aggressive plays for 66 of the points. And he won 31 additional points that Federer handed to him.

Now let's change Federer's UE to 41:

Federer won 68 points and made 41 UE's - AM of 6.1%.
Murray won 97 points and made 17 UE's - AM of 23.6%.

Now Murray only has to make aggressive plays on 56 points; 41 points are now being handed to him.

Let's say Murray tanks, and hands Federer all of the points in the match. Federer wins 6-0, 6-0, 6-0. Eighteen straight games, 72 straight points -- all of them won by Federer on Murray's unforced errors. Federer didn't have to lift a finger to win anything.

Murray's AM in that scenario would be negative: -100%. Federer's AM would be exactly 0.0%.

I think you've put your finger on some good issues, but I don't think it's valid to say that Murray's AM was high in the Olympic final because Federer's was low. If your opponent is handing you the match, you're getting free points rather than making aggressive plays, so your own AM has to go down too.

[krosero]

Originally Posted by abmk

I watched the sampras stich wimbledon 1992 QF match again :

I had earlier done the stats for it , but those seem to be off a tiny bit ..

http://tt.tennis-warehouse.com/showthread.php?t=327265

sampras won 84 points , stich 61

sampras had 2 FH UEs, 3 BH UEs and 2 DFs - 7 UEs overall
stich had 2 FH UEs, 0 BH UEs and 3 DFs - 5 UEs overall

so a total of 12 UEs in the match ..

sampras' AM = (84-12)/(84+61) = 49.66%
stich's AM = (61-12)/(84+61) = 33.8%

again, these are my own stats and not official ones ...

Great, thanks for these! I need to see this match sometime, too.

Stich did not play that badly (he actually had fewer UE's than Sampras); but he got straight-setted because Sampras played so well.

Yet it's still surprising, because Sampras said that Stich was one of his most difficult opponents.

I think when you give your UE count, your terms "FH" and "BH" include both ground strokes and net strokes?

I presume that both players came in behind all of their serves, so there cannot have been many opportunities for either player to make unforced errors with groundstrokes.

[abmk]

Originally Posted by krosero

Great, thanks for these! I need to see this match sometime, too.

Stich did not play that badly (he actually had fewer UE's than Sampras); but he got straight-setted because Sampras played so well.

Yet it's still surprising, because Sampras said that Stich was one of his most difficult opponents.

I think when you give your UE count, your terms "FH" and "BH" include both ground strokes and net strokes?

I presume that both players came in behind all of their serves, so there cannot have been many opportunities for either player to make unforced errors with groundstrokes.

yes, basically FH and BH UE's for any kind of unforced errors of those wings ..yep, not that many chances to make errors with groundstrokes ...

the match is available on youtube

sampras stich wimbledon 1992 QF

stich did play some good tennis, but wasn't at his sharpest and sampras' playing his finest tennis pounced on it ....

[abmk]

Originally Posted by krosero

I agree with a lot of this, though I would question whether the lower AM's in the final were due to better returning.

This gets to the issue we've been debating in recent days in this thread, about whether it's possible for two players to lift their play, from one match to the next, but for their AM's to stay the same or even to go lower.

I doubt that better returning can result in lower AM's. Let's say I make 20 unforced errors on the return, in one match. I cut that down to 10 in the rematch. The quality of play has gone up, and my AM will be higher, because I've cut down on my UEs.

You said that the better returning produced a greater number of rallies, and that long rallies tend to end in UEs. I agree that long rallies often end in UEs. But not all of them. Some of those points -- in fact, a decent number, even among insanely good defenders like Nadal and Djokovic -- will end in winners, or in forcing shots.

Let's do the math. On the return, I reduced my UE's by 10. Yesterday on those 10 points I didn't put the ball back in play; today I made rallies out of those 10 points. Let's say I make UE's on 6 of those rallies. I still have an overall improvement of 4 UE's.

If I reduced my return UE's by 10, and ALL TEN rallies ended in UE's, then my AM would stay the same. But the chances of that happening are small. If even one of those points ends with me making a winner or a forcing shot, then I've come out ahead overall, and my AM will rise from the previous day.

As for the real matches at Wimbledon, I'm not sure that the returning was better in the final than in the semi.

Unfortunately I only saw the middle of Fed-Djokovic. (I saw all of the final). I did notice Djokovic having trouble with Federer's serve, and I wondered whether Djokovic was just having an off day, or whether his trouble was due to Federer's serving. I really thought Djokovic looked handcuffed by Federer's serve, for the most part.

Wimbledon.org had Djok making only 8 UE on the return, which is pretty good returning in a four-set match. They had Federer making only 5 UE on the return -- so Federer really could not have returned much better, if at all, in the final.

Fed/Djok had the roof closed throughout. Federer's serve traveled faster by a few mph whenever they closed the roof. In the final, when they closed the roof one of the announcers said that his average 1st serve went up by 5 mph.

But the Fed/Djok semi was played entirely under the roof, and if you look at the Wimbledon boxscores Federer did have more mph on his serves in the semifinal, as compared to the final. His average 1st serve was 2 mph higher in the semis. His average 2nd serve was 4 mph higher, which is a significant margin -- and that partly explains how Federer managed to win 72% of his 2nd serve points in that match. That's just phenomenal; I fully agree with you that he served extremely well against Djok.

But I think that means that Djokovic looked like he was returning worse than Murray, because Djokovic had to deal with better serves.

The same seems to be true on the other side of the coin. Did Federer improve his returns, going from the semis to the finals? Or did his returning just look worse against Djokovic, because Novak was serving better than Andy?

I think Novak was serving better. Perhaps the indoor conditions helped the mph on his serve. Murray barely had a faster 1st serve, on average, than Djokovic (121 mph vs 120 mph). But on second serve Djokovic was clearly superior: 95 mph vs 88 mph.

So Federer's returning may have looked better in the final because he had more of a puffball to deal with, on second serves.

7 mph is a significant difference. And yet Federer, dealing with these faster second serves from Djokovic, made only 5 ue's errors on the return.

Federer cannot have cut that number down by much, if at all, in the final.

So I think if the AM's are lower in the final than in the semifinal, we may need to look for another cause.

The mph figures, and the return UE's counts, actually suggest that the semifinal had better serving and returning than the final did.

And if so, that's reflected in the AM's.

firstly, I think federer *did* make more unforced errors than what is shown in the semi stats ... maybe I'd have to watch the match again ..

again, just the UEs made on the return aren't the only measure of the quality of the return, right ? its ok, if you make a few unforced errors here and there if you are trying to attack , but if you can put pressure on the server by attacking, that'd have achieved the purpose.

Also, if a player is returning well, points that could've been forced errors on the return get back into play and those rallies could end in unforced errors

in the final, avg serve speeds :

federer :

1st set : 1st serve : 115 mph, 2nd serve : 96 mph
2nd set: 1st serve : 116 mph, 2nd serve : 101 mph
3rd set : 1st serve : 117 mph, 2nd serve : 100 mph
4th set : 1st serve : 116 mph, 2nd serve: 99 mph

only in the first set, speed was less, so I don't think it had to do with the roof .... probably had to do with some back pain/ache for federer or his back being stiff - he removed the extra shirt mid-way through the second set - probably because he felt better ..

federer had back problems in the benneateau match and only upped the ante on his serve when he had to, after being down two sets to love .... then in the malisse match, it was obvious that even his movement was hindered because of the back problem ...given that I was surprised at how well he served in the djokovic match

federer did obviously find it easier vs murray's 2nd serve, but first serve ? I think murray's was better ... not that much of a difference in federer's returning in the semis and finals, but I think he was a bit better in the finals than in the semis

anyways djoker's returning did look worse because federer served quite a bit better in the semis, but I do think the other part of it was that murray did in fact return quite a bit better ( even if it was against a worse serve )

I remember murray going deep into almost every service game of federer's, but in the djoker match, I remember several games that just flew by, at love or 15 ...

one good stat to look at might be the % of returns in ...

[krosero]

For now just some more stats. I've posted AM's for:

Berdych d. Federer, 2010 Wimb
Federer d. Hewitt, 2005 Wimb
Federer d. Gaudio, 2005 TMC (double-bagel)
Federer d. Blake, 2006 TMC final
Nadal d. Youzhny, 2010 USO
Johansson d. Roddick, 2004 USO

That last one is so strange. Johansson won 24 fewer points than Roddick did (128-152), but won the match. So the AM's look like this:

Johansson - 20.4%
Roddick - 28.9%

[MGK1965]

He could retire. I wish he'd try switching shoes. My knees felt much better after I switched from Nike to Prince. I tho.nk Nike are terrible for support.

[krosero]

Originally Posted by Bursztyn

Perhaps AM should be viewed as a way to describe a tennis match rather than as a measure of its quality?

The ultmate goal in tennis is to win, isn't it?

So Borg beat Vilas in RG 1978 final because of his consistency and his stamina (perseverence, patience, discipline?) - both are very important for a tennis player.

He used his weapons to beat his opponent, however he employed a bit unusual tactics.

Should his performance be regarded as inferior (when compared to some of his other matches) because he decided to use less spectacular method to win?

No it shouldn't necessarily be regarded as inferior just because Borg employed a less spectacular method. You can have a high AM by lowering your UE's and playing un-spectacular tennis.

Lendl, for example, had a lower UE count in his '87 USO final than he did in the '88 final (both against Wilander). He also struck fewer winners in '87, because he was employing, as you put it, a less spectacular strategy.

But Lendl's AM is higher in '87. So the AM is not punishing the less spectacular performance. In this case the less spectacular is actually being rewarded.

The Borg/Vilas/Pecci example is trickier, because Borg’s less spectacular performance against Vilas does have a lower AM than his performance against Pecci.

That's because Borg actually made more UE's against Vilas. In that very short match (short in terms of number of points played), Borg made 19 UE, by my count. In the much longer Pecci match, Borg made only 17 UE (Bud Collins' count).

It's difficult to use AM's when the players you're comparing have radically different styles. Pecci was always coming into net, always forcing the issue. Borg had fewer opportunities to make an UE, when facing Pecci. On many points Borg was either forced into an error, or he made a winning pass. When Borg played Vilas, almost every point was an opportunity to make an UE because both players stayed back and rallied endlessly.

But it was a smart way to play Vilas -- at least for Borg.

So was Borg's '78 performance inferior to '79? I'm not certain. Borg was in peak form both years, and '78 was actually the year that he set the record for fewest games lost in the tournament.

On the one hand, I know that Borg had more UE's errors in '78 not because he played below form but because long rallies against another baseliner present constant opportunities to make UE's.

On the other hand I don’t think that Vilas forced Borg to dig deep or to do anything special. Against Pecci, Borg had to run more, make precision passes, keep Pecci from attacking his serve, etc.

I might have to say that Borg’s performance against Pecci was the higher performance. It’s not specifically his higher AM against Pecci that makes me say that, because AM's are a bit deceptive in comparisons that involve radically different styles. I say that just because I think that Pecci played a better match than Vilas. Pecci was in better form in ’79 than Vilas was in ’78. And in fact Pecci beat Vilas on the way to the '79 final, in straight sets. Pecci was a very good claycourter in peak form, and he could get "hot." Vilas in ’78 was struggling a bit with his game – and had a bit of a mental block against Borg.

[krosero]

1991 USO
Connors - 8.3%
Krickstein - 11.5%

Now here was one of the most exciting matches in recent years, yet the AM's are relatively low.

It was exciting because there were so many winners; but the AM's are low because the UE's were high.

Officially, Connors had 86 winners and 106 UE. Krickstein had 38w, 44ue.

So both men had negative winner/error differentials: Connors at -20, Krickstein at -6.

Connors won the match but has poorer stats than Krickstein because he trailed Krickstein by a large margin in total points won: 181-193.

[piece]

Originally Posted by krosero

Great, thanks for these! I need to see this match sometime, too.

Stich did not play that badly (he actually had fewer UE's than Sampras); but he got straight-setted because Sampras played so well.

Yet it's still surprising, because Sampras said that Stich was one of his most difficult opponents.

I think when you give your UE count, your terms "FH" and "BH" include both ground strokes and net strokes?

I presume that both players came in behind all of their serves, so there cannot have been many opportunities for either player to make unforced errors with groundstrokes.

I'm curious as to whether you think the issue implicit in the bolded section of your post undermines the usefulness of AM as a means of measuring quality of play. It seems to me that in matches where both players regularly attack the net, their respective AMs will tend to be higher than those of players who stick more to the baseline. The reason for this, as you alluded to, is that when a player approaches the net he takes away any opportunity for his opponent to hit an unforced error, because attempted passing shot errors are counted as forced by convention. This effectively halves the probability that an unforced error is hit during a point where one player is at net when compared to a point where both players remain at the baseline, as in the former only one player can potentially make an unforced error, while in the latter either player can. This fact is important because any reduction in unforced errors necessitates a corresponding increase in the fraction of points won by aggressive plays, and hence an increase in AM.

Was the Sampras-Stich match really as high quality as the AMs make it out to be, for instance? I have my doubts; it seems that these figures might, to some extent, be an artifact of the playing styles of Sampras and Stich.

I feel this is an issue, not least of all because I see no reason to think that matches involving lots of serve and volleying or lots of chip and charging are of higher quality in general than matches played primarily from the baseline.

Do you have any thoughts on this?

[krosero]

Originally Posted by piece

I'm curious as to whether you think the issue implicit in the bolded section of your post undermines the usefulness of AM as a means of measuring quality of play. It seems to me that in matches where both players regularly attack the net, their respective AMs will tend to be higher than those of players who stick more to the baseline. The reason for this, as you alluded to, is that when a player approaches the net he takes away any opportunity for his opponent to hit an unforced error, because attempted passing shot errors are counted as forced by convention. This effectively halves the probability that an unforced error is hit during a point where one player is at net when compared to a point where both players remain at the baseline, as in the former only one player can potentially make an unforced error, while in the latter either player can. This fact is important because any reduction in unforced errors necessitates a corresponding increase in the fraction of points won by aggressive plays, and hence an increase in AM.

Was the Sampras-Stich match really as high quality as the AMs make it out to be, for instance? I have my doubts; it seems that these figures might, to some extent, be an artifact of the playing styles of Sampras and Stich.

I feel this is an issue, not least of all because I see no reason to think that matches involving lots of serve and volleying or lots of chip and charging are of higher quality in general than matches played primarily from the baseline.

Do you have any thoughts on this?

Yes I think this is an issue with any stats that use Unforced Errors: namely methods like the Aggressive Margin, winner/error differentials, ratios, etc.

On grass you will commonly see very high winner/error differentials: +50 or more. On clay it's hard to get even a positive differential, and negative ones are common.

Perhaps I should have listed the AM's grouped by surface, because that is the factor that seems to have the greatest impact on how people play -- and therefore an impact on how UE's are scored in those matches.

On grass two SVers coming in behind every serve will have very few opportunities to make UE's. On clay two baseliners engaged in a war of attrition are going to have many such opportunities.

That's one reason that AM's are higher on grass. But leaving aside the question of playing styles, it's also just generally true that you can make more winners and forcing plays on a fast surface. That has nothing to do with styles per se, because you could have two baseliners playing each other, and they will find it to be true: they will have an easier time forcing their opponent when they're playing on a fast surface.

On clay you can hit the ball just as hard, or just as wide, but the defender gains an advantage, as compared to grasscourt matches. Any surface that slows down the ball will make it easier for the defender to catch up to it, and harder for anyone to hit winners or to make forcing plays.

And that's what the AM is about: making forcing plays while keeping your unforced errors down. That is just inherently easier on a fast court.

So that makes it really impossible to compare AM's -- or winner/error differentials and ratios -- across surfaces (except for limited purposes like I did with the Federer-Delpo matches).

Take the top AM's in my opening post as examples. McEnroe has 53% on grass, the highest AM of all. The highest AM on clay belongs to Nadal, with 35%. No one would suggest that McEnroe's level on grass is a greater level than what Nadal achieved on clay. I think it's fair to assume that both of those performances were GOAT level for their respective surfaces.

Sampras/Stich should be compared to other grasscourt matches of the time period. It was very common then for players to follow all, or nearly all, of their serves to net. Sampras/Stich could easily be compared with Sampras/Becker, for example.

But if you try comparing it to this era's matches at Wimbledon, which commonly feature two baseliners slugging it out, you're going to have problems.

At the very least you could not use the numbers in such a comparison literally, if at all.

But that all results from surface being such a large factor in tennis. I'm not sure there's any good or efficient way around this. I, for one, would not want to chuck the distinction between a forced and an unforced error; I would just restrict comparisons by surface.

But maybe we need to think more about how to score UE's.

[abmk]

krosero,

I rewatched this match ...I take back my statement that federer returned better in the final .. I think he was just as good in the semis as he was in the finals as far as returning goes ... However still think murray did return quite a bit better than djoker ...( even though it was against a weaker federer serve )

at the tail end of the match, a stat was shown, 57% returns in for djoker and 68% in for federer

I did the UE stats for this as well,

wimbledon.org has UEs by set :

federer : 2,3,4,1 - total of 10
djokovic : 5,3,9,5 - total of 22


I have significantly different :

federer : 4,4,9,1 - total of 18
djokovic : 6,3,11,7 - total of 27

which brings their AMs to :

federer : 31.95%
djokovic : 26.4%


funnily enough, I have 4 UEs on the return for fed,7 for djoker
wimbledon.org has 5 for fed and 8 for djoker

so I don't know how lenient the statistician was on strokes apart from the return

My initial impression was justified, there were more unforced errors than what is on the site and difference is non-trivial ..

If by any chance, you are planning to watch this, can you please keep track of the UEs and see how many you end up with ?

[corners]

Originally Posted by abmk

krosero,

I rewatched this match ...I take back my statement that federer returned better in the final .. I think he was just as good in the semis as he was in the finals as far as returning goes ... However still think murray did return quite a bit better than djoker ...( even though it was against a weaker federer serve )

at the tail end of the match, a stat was shown, 57% returns in for djoker and 68% in for federer

I did the UE stats for this as well,

wimbledon.org has UEs by set :

federer : 2,3,4,1 - total of 10
djokovic : 5,3,9,5 - total of 22


I have significantly different :

federer : 4,4,9,1 - total of 18
djokovic : 6,3,11,7 - total of 27

which brings their AMs to :

federer : 31.95%
djokovic : 26.4%


funnily enough, I have 4 UEs on the return for fed,7 for djoker
wimbledon.org has 5 for fed and 8 for djoker

so I don't know how lenient the statistician was on strokes apart from the return

My initial impression was justified, there were more unforced errors than what is on the site and difference is non-trivial ..

If by any chance, you are planning to watch this, can you please keep track of the UEs and see how many you end up with ?

I re-watched this match a couple days ago and had the same reaction I had the first time I watched and saw the very low UE count flash on screen: What!?

BTW, I didn't score it because I don't know how. Can anyone suggest an authoritative guide to differentiating UE, winners and forced errors?

[krosero]

Originally Posted by abmk

one good stat to look at might be the % of returns in ...

Semifinal
Djokovic got back 57 of Federer’s 101 good serves: 56.4%
Federer got back 76 of Djokovic’s 113 good serves: 67.3%

Those stats are from Wimbledon.com and the New York Times tennis blog; I also used ABMK's late-match graphic.

Final
Murray got back 99 of Federer’s 128 good serves: 77.3%
Federer got back 107 of Murray’s 156 good serves: 68.6%

I got stats for the final by using a graphic that was shown in the fourth set. At 2-all, 15-love, right after Murray forces a BH return error, the graphic had Murray with 43 unreturned serves and Federer with 25. I counted to the end – including aces in my count, which I think is the way this is done though I'm not absolutely certain – and finished with Murray at 49 and Federer 29.

The graphic makes it look like Murray was doing nearly twice as well as Federer in serving unreturned serves. That's because it's a simple count of how many serves were unreturned; it does not take into account the fact that Murray served a lot more points in the match than Federer did. When you convert the counts to the percentages above, there was not much of a difference between the two players.

ABMK, even though Federer had more mph on his serves in the semifinal, compared to what he served at Murray, I think these numbers support your argument that Murray returned better than Djokovic (btw I'll have more extensive replies to your posts later).

Federer's second serve was really humming in the semifinal, so if anyone counts the return errors in this match it would be useful to know how many of his second serves went unreturned. If a LOT of Djokovic's return errors occurred on second serve then we might have to give more credit to Federer for making Djokovic's returning look worse than Murray's.

[Bursztyn]

We have been debating whether it is possible for the two players to raise their level of play without changing the AM's.

I would like to present a hypothetical scenario in which two players playing on two occasions on the same surface exhibit different level of play yet their calculated AM’s stay the same.

It is open to debate if such a scenario is realistic.

On the first day they played an average match; winners, forced and unforced errors were counted, and the AM's were calculated.

On the second day they played an excellent match, however the number of winners, forced and unforced errors was the same as on the first day (for the sake of simplicity I assume that the number of points in both matches was the same). How is it possible?

Winners;
The first player was hitting great shots, however the second player was also defending better than previously (had the second player been playing as well as on the first day, the first player would have had more winners than on the first day, had the first player been playing as well as in the first match he would have had less winners than in the first match). As a result the first player had the same number of winners on both occasions, and the same applies to the second player.

Forced errors;
Generally the same logic applies to forced errors. Both players committed the same number of forced errors as during the first match, however the shots that forced these errors were of the higher quality than the shots that forced errors in the first match.

Unforced errors;
Generally the same logic applies to unforced errors. Both players committed the same number of unforced errors as during the first match, however the points which ended in unforced errors were of the higher quality than points which ended in unforced errors in the first match. We have already mentioned that points that end in an unforced error may be of different quality (e.g. long neutral rallies versus double faults or return errors or short neutral rallies).

[piece]

Originally Posted by krosero

Yes I think this is an issue with any stats that use Unforced Errors: namely methods like the Aggressive Margin, winner/error differentials, ratios, etc.

On grass you will commonly see very high winner/error differentials: +50 or more. On clay it's hard to get even a positive differential, and negative ones are common.

Perhaps I should have listed the AM's grouped by surface, because that is the factor that seems to have the greatest impact on how people play -- and therefore an impact on how UE's are scored in those matches.

On grass two SVers coming in behind every serve will have very few opportunities to make UE's. On clay two baseliners engaged in a war of attrition are going to have many such opportunities.

That's one reason that AM's are higher on grass. But leaving aside the question of playing styles, it's also just generally true that you can make more winners and forcing plays on a fast surface. That has nothing to do with styles per se, because you could have two baseliners playing each other, and they will find it to be true: they will have an easier time forcing their opponent when they're playing on a fast surface.

On clay you can hit the ball just as hard, or just as wide, but the defender gains an advantage, as compared to grasscourt matches. Any surface that slows down the ball will make it easier for the defender to catch up to it, and harder for anyone to hit winners or to make forcing plays.

And that's what the AM is about: making forcing plays while keeping your unforced errors down. That is just inherently easier on a fast court.

So that makes it really impossible to compare AM's -- or winner/error differentials and ratios -- across surfaces (except for limited purposes like I did with the Federer-Delpo matches).

Take the top AM's in my opening post as examples. McEnroe has 53% on grass, the highest AM of all. The highest AM on clay belongs to Nadal, with 35%. No one would suggest that McEnroe's level on grass is a greater level than what Nadal achieved on clay. I think it's fair to assume that both of those performances were GOAT level for their respective surfaces.

Sampras/Stich should be compared to other grasscourt matches of the time period. It was very common then for players to follow all, or nearly all, of their serves to net. Sampras/Stich could easily be compared with Sampras/Becker, for example.

But if you try comparing it to this era's matches at Wimbledon, which commonly feature two baseliners slugging it out, you're going to have problems.

At the very least you could not use the numbers in such a comparison literally, if at all.

But that all results from surface being such a large factor in tennis. I'm not sure there's any good or efficient way around this. I, for one, would not want to chuck the distinction between a forced and an unforced error; I would just restrict comparisons by surface.

But maybe we need to think more about how to score UE's.

Your point about surface affecting AM is well taken. You say, for instance, that fast/low-bouncing surfaces tend to have higher AMs. The reason you give for this seems to be twofold: (1) it's easier to execute aggressive plays on fast/low-bouncing surfaces; and (2) netplay is more common on fast/low-bouncing surfaces. Both of these two aspects (easier to force plays, more netplay) seem to positively affect AM, yet you talk as if matches played on the same surface (with the same speed) are roughly comparable with respect to AM. This would only be true if the amount of netplay is roughly constant between matches on a given surface.

It's not clear to me that their is this constancy - especially if we look across eras. Compare Borg-Vilas or Djokovic-Nadal at the FO to Laver-Rosewall. The former pairs remained almost exclusively at the baseline, while the latter approached the net in most points. I include Djokovic-Nadal because, while both camped out on the baseline, they also both go for winners when they have an opening, perhaps in contrast to the more conservative play of Borg-Vilas. This is worth mentioning because I'm going to guess Laver-Rosewall would have the highest AMs of the three pairs due to less unforced errors, but I don't want this attributed to the 'rally-all-day' style of Borg-Vilas. Rather, I think, Laver-Roswall will have the higher AMs than the other pairs because they had less opportunity to hit unforced errors due to the frequent net approaches. This would be the case even though the surface is held relatively constant.

You've talked about this yourself, and the lesson I think is that one can only use AM to compare match quality when holding the surface and level of net play relatively constant. Otherwise AMs can be easily inflated or deflated by factors unrelated to quality of play.