In 1988, a Swede named Kent Carlsson ended the year ranked No. 6. The five guys ahead of him were Mats Wilander, Ivan Lendl, Andre Agassi, Boris Becker and Stefan Edberg. Jimmy Connors was No. 7. Other than Carlsson, all of them finished their careers with six to eight singles Slams. Carlsson had a 50-6 record that year (just a tad worse than Federer’s 52-5 in 2017). All of those matches were on clay. If Carlsson had decided to play the 1989 Australian Open, he almost certainly would have been seeded fifth (Agassi didn’t enter), even though he hadn’t played on a hard court in the previous 23 months.
I was thinking of Carlsson — but in a bizarro world kind of way — when I saw that Sam Querrey, No. 13 in the ATP rankings, was seeded first at this week’s tournament in Geneva. Then I looked at the draw. Of the 27 other players in it, I probably wouldn’t pick Querrey to beat any more than six of them in a match on European red clay. I imagine that most knowledgeable tennis fans would have made Querrey a slight underdog against his first-round opponent, Guido Pella. Sure enough, Querrey lost 7-6 in the third, a rather predictable outcome.
This happened as a debate was simmering over how much flexibility tournaments should have in seeding players. The discussion centers around whether the French Open would seed Serena Williams (it didn’t), who has played only two tournaments this year after returning from childbirth; and to a lesser extent Victoria Azarenka, who has played only four tournaments due to her child custody dispute.
One question around Williams is how long the WTA should protect a player’s ranking after she comes back from having a baby, and whether the protected ranking should also include a protected seeding, which it currently does not. I’ll leave aside for now the specific return-from-pregnancy question, and instead discuss the ranking-seeding relationship more broadly — because this issue is also about how much flexibility tournaments should have when seeding players.
My view is that seedings, first and foremost, need to pass the smell test. There should be criteria, but criteria with sensibility built in. Querrey is a good player. He has earned his No. 13 ranking — generally. But he should not be seeded No. 1 in a European red clay tournament. Any decent tennis fan would look at the Geneva seedings and think, “C’mon, that doesn’t make sense.”
So how can that be fixed? First, we need to start ditching some long-standing assumptions. It is understood in the World of Tennis that a ranking number corresponds directly to a seeding number. It’s the way things have been for a long time, so players, fans and journalists have come to accept it. The thinking is that if you’re ranked, say, No. 7, you’ve earned the right to be seeded at least No. 7 at Roland Garros, even if no knowledgeable tennis follower thinks you are anywhere near the seventh-most likely winner of the tournament. I would argue that the assumption should be different. The assumption should be that if you’re No. 7 and you don’t have good clay results, the tournament can knock you down several spots.
I’m not saying there should be no relationship between ranking and seeding. Of course there should be. But the tie should be much less direct than it is now. Perhaps rankings should account for about two-thirds of a seeding criteria, with the other third coming from a mix of surface aptitude and recent results. Tournaments should have to justify their position. Show their work, so to speak.
Wimbledon has a formula that allows straying slightly from the rankings, but I’m suggesting even more flexibility. Other than Wimbledon, seeding is a rules-based, or bright-line system. You could also have a principles-based system with no hard rules, which would be too loosey-goosey. Or you could have a hybrid, which is what I’m advocating.
Look at the NBA draft lottery. The team with the worst record doesn’t necessarily get the top draft choice. It’s a weighted lottery system. I know this isn’t a great analogy, but I use it to make the point that you can change the expectations of the participants. Before the 1980s, the teams drafted in precise reverse order of win-loss record. Similar to today’s tennis seedings, NBA teams back then knew their exact draft position as soon as the regular season was over. It was a direct relationship. But the system was changed almost three decades ago. Now people are used to the weighted system and barely remember how it was done before.
Over time, you can change the expectations of players. Today, if a player goes into the French Open ranked, say, 29th, he knows he will be among the top 32 seeds. But there shouldn’t be any such guarantee. A player ranked 29th going into the French should think: “Well, that’s probably good enough to be seeded, but it’s not a sure thing.” There should probably be a cut-off at which you know you’ll be seeded. Maybe around 20th. So we just need to change the expectations. If you want to ensure a seeding, you need to go into the Slam ranked in the top 20. Anything lower and you can’t be sure. When they go to 16 seeds, I’d say top 10 should assure you a seeding. But not anything below that.
I reject the notion that a seeding should be a reward for a good ranking. Querrey has not earned being seeded No. 1 at a European red clay tournament. His ranking isn’t based on his play on red clay. On hard courts and grass he has earned a seeding that reflects his ranking. On grass, perhaps even better. If Wimbledon bumps him up from No. 13, that would make sense. But on red clay in Geneva, he should be dropped several notches.
I should reiterate that if tournaments stray from the rankings, they should have to show how they came to that decision. Rome can’t give Fabio Fognini a top-four seed just because they want to help the local. Wimbledon can’t shoot Andy Murray to the top of the seedings. But if Roland Garros wants to seed No. 14 Roberto Bautista Agut ahead of Querrey, it should be able to. It can be justified based on surface. And most important, it would make sense.