It's accepted wisdom in tennis circles that surfaces have been made slower and more similar to one another—which is why the top players generally do well in most tournaments, no matter the surface.
It's also accepted wisdom that the serve-and-volley style has died—partly due to this slowing of the surfaces and partly due to racquet and string developments that have made the job of the baseliner easier. (I highly encourage you to read this article on the death of serve-and-volley.)
The most venerable serve-and-volley venue was Wimbledon, but even there, the style died in 2001.
Both of these developments should, in principle, lead to the good return players winning more games.
However, I've come across some highly surprising data that seems to contradict this idea.
While replying to a reader regarding my article about the best returners of the last 20 years, I noticed a very interesting pattern: Players in the 90s won, on average, a higher percentage of the return games than players in the 00s.
In other words, at a time where the serve-and-volley was still a viable and normal tactic, the returners were winning a higher percentage of the return games than they are now.
What's the explanation?
I've come up with a few, but I'll first present my statistical findings.
On average, the 90s had almost 15 players winning 30 percent or more of their return games. There is only one year, 1999, where fewer than 10 players won at least 30 percent of their return games.
In the 00s, only a few elite players have had returning percentages in the 30s. Since 2007, only five or six players each year have finished with such a number.
The number is slightly better if we go back to the early part of the decade: 2003, 2004 and 2006 saw 10, 10 and 11 players, respectively, with a return percentage of 30 or greater.
If we lower the percentage threshold to 25 percent, the pattern is the same. A greater number of players won 25 percent of more of their return games in the 90s than in the 00s.
So, what is the explanation?
If the courts are slower and the strings are better, shouldn't the return player be winning more, not less?
Were the players simply better at returning serves in the 90s? Are the servers significantly better today?
That seems an unlikely explanation to me, given string development and the death of serve-and-volley.
We need to take a closer look at the specific surfaces to understand this strange pattern.
While most surfaces have generally gotten slower, clay has gotten faster.
We no longer see a range of low-ranked clay court specialists beating the top seeds tournament after tournament. At Roland Garros this year, the Big Four all made it to the semifinals, even though only Nadal can be said to prefer clay.
If clay plays more and more like the faster surfaces, the server's advantage increases. So what do we see here?
Here are the yearly number of players who've won at least 30 percent of their return games from 1991-2011:
51 (1991), 37, 43, 43, 41 (1995), 35, 39, 37, 42 (1999), 29, 34, 31, 35 (2003), 29, 30, 33, 19 (2007), 21, 17, 14, 24 (2011).
This show three periods. From 2007 onwards, the average is just below 20 players. From 2000 to 2006, the average is just above 30 players, and in the 90s, the average is just above 40.
Judging by these numbers, clay courts have clearly become faster through the years.
Perhaps fewer players are winning a high percentage of their return games on clay because the surface has become faster, despite the slowing of the other surfaces and better string technology.
But what about hard court and grass—slower surfaces where new string technology should help the baseliner and returner?
Do we see that?
Not on hard court.
Here are the number of players with a 30 percent or better return game, every other year from 1991-2011:
10 (1991), 13, 7, 5, 4 (1999), 4, 6, 6 (2005), 3, 5, 7 (2011).
If we include players who've won at least 25 percent, there's still no evidence of any improvement for the returner:
31 (1991), 36, 22(1995), 19, 21, 26 (2001), 21, 19, 23 (2007), 18, 25 (2011).
I've checked the other years in this period too; if anything, the best years of the hard-court return game were the first four years of the 90s, where at least 30 players were able to win 25 percent of their return games.
From 1995 to 2011, we haven't reached the 30-player mark.
Finally, what about grass?
The problem with grass is that so few matches are played on the surface that the statistics are less reliable when you view each year in isolation. Nevertheless, there is a clear pattern and it is highly surprising.
In the 90s—when Pete Sampras was ruling Wimbledon with serve-and-volley—more players had high returning percentages than they did in the 00s.
Here are the numbers of players clearing 30 percent every other year, followed by the number of players clearing 25 percent (note: I've included the each year from the early part of the 00s, as this is when serve-and-volley presumably died).
12 (91), 7, 10 (95), 9, 6, 5 (2000), 10 (2001), 13 (2002), 15 (2003), 6 (2004), 6 (2005), 2, 4, 5 (2011).
32 (91), 21, 35 (95), 29, 23, 20 (2000), 30 (2001), 34 (2002), 25 (2003), 20 (2004), 16 (2005), 13, 16, 18 (2011).
If anything, these numbers indicate that the the return player has had a more difficult time at Wimbledon after 2004.
2001—the last time a serve-and-volley player won Wimbledon (Goran Ivanisevic)—was one of the better years for returners (though you could argue that Roger Federer was at least partly a serve-and-volleyer when he won in 2003). So are 1995 and 1997, years when Sampras was at his peak.
2002 (featuring the first pure baseliner Wimbledon final) and 2003 were also good years for returners—but after 2004 the numbers drop.
So, what are we left with?
We can probably claim that clay has become faster, especially after 2007, making it harder for the return player to win games.
But on grass and hard court, the two surfaces more benevolent to the return player—and the two surfaces that have supposedly become slower—we still don't see any sign of a more level playing field
If anything, since 1995, hard courts have been harder for the returner—the same applies to grass since 2004.
How can this be?
There are, as I see it, two or three possible explanations, but I'm open for other suggestions as I'm genuinely puzzled by the data.
1) The courts aren't actually slower (which is, according to Federer and others, false).
2) Since string and racquet technology has indeed improved the possibility of taking larger swings at the ball at times, the server actually has benefited more than the returner.
3) The game has simply evolved—players are fitter and can run all day, whereas in the past they needed to rush the net in order to preserve energy. But why has this happened?
4) Donnie Brasco's explanation: It's easier to break serve in a serve-and-volley game (compared to a baseline game) because the serve-and-volley game is inherently more risky. String technology has enabled better returns and thus forced the death of serve-and-volley.
But it's also forced players to combat the entire baseline game of the server, rather than play serve-and-volley.
I don't have data for the average serve speed over the period, but the number of aces players make during a year is more or less constant, with about 10 players a year hitting at least 500 aces.
Nevertheless, serve speed has probably increased, as is evidenced by Pete Sampras, universally regarded as one of the best servers of all time, clocking serves at up to 132 mph at the US Open 2002—something that Nadal, not known as the best server, replicated during his 2010 US Open campaign.
Is this the reason why return-game players, statistically speaking, haven't gotten better during this decade of slower surfaces and better string technology?
Perhaps it's part of it, but I'm still bewildered and surprised by the data.
Is Donnie's explanation the best we've got?
Given that grass and hard court used to be faster, we would expect the elite to have had better hold games on those surfaces than they do today.
However, this is not the case on grass (I don't have time and patience to plot data for every single year again, but I've looked at the statistics, and the hold game has become better both for elite and average players. You can play with the numbers here.)
It also isn't the case on hard court, though the elite hold-game players in the 00s weren't much better on this surface than on grass.
On clay, as expected, the hold-game elite have been better in the 00s than in the 90s—especially after 2007. We are starting to see hard court-like numbers.
Is the explanation simply that serve-and-volley is more risky and leads to more breaks? But why did the evolution to baseline tennis come later, rather than sooner?
Have string and racquet technology developments aided the server more than returner? Enough to reverse the effects of the slower hard- and grass-court surfaces?
I'm still not sure what the answers are, but the facts are contrary to what I expected prior to my research.
I'm keen to have a debate on it—hopefully we can come to some collaborative conclusions.