By Nick Carter
Tennis rankings came up in conversation after a bizarre all-caps tweet from Nick Kyrgios.
The Australian was sharing a Tennis Channel graphic that suggested that according to the Universal Tennis Ranking (UTR) system, he would be ranked 2nd in the world behind Novak Djokovic, as opposed to 22nd.
This of course caused some debate over what kind of qualities rankings should reflect. UTR uses the below method, explained by Josh Vernon of Universal Tennis:
“For each match, the algorithm calculates a match rating and a match weight for each player. A player’s UTR Rating is the weighted average of up to 30 of their most recent match ratings. Only matches within the last 12 months count toward a player’s UTR.”
So, it’s understandable why this would benefit Kyrgios, whose most recent results were from the mid-summer period where he had some strong runs on grass and U.S. hard courts. The UTR system benefits him because it doesn’t require players to compete through the year (or at least half of it) like the ATP rankings.
For comparison, the ATP ranks players based on their best 19 performances across a 12-month period. Performance is not decided on a match-by-match basis, but by assigning points to tournaments based on stage reached (so the further you get, the more you score, on an increasing rate with each win), with some having more weight based on status (e.g. you gain more points from a major based on any other event).
I would be remiss at this point not to mention ELO as a ranking system, which is favoured by Jeff Sackman who runs Tennis Abstract. The full breakdown is here but the simple explanation is that it is not about which matches a player wins in terms of stage of a tournament, rather it is based on the quality of opponent. So, a result’s weight is based on whether the victor has beaten an opponent with a higher, equal or lower rating. Sackman’s rankings also take into account a 12-month period. Most systems use this timeframe, which makes sense as a 12 month period encapsulates an entire season and its many diverse conditions, making it a good indicator of “form.”
Let’s look at the four qualities Kyrgios has highlighted. ‘Consistency’ is an underrated trait in tennis. The mark of a top player is not just that they can win events and beat the best in the world. All professional tennis players have the capability of doing this at least once in their career, or at least those who crack the top 100 at one point. Their ‘skill’ level is going to be high and the margins between the top players are very small. The fact you can beat most players, most days, is just as much a sign of being one of the most elite players as anything else. Jessica Pegula on the WTA Tour is the perfect example of this, regularly featuring in at least the quarter-finals in most tournaments that she played in and half the time only losing to the eventual champion if she didn’t win herself. This is what got her #3 singles ranking on the WTA (which is based on a similar premise to the ATP, but takes the best 16 results rather than best 19). A player can have the capability of charging through the draw and being in contention based on past results (like Kyrgios, Grigor Dimitrov or Dominic Thiem on the ATP or Naomi Osaka, Garbine Muguruza or Eugenie Bouchard on the WTA) but it’s more impressive when you are doing this on a regular basis (like Casper Ruud, Stefanos Tsitsipas, Taylor Fritz, Pegula, Daria Kasatkina and Iga Swiatek).
‘How much you play’ is very much linked to the above. Now of course favouring this in a ranking does hurt players who are physically injured or need to take time out for their mental health, parental leave, or other reasons. The Protected Ranking system mitigates some of the disadvantages, but given how most players come back to the tour not quite the same, at least initially, they should have to prove themselves again. Even if someone plays week in week out, and has a great last few weeks on tour, they have to show that they can still maintain that kind of performance whenever they step on court. However, this goes back to the same debate. Do you favour rewarding peak results or consistent results?
Let’s take a step back here and ask another question: what are rankings for? Why does tennis use them? First, they’re useful for structuring draws, especially to set up the most interesting final possible. This is not just because of the potential that it would be the best two players on paper, but also means that for events with bigger entry lists, the leading competitors are usually better spread throughout (though not always). This leads to the second reason, which is foundational to the first: It helps everyone make sense of the sport. ‘Everyone’ includes players to help them understand where they are in their development, but it is mainly for any off-court observers, especially fans and the media. We want to understand if one player is significantly ‘better’ than the other, which should inform us if their upcoming match will be interesting and for what reasons. Without a formalised computer system, we’d be using one based on reputation based on previous results, even if it was just the previous year’s results from the same event. Rankings set the context of the stories that unfold in matches and across the tournament, looking at who the favourites and underdogs are. People don’t usually like not knowing things and want to have some comfort in understanding something about what we’re going to see, so rankings give us something to hang on to.
So let’s go back to the earlier question: Should a ranking system be based more on peak results (ELO and UTR) or consistent results (ATP and WTA)?
People who maybe are less familiar with the tour often believe things are based on peak results over a career, so it’s probably simplest to understand. And it’s odd that given we know the world-class tennis they can produce, that Novak Djokovic and Rafael Nadal are not always number one and two and swapping the top spot between them. But then would it have been fair to rank Serena Williams number one in every tournament she played in 2022 given how rusty she was? That’s why there’s a 12-month window so that more recent results can influence things and help it make more sense to those who follow the sport a little more closely. After all, that’s how we got that magnificent Australian Open final where somehow Rafael Nadal won despite being an underdog against a gifted but nowhere near as decorated opponent in Daniil Medvedev.
This is where my categorisation of the different systems is shown up to be a little too simplistic. The Tour ranking is weighted in favour of peak results in terms of the value of tournaments won (e.g. winning a major in a season more or less guarantees someone a top 10 spot in the rankings). Likewise, ELO does reward consistency–a ranking goes down following a loss, how much depending on the ranking of the opponent. So, it still rewards those who go deep in tournaments regularly.
Often (though by no means always), the winner of a major has to have beaten a highly motivated elite player on the way. The next question is whether you define ‘peak’ results on who you beat or what events you win? Often it ends up being more or less the same thing. For example, in 2022 Carlos Alcaraz’s biggest point yields were the US Open title, followed by his Madrid and Miami trophies. However, if ELO was used it would be his Madrid title that was his biggest yield, then his wins at Miami and the US Open. The difference might not be major, but certainly emphasises different things.
Regardless, it makes sense to think that someone with higher potential should be favoured, but then there’s the question of how likely are they to meet that standard at an event. If their 12-month form is patchier, where let’s say they’ve won a Masters 1000 but struggled to get past the round of 16 most other events they’ve entered, it’s hard to gauge if they’re a top 4 or top 16 contender at a major. For example, if Nick Kyrgios is 2nd best according to UTR but hasn’t won a tennis match he’s played for 3 months building up to the Australian Open (hypothetically of course), how much of a contender do observers perceive him as? This is where we see the value of consistency as a trait to be rewarded.
The rankings for the most part strike a decent balance as they are. With one exception: ‘form’. In this case, ‘form’ means more recent results (for example the previous month or 6 weeks). Often, these influence perceptions of a player about to enter, say, a major. There is some benefit to players as they get a boost from a strong upturn in ‘form’ but how much depends on the level of the tournament.
A suggestion would be to weight 12-month rankings more heavily in more recent results. Essentially in a 12-month ranking, results from over 3 months would reduce in worth by 25%, 50% for over 6 months and 75% for over 12. So, this would mean players like Beatriz Haddad Maia at Wimbledon or Caroline Garcia ahead of the US Open would seem a lot less like dark horses and more like genuine contenders, as their strong recent results would be weighted more heavily. Likewise, majors that just ended would have more of an influence. In this version, Rafael Nadal would become world number one after Roland Garros in 2022. In fact, the rankings at the end of the season would be led by Alcaraz (U.S. Open champion), closely followed by Djokovic (World Tour Finals champion). The benefit is that consistency is still rewarded with this system because results through the year will create a stronger ranking, whilst a peak would have strong benefits for the short term but lose its value as the year progresses.
The main purpose of this piece was to get everyone thinking, and asking themselves what they would prefer to prioritise in a ranking system: consistency or peak performance, and whether you reward big matches or trophies more. Let us know which system you prefer, why, and what’s important to you. It’s not about pushing for a particular system, although I will say the method the tours use works well from my perspective.