The subject of longevity
The brief for this blog is to take a sideways, statistical look at the men's game in previewing each of the Grand Slams, hoping to identify some value bets. Armed with this brief, I've come back to the topic of career longevity - and how it links to match performance - on a couple of occasions over the years.
As we now find ourselves at an unusual point - historically speaking - in the men's game, it seemed worth revisiting it again, this time focusing on the setting career of Novak Djokovic.
Djokovic's place among the greats
In terms of longevity, Djokovic occupies rarefied space. Taking a list of the best players of the modern era according to their peak Elo rating, I have charted how long players keep winning top-level titles, starting the count in the season they win their first and ending it in the season they win their last.
Using this measure, Rod Laver tops the list, winning titles across 21 seasons. Djokovic currently sits fifth at 17 seasons, two behind Roger Federer, and one behind Andre Agassi and Rafael Nadal. In other words, Djokovic is achieving something exceptional. In doing so, however, it's reasonable to expect that his time must soon be up, if the careers of other greats are anything to go by.
One way of concluding this is to look at how long players keep winning titles after their peak. Taking the "peak" as the season in which players win the most titles, there is a remarkable consistency in how quickly careers dwindle. Eight seasons is the mean, mode and median, with the outliers being Borg, who unexpectedly retired when only 26, managing only two more seasons of winning titles, and Federer and Laver, who prolonged their competitiveness much longer than anyone else.
Assuming that 2015 was Djokovic's peak, when he won 11 ATP titles, it's reasonable to expect that he might well disappear from the winner's enclosure soon, unless he can replicate the outliers of Federer and Laver. Examining the success achieved by the others on the list in their final seasons would also suggest that the next couple of seasons will be relatively fallow for Djokovic: he may only pick up a handful of titles.
A Grand Slam winning paradox
A paradox sits at the heart of this analysis, though. Whilst it may seem as if Djokovic's powers are waning, it hasn't stopped him winning Grand Slams. He has won all of them this season, and won eight of the last 12.
That's some record for a player on the decline, and highlights the crudeness of career-longevity analysis: contextual factors, like how strong your competitors are in your twilight, will have a significant influence on how long the "tail" of your winning career is.
Elo gives more precision to the analysis
This is exemplified by a more precise analysis of a player's ability over time that compares how their current Elo rating matches their peak rating.
As an example, the last time Federer won a title in 2019, his Elo rating was at 91% of its peak. Performing this same analysis for the other greats on my list requires some estimation, as the data I am using to create historic Elo ratings is not perfectly complete or comparable. However, a broad picture does seem to emerge: when players hit around 90% of their peak rating, they don't win another title.
Despite his Grand Slam successes this year, Djokovic finds himself at a career-low Elo rating that is just 88% of its best. Nadal is on the same percentage, with Federer at 86%. Other lesser-lights, like Kei Nishikori, Milos Raonic and Nick Kyrgios, also find themselves languishing in the 80-per-cent brigade.
These figures are in stark contrast to the likes of Daniil Medvedev, Alexander Zverev, Stefanos Tsitsipas and Matteo Berrettini, who are all at, or within a couple of percentage points of, their peak Elo rating.
It's easy to pick holes in this analysis, arguing that it doesn't matter where a player's ability is relative to their own career, as long as they are better at that moment than the player they face across the net. And with Djokovic specifically, it could be argued that his decline-relative-to-himself is of little consequence in Grand Slams, where his experience and psyche count for more than his recent form.
These are fair points, but they don't fully repudiate the analysis, which at least suggests that Djokovic is not going to get better than he is now: other players may over-perform compared to their rating; Djokovic is much more likely now to underperform.
And this is his problem going into the US Open. I have his current Elo rating at around 2,150. Within 50 points of this rating are Medvedev, Zverev and Tsitsipas. To put that in context, should Djokovic meet any of those players, the pure ratings would suggest he would only have a 57% chance of winning, or put another way, that he would be the 1.768/11 favourite.
It's surprising, then, to see him as the 1.875/6-favourite for this US Open as a whole. His experience is one thing, but Djokovic is likely to see himself consistently challenged in matches he would normally have won comfortably, and that will make it increasingly hard for him to win Grand Slams, especially as younger players offer the promise of further improvement.