David Ferrer and Defiance of the Aging Curve (+Updated WTForecast)

Italian translation at settesei.it

At the end of 2009, aged 27, David Ferrer finished the year with an ATP ranking of 17.  It had been a rough 15 months.  A poor pair of Masters events at the end of 2008 knocked him out of the top five, all the way down to 12.  An indifferent season saw him fall out of the top 20 for a few weeks.  Many players never improve upon their mid-20’s form, so had things gone according to script, Ferrer might still be kicking around the high teens.  His near contemporaries Mikhail Youzhny and Tommy Robredo have followed paths of that nature.

Instead, the Spaniard has only gotten better.  He finished 2010 back in the top ten, at #7.  At the end of 2011 and 2012, he sat at #5.  He’s likely to conclude 2013 at his career-high position of #3.  All this at the age of 31, when many players have shifted focus to their golf games.

This is unprecedented.  Ferrer is only the 12th player of the last 30 years to string together four consecutive year-to-year ranking improvements starting at age 24 or later.  He’s only the second to do so starting at 27, and no one has done it from a more advanced age.  The only other man to match Ferru’s current streak doesn’t really compare: Wayne Arthurs improved his ranking from 1998 to 2002 up to an ’02 year-end position of #52.

Admittedly, this streak is a bit of a sideshow curiosity.  But the underlying issue it reveals is more significant.  Even in an era of 30-something stars on the ATP, tennis is a young man’s game.  At the age when Ferrer began his resurgence, most players are fading, if they’re not already gone.

The exact trajectory of the aging curve depends on the data you choose to examine.  I ran the numbers twice: first with all players in the top 300 since 1983, then limited to players born in 1975 or later.  With the bigger dataset, the apparent peak is at age 23-24.  The average player maintains their level from their age 23 season to their age 24 season, but every year beyond 24 brings with it an increasing decline.  For instance, if we set aside those who disappear from the top 300 entirely, 45% of players improve their ranking in their age 25 season, while 2% maintain it and 53% decline.  At age 26, it’s 38%, 2%, and 60%, while at age 31, it’s 30%, 1%, and 69%.

The following graph shows the percentage of players who improve and decline in the rankings at each age.  While there are still a few guys like Ferrer who post a year-to-year improvement at any age, they are harder to find at each successive age.  Also, keep in mind that the later-career numbers include players returning from injury–Lleyton Hewitt, for example, has improved his ranking each of the last two years.

ferru2

Limiting our view to those players born in 1975 or later, we have a smaller dataset, but one that should better reflect the current state of affairs.  Here, the peak is one year later, at age 24-25.  Despite the Ferrers, Roger Federers, and Radek Stepaneks who seem to be rewriting the rules, it is still the case that only 42% of 26-year-olds improve their rankings from their age 25 season, while 3% maintain and 55% decline.

Another way of looking at the decline is by measuring and then aggregating the magnitude of ranking changes.  In the dataset limited to 1975-and-later births, he average player loses roughly 2.5% of his ranking from his age 25 season to his age 26 season, and almost 19% of his ranking from age 31 to age 32.  Using this metric, here is a graph of two “decline curves”–ranking position lost at each age.  Both the overall dataset and the more limited, recent dataset are shown:

ferru3

While the overall direction hasn’t changed from the 80s to the present, the trend in magnitude is clear. At every age in the decline phase, the curve has flattened out, making it a bit more likely that someone like Ferrer would improve throughout his late 20s.

Keep in mind that we’re only measuring those players who remain in the top 300.  Those who retire or fall out of the rankings due to injury aren’t considered, so the actual effect of age–in either dataset–is more severe than these numbers represent.  However, without forcing those guys to play, we can only estimate their aging patterns based on those who do stick around.

Having determined the percentages of players in the current era who improve and maintain their rankings at each age, we can calculate the likelihood that someone would do what Ferrer has done, keeping his ranking moving in the right direction from his age 27 to his age 31 season.  For any individual year, the chances are about 40%, giving us an overall probability of roughly 2.5%, or 1 in 40.  Even limiting our scope to the pool of players in the ATP top 300 at age 27, that seems reasonable–Ferru is, at the very least, a 1-in-40 aberration.

Ferrer’s biggest test yet will be his age-32 season in 2014.  Of players in the current era, 18% of 31-year-olds fall out of the top 300 by the end of their age-32 season.  (In the bigger dataset going back to 1983, 27% disappear.)  Of those who remain, only a quarter improve, and the average ranking change is strongly negative.

Eventually, nature will stop David Ferrer.  Precedents or no precedents, though, he’s a hard man to bet against.  He hasn’t been particularly constrained by nature thus far.

London forecast: After today’s Group B matches, Djokovic is guaranteed a spot in the semis, while Federer’s match on Saturday with del Potro will determine the other semifinalist.  My ratings consider those to be nearly equal on this surface, giving the slight edge to Delpo.  Here is the complete forecast:

Player     3-0  2-1  1-2  0-3        SF      F      W  
Nadal      70%  30%   0%   0%     98.5%  57.9%  33.3%  
Djokovic   73%  27%   0%   0%    100.0%  65.8%  36.3%  
Ferrer      0%   0%  54%  46%     14.7%   5.1%   1.9%  
Del Potro   0%  52%  48%   0%     52.3%  23.3%  10.7%  
Federer     0%  48%  52%   0%     47.7%  20.1%   8.8%  
Berdych     0%  30%  70%   0%     35.8%  11.9%   3.9%  
Wawrinka    0%  46%  54%   0%     51.1%  16.0%   5.2%  
Gasquet     0%   0%  27%  73%      0.0%   0.0%   0.0%

My algorithm doesn’t capture all the complexity of the tiebreak rules, so it’s got Group A a bit wrong right now.  Nadal has locked up a spot in the semis. To clear up any remaining confusion, we’re lucky to have Anna, who lays out the qualification scenarios very clearly for both Group A and Group B.

Today’s matches: I charted both Group B matches today, so there are detailed serve, return, and shot-by-shot stats for each one.  Here is Federer-Gasquet, and here’s Djokovic-del Potro.

Finally, it’s already time to look ahead to Melbourne, as Foot Soldiers of Tennis is monitoring the players on the cusp of direct entry.

Discover more from Heavy Topspin

Subscribe now to keep reading and get access to the full archive.

Continue reading