Success on second serve

[krosero]

Originally Posted by Limpinhitter

I'd like to know what Gonzeles success on 1st and 2nd serves was. Virtually everyone who played Gonzales, including Kramer and Trabert, has said that his serve was the best they'd ever seen.

I'd like to see such a stat for Von Cramm, given what was said about his second serve.

Here's a page from A Terrible Splendor where Fisher describes von Cramm's second serve as possibly better than his first: http://books.google.com/books?id=Or9...0serve&f=false

And yet here's a description of Budge taking von Cramm's second serve on the rise during the fifth set of their Davis Cup meeting:

"Budge, after holding serve to make the score 4-2, decided he must gamble to pull himself back from the abyss. The baron's serve, particularly his second delivery, tended to kick high off the grass and at a tricky angle. To nullify that high hopper, Budge moved a step closer to the net, hoping to catch the ball on the rise with his superb backhand, which may have been the best the game has ever known. Luck was also with Budge, for Von Cramm, in his eagerness to close out the match, began missing his first service. Only once in the critical seventh game did the baron get his first serve in, and that was the only point he won. Budge took each second serve on the rise and drove Von Cramm deep, setting up a volley."

http://sportsillustrated.cnn.com/vau...11/4/index.htm

[krosero]

Here's a reference to the strength of Tilden's second serve:

"The champion’s second ball was a big thorn in Johnston’s side, breaking too high and wide for him to do more than to get the tip of his racquet on it and send back a short return, which Tilden put away for a point."

(That's the NY Times report on the 1924 final at Forest Hills which Tilden won over Johnston 6-1, 9-7, 6-2).

However because of the strength of Tilden's first serve I'd be surprised if he ever had greater success on second serve.

It's tough to say because I haven't found any boxscores from back then which calculated that stat -- or even first serves successfully made.

For example the Times boxscore for this match has the game scores. From that you can calculate the points won on service:

Tilden won 60 of 86 points on serve (70%)
Johnston won 47 of 90 points on serve (52%)

... but nothing about the difference between first and second serves.

[Bursztyn]

Comparing success on first and on second serve we have to bear in mind that samples sizes should be more or less the same (that is the number of first/second serves successfully executed during a match). Usually this is the case however sometimes (as Wilander Leconte, FO 198, players may have unusually high percentage of first serves in (Wilander 97%), which implies that the number of second serves executed was very small (2), and calculated success on second serve (50% in this case) is not very informative. Of course in that match Wilander success on first serve was higher that on the second serve.

[krosero]

Originally Posted by Bursztyn

Comparing success on first and on second serve we have to bear in mind that samples sizes should be more or less the same (that is the number of first/second serves successfully executed during a match). Usually this is the case however sometimes (as Wilander Leconte, FO 198, players may have unusually high percentage of first serves in (Wilander 97%), which implies that the number of second serves executed was very small (2), and calculated success on second serve (50% in this case) is not very informative. Of course in that match Wilander success on first serve was higher that on the second serve.

Right, if you change just one point in Wilander's match he would have had 100% success on second serve. The comparison with first serve in that case would be flawed. However, as you say, his success on first serve was pretty high (70%), higher than most success rates on second serves.

I don't know if I would say that the sample sizes of first and second serves should be more or less the same. Not if what you mean is a first-serve percentage of around 50%. I think a player could make around 70% of his first serves and you'd still be left with a sizable number of second serves.

But I think you're raising an interesting point, and it might be worth looking at whenever someone makes more than 80% of their first serves, and certainly when they make more than 90%. The sample of second serves might be quite small.

[NLBwell]

Graf – 69% to 100% (boxscore) – defeated Seles – 1992 Wimbledon

100% won on 2nd serves - pretty good!

I was surprised how many of the matches were not on clay. That would have been the surface I would have thought most of those occurances would have been on. Back in the 70s a lot of the claycourters just looped both serves in.

[krosero]

Originally Posted by NLBwell

Graf – 69% to 100% (boxscore) – defeated Seles – 1992 Wimbledon

100% won on 2nd serves - pretty good!

That's right, she won 25 of 36 points on first serve (69%) and 11 of 11 on second. She made 77% of her first serves.

[krosero]

From a 1957 profile of Pancho Gonzales ("Anyone For Tennis?", New York Times, Allison Danzig):

_____________________________

Francisco (Pancho) Segura, a little South American from Guayaquil, Ecuador, who has the best two-handed forehand the game probably has known, won the Australian professional championship recently. He rates Gonzales the world's No. 1 player, but only because of his service.

Little Pancho goes so far as to declare that big Pancho would not be an outstanding player were it not for his serve. He says that on any surface Rosewall is a better player than Gonzales, except in the service department. "Rosewall," says Segura, "is sounder off the ground. He returns service more consistently and his passing shots are better. Gonzales has more great shots but he is not as reliable. He uses a little too much wrist. He flicks the ball and has no long follow-through."

But no one, Segura concedes, is more consistent than Gonzales in getting his first service into play. When he does, he adds, "You can't attack him. You can't make him work indoors. Outdoors, you can return his service with more length and make him work more to win the points."

[krosero]

Incidentally the same article reports that Pancho's "service, recorded electronically, travels 112.88 miles per hour."

[krosero]

John Kieran wrote in the New York Times, Sept. 1939: "[Adrian] Quist has a fair service, just as good or perhaps a little better on the second ball than on the first."

[Bursztyn]

Originally Posted by krosero

I don't know if I would say that the sample sizes of first and second serves should be more or less the same. Not if what you mean is a first-serve percentage of around 50%. I think a player could make around 70% of his first serves and you'd still be left with a sizable number of second serves.

I agree. One may say that the smaller sample should not be really small. If it is not very small then we may calculate percentages (success).

[krosero]

Today Djokovic had 60% success on first serve and 61% on second.

http://www.rolandgarros.com/en_FR/sc...18/1602ms.html

So Federer was not doing too well when he got a look at a second serve. Fed also had a poor conversion rate on break points -- 4 of 25. That's 16%, compared to the 40% he won on receiving points overall. (Djokovic converted 4 of 13 break points, or 31%, compared to 33% he won overall.) So I wonder how many of those break points that Federer lost went to second serve.

It would also be an interesting stat from Djokovic's viewpoint, just to see how often he made his first serve when down break point. He made 67% of his first serves overall.

[krosero]

It almost happened again today. Li had 70% success on first serve, 69% on second.

http://www.rolandgarros.com/en_FR/sc...19/2701ms.html

[krosero]

Originally Posted by Bursztyn

I agree. One may say that the smaller sample should not be really small. If it is not very small then we may calculate percentages (success).

I have another match with a really small sample like you're talking about. I'm pretty sure Borg had only 8 second serves in the '79 RG final (d. Pecci). I know that he lost only one of those points, which means that he must have won 69 of 99 points on first serve (you can see his full service numbers here: http://tt.tennis-warehouse.com/showthread.php?t=187098).

That's a success rate of 69.7% on first serve, and 87.5% on second serve.

So what does that mean? If you read the articles of the time, they say it was a smart strategy for Borg to spin his first serve in and deny Pecci the chance to attack second serves. But you look at the success numbers and it almost suggests the opposite: that Borg would have been no worse off, and might even have done better, if he'd missed his first serve frequently.

But that doesn't make sense. Even spinning in those first serves, they were stronger and deeper than the second balls; occasionally he went for the big one on the first serve, too, with success.

I think in a case like this something odd may be going on with the second serve points. I don't know for sure what it might be but here's a suggestion.

If a player like Pecci wants to attack second serves, he might do that with some regular success if he gets a look at second serves with some regularity. Then he can put his plan of attack, whatever that might be, into motion. He can figure out how to do it, and get into some kind of groove.

But if he sees a second serve only once or twice a set, and it comes as a complete surprise each time, he's not suddenly going to be able to switch into the attack mode that he had drawn up on paper. He's been busy playing points on first serves and doing whatever he does to win those points. He can't just switch from that and suddenly do something different. And if he tries he may find himself unsuccessful.

I'm not saying for sure that's what happened in this particular match, but in general it could explain why the returner might win so few points on second serve when you expect him to be able to take advantage of second balls.

[krosero]

Nadal in his loss to Federer, 2010 WTF final in London: 64% on 1st serve, 66% on 2nd.

[Bursztyn]

I do agree that there may be some psychology involved in the "mystery of second serves".

If the second serves were more effective in general why would players bother to play first serves at all?

Players may be more conservative dealing with the first serves and more aggressive when returning the second delivery. Sometimes they may be too aggressive, make a lot of errors thus increasing server's success behind the second serve.

[krosero]

Vilas in the 1978 RG final (lost to Borg): 41% on first serve and 43% on second.

That margin is somewhat small, but 5 of those points on second serve were double-faults. If you exclude those and just look at points where the second serve landed good and started rallies, there was an even stronger tendency for Vilas to do better on his second serve than he did on his first (and note: he had no aces).

[krosero]

Connors – 47% on first serve and 52% on second (1985 Wimbledon, lost to Curren)

[thor's hammer]

Originally Posted by krosero

Each name here is followed by the success on first serve, and then the success on second.

Sampras – 86% to 92% (my count) – defeated Rafter – 1997 Grand Slam Cup

That certainly stands out!

[krosero]

A New York Times article published today noted how Davydenko "is consistently among the best on the ATP Tour in percentage of points won off opponents’ first serves. He has the feel, flexibility, footwork and anticipation that are the distinguishing characteristics of the best returners. This year, the most universal of tools has forsaken him. Since Davydenko switched racket manufacturers, he said, his return has not been the same."

I looked up his stats vs. Federer to see if he ever kept Federer's success on first serve low enough to fall below success on second. It happened once, but actually Davydenko is the one more often who has higher success on second serve.

Federer – 72% on first, 76% on second, d. Davydenko, 2005 Doha
Davydenko – 65% first, 66% second, lost to Federer, 2005 Rotterdam
Davydenko – 56% first, 58% second, lost to Federer, 2005 Hamburg
Davydenko – 51% first, 60% second, lost to Federer, 2006 USO
Davydenko – 53% first, 60% second, lost to Federer, 2010 Cincinnati

[krosero]

In the Davydenko-Nadal rivalry, which Davydenko leads 6-4, both players have always had better success on first serve, with one exception:

Nadal - 59% on 1st serve, 70% on second, d. Davydenko, 2008 Monte Carlo