The Tennis 128 will return tomorrow with player #126.
* * *
The good news: Elo ratings for tennis are popping up in more and more places, exposing an increasing number of fans to an alternate (and superior) system for ranking tennis players.
The bad news: A lot of people don’t yet understand how Elo ratings differ from the traditional points-based rankings. Many newcomers criticize either specific ratings or the system as a whole, often because they expect Elo to be just like the official rankings, only with slight tweaks to better match their own beliefs.
The purpose in one sentence: Elo ratings are designed to estimate each player’s ability level right now.
That’s it. The ATP and WTA systems are a hodge-podge of arbitrary decisions meant to balance results, willingness to play lots of tournaments, and performance in the latter stages of particular events. That doesn’t mean they are wrong–there’s often not much difference between the official rankings and the Elo list. You can devise an awful lot of methods for ranking women tennis players right now, and almost all of them will put Ashleigh Barty at the top of the list.
When the results do differ, it’s important to remember what the official rankings prioritize. In the rankings released today, Naomi Osaka is 85th on the WTA computer, against 12th in my Elo ratings. Despite the enormous gap, in a way, they’re both right. Osaka has played only seven events since last year’s Australian Open, so the WTA method treats her as a part-timer with a handful of decent results. Elo, on the other hand, recognizes that she won two of the last six grand slams. It rates her lower than it did a year ago, but Elo doesn’t simply forget about extremes of form because a magic 52-week window expires.
If you’re interested in what a player deserves (whatever that means), the ATP and WTA formulae are probably what you’re looking for. You may have some quibbles with the system, but everybody knows (approximately) how it works. If a player wants to crack the top ten, she understands what she needs to do, and at which events, to accomplish that.
If you’re interested in who will win tomorrow, Elo is almost always your better bet. The official rankings don’t even try to estimate a player’s current level. By definition, they serve as the average of a player’s performances over the last 52 weeks (unless a pandemic changes the rules), so the ranking is a decent approximation of their level five or six months ago. Osaka may not “deserve” a spot in the top 80, but most of us would be ecstatic to make an even money bet that she would beat #84 Anna Bondar or #86 Xinyu Wang. Elo’s estimate of #12 suggests a much more plausible range for how well she will play the next time she steps on court.
Just as the official rankings don’t try to estimate a player’s current form, Elo doesn’t concern itself with what players “deserve.” You might think that Gael Monfils has earned his spot in the top 20–after all, he won a tournament to start the year and followed it up with a run to the Australian Open quarter-finals. Elo had him several places lower, even before his opening-match loss to Mikael Ymer last week. Now he sits at #32. A major quarter-final is a nice achievement, but Elo recognizes that Monfils’s eight wins in Adelaide and Melbourne were against mediocre competition, including several players who were playing on their weaker surface.
Personally, I don’t care much about what players “deserve” from a system that–while adequate and widely accepted–is slapped together and incoherent. I’m more interested in who’s playing the best tennis, so Elo is exactly what I want to see. But that doesn’t mean you have to feel the same way. If you want your rankings to measure something else, that’s fine. Just don’t get mad at Elo.