Taking his usual sideways look at the data behind the men's game, Jack Houghton looks at a funky new metric to help predict whether a player is getting better or worse...
"Players I will be more aggressive with include Miles Raonic, Andy Murray and Kei Nishikori. Raonic reached his peak rating in the semi-finals at Wimbledon this year and has broadly maintained that level since, whilst Murray and Nishikori have never played as well as they are playing right now - both peaking in Cincinnati..."
I've always been interested in the arcs of tennis players' careers: where they start; the heights to which they rise; and the timing and rate of their decline. A few years ago I wrote this analysis of the career of Roger Federer and suggested that, when comparing his record to that of tennis greats from the past, his title-winning days were largely over.
Messing around with my tennis ratings in the last week, I've found myself returning to this question of how players improve and decline over the course of a season, and the course of their careers.
The method I have been using is simply to note when a player reaches their peak rating, and how their current rating matches up to that peak. Although not an especially sophisticated analysis, it throws up some interesting talking points.
Federer, absent from the US Open with continuing knee and back problems, was at his best in 2007, and hasn't got close to recovering that best since: he declined slowly over the course of a few seasons, and has been consistently around 92% of his peak rating for the last five seasons.
Federer's numbers offer no great insight - anyone with a passing interest in the men's game could predict what the statistics would say - but other players possessing relatively old peak ratings might be more of a surprise.
Although now 31, Jo Wilfred Tsonga has seemed to be a better player in recent seasons, reaching the semi-finals of both the French and US Opens in 2015. The data suggests otherwise, though, showing that his peak came in 2009, and that he is only 93% of the player he was back then.
Marin Cilic, only 28, and the winner of the US Open in 2014, is seen by some as a viable outsider to reclaim his title, especially after he beat Andy Murray to win in Cincinnati last week, but his best came in 2010, and he is currently only playing at 95% of that best.
Juan Martin Del Potro has had some strong performances this year, most notably beating Wawrinka at Wimbledon and in winning a silver medal at the Olympics. This has led to many suggesting that he is reclaiming the best of his form and perhaps explains why he is the fourth-favourite at around 27.0026/1 to win the US Open. His rating tells a different story, though, suggesting that, despite a few good results, his overall form is only around 92% of what it was when he won this title back in 2009 as a 20-year-old.
The usefulness of this percentage-of-peak-rating metric is questionable. Many would argue 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.
From a purely statistical point of view, this is true, but it's worth remembering that ratings are a retrospective analysis - they make an assessment of a player's ability based on all of their matches up to that point - and, whilst they are pretty good at predicting how a player might perform in their next match, they are unable to predict with any certainty whether that player's rating will rise or fall in the future.
In an admittedly crude piece of analysis, I've looked at various player's ratings over time as a percentage of their peak rating and, when a player begins to decline below about 96% of their peak rating, they are unlikely to ever recover to anywhere near their best, especially if they are in their late-20s or older.
For a punter, this is significant, as it helps us to understand which players are most likely to outperform, or underperform, the rating they are currently on. In the case of Jo Wilfred Tsonga, Marin Cilic and Del Potro, I'll be erring on the side of caution when pricing up their matches at this year's US Open, in the expectation that they may not be able to play to the level their rating suggests.
Players I will be more aggressive with include Miles Raonic, Andy Murray and Kei Nishikori. Raonic reached his peak rating in the semi-finals at Wimbledon this year and has broadly maintained that level since, whilst Murray and Nishikori have never played as well as they are playing right now - both peaking in Cincinnati.
Djokovic and Wawrinka have also never been as good as they have in 2016, although for both players their peak ratings came much earlier in the season. Given their ages - 29 and 31 respectively - is this the start of a decline for both of them?
Back Andy Murray @ 3.1085/40
Back Miles Raonic @ 22.0021/1
Back Kei Nishikori @ 30.0029/1