Since 2017, Serena Williams has held 23 major titles, leaving her just one shy of Margaret Court’s 24. The Williams-Court comparison forces us to think across eras in the same way that Federer-vs-Laver does, with the additional complication that Court has earned herself extreme dislike among many fans and fellow champions.
Let’s set aside the off-court stuff and work this out. The pro-Court case is simple: 24 is greater than 23, and you have to evaluate players relative to their own eras. The pro-Serena side is equally straightforward: 11 of Court’s 24 titles came in Australia, before Melbourne was a mandatory tour stop. Regardless of the era, Court’s home event was weaker back then.
As much as possible, I’m going to try to hold to the “relative to their own era” assumption. Everyone seems to accept it when it comes to Laver-vs-Federer. Plus, if we drop that constraint, the whole exercise is meaningless. With improved technology, fitness, and coaching, of course today’s players are better. But that’s not what people are talking about when they pick a side of Serena-vs-Margaret or Rod-vs-Roger.
Attentive readers of this blog might recall I took a stab at this problem back in 2019. That attempt relied on some extreme approximating due to the lack of pre-Open Era women’s tennis data. Regular readers will also know that the state of pre-Open Era women’s tennis data has vastly improved in the last few months. Tennis Abstract, plus the associated GitHub repo, now contains thousands of match results back to the mid-1950s.
Adjusting Australia
Let’s be clear: I’m not about to settle whether Margaret Court or Serena Williams (or someone else) is the GOAT of women’s tennis. That debate depends on much more than grand slam titles.
Today’s question is: How do Williams’s 23 titles stack up against Court’s 24?
That boils down to an even simpler question: How do Court’s 11 Australian titles measure up against other slams, then and now?
The anecdotal evidence is strongly anti-Margaret. As I mentioned in this morning’s Expected Points, the 1960 Australian Championships–Court’s first major title–had a 32-player draw (strike one), and 30 of those players were Australian (strikes two and three). Yes, it was a strong era for Australian women’s tennis, especially a few years later, but the tournament was hardly a showcase of international superstars. As such, it isn’t what we think of as a “major” tournament these days.
I’ve done a lot of “slam adjustments,” mostly to track the difficulty of the majors won by Djokovic, Federer, and Nadal. (Here’s the most recent.) The basic approach is simple. For each tournament, take the winning player’s draw, and for each match, calculate the chance that an average slam winner on that surface would beat that set of opponents. (Odds are determined by my Elo ratings, which are based on results before the event.) Take the resulting probabilities–on average, around 14% between 1952 and 2020–and normalize them, so that a mid-range slam draw is 1.0. Tougher draws are higher than 1, and easier draws are lower.
Equalizing the eras
This type of adjustment gets us most of the way there, but it doesn’t directly confront the “relative to the era” issue. The field in general was more lopsided in the 1960s than it is now, with a handful of very strong players swatting away a pack of also-rans who struggled to win more than a game or two per set against the elites. That in itself is a point in favor of Serena (and modern players in general), but again, on the Laver-vs-Federer principle, that’s not what we’re talking about today.
The easiest way to express this idea that all eras are equivalent is to use as a standard each season’s Wimbledon, the one tournament that everybody always wanted to play, and almost everyone actually did play. To avoid year-to-year fluctuations based on short-term injuries, we’ll make things a bit more resilient and compare the strength of each year’s Australian draw to the average strength of that year’s Wimbledon and US draws.
For example, my slam adjustments consider 1960 to be a strong year. Maria Bueno’s Wimbledon title was 40% more difficult than the average slam draw, and Darlene Hard’s US victory was about 30% tougher than usual. Court’s Australian title that year comes out as exactly average, so we compare Australia’s 1.0 to the average of Wimbledon and the US ( (1.4 + 1.3) / 2 = 1.35), and the 1960 Australian title, relative to the era, measures as:
1 / 1.35 = 0.75
The mostly-Australian field wasn’t as weak as the caricature makes it out to be, but it was weaker than the marquee majors that year.
Here is how the strength of the Australian draw has evolved relative to the other grass- and hard-court slams from 1952 to the present:
Except for an outlier in 1965, when Bueno, Billie Jean King, and several other international stars turned up, the Australian Championships was a second class member of the grand slam club until around 1980. It’s had plenty of weak years since then, as well, partly because of players who skipped due to injury, and partly due to contenders losing early, giving the eventual winners easier paths.
The main event
Margaret Court won the Australian 11 times. By this measure of relative strength, those titles were worth 62% as much as the other majors in those years. The strenght of individual titles ranged from a low of 0.29 in 1961, when no international elites made the trip, to a high of 1.02 in 1965, when the field was positively star-studded.
Serena Williams has won the Australian seven times. It is tempting to leave that “7” as is, because Melbourne is now a mandatory tour stop and virtually every woman on tour considers it one of the top targets in her season. However, we should treat Serena’s seven the same way we adjusted Court’s 11. For all the era differences, some things remain the same, like jetlag and the difficulty of playing top-flight tennis only a few weeks into the season.
Williams’s seven were worth, on average, 88% as much as the other majors in their respective years. The weakest of the bunch was her last, in 2017. So many top players lost early that Serena never faced a top-eight opponent.
Court’s 11 titles, then, are equivalent to about 7 non-Australian majors–a penalty of four. Serena’s 7 are worth about 6 non-Australian majors–a penalty of one.
The final, adjusted tally: Williams 22, Court 20.
Margaret Court was one of the greatest players of all time, but her position the all-time grand slam singles list depends too much on the shifting status of her home event. When we properly account for the Australian tournament’s position for decade as the most minor major, Court loses her remaining claim to the top spot. Serena may yet win 24, but to match or exceed Court, she shouldn’t have to.