Sportico is proud to partner with The Harvard Sports Analysis Collective, a student-run organization dedicated to the quantitative analysis of sports strategy and management.
The Australian Open, ensconced in turmoil and controversy over Novak Djokovic’s immigration status, still promises to be one for the history books with many competing storylines. Djokovic, deported after the country canceled his visa, will not be able to defend his title. Meanwhile, that creates an opportunity for Rafael Nadal to pass Djokovic and Roger Federer by being the first reach 21 Slam titles while playing in his first of the big four since being ousted by Djokovic in the semis of last year’s French. The door is also open for others like Daniil Medvedev, coming off his first Slam win at the U.S. Open, where he dominated Djokovic in three sets. Alexander Zverev—who won the ATP Year End finals, beating Djokovic and Medvedev in the semis and finals, respectively—is a threat. Of the Top 10 players in the world, eight of them are 25 or younger. Since 2006, only one player (Stan Wawrinka in 2014) has taken the title outside the Big 3, but 2022 seems to be the most open it’s ever been for someone to capitalize.
Trying to predict outcomes with probabilities is common in sports. But it is much easier to do in a single game between two teams or even your standard 16-team playoff bracket than it is for a Grand Slam tennis tournament with a draw size of 128. This is because a lot of the probabilities are interdependent and can change meaningfully with each result, particularly unexpected ones. For example, with Djokovic out of the tournament altogether, Medvedev’s probability of winning went drastically up since he would no longer have to play the world No. 1 in the finals.
Nevertheless, we can still try to predict the outcome of the tournament through a simulation. For my simulation, I used Tennis Abstract’s Elo ratings, which can be adjusted for different surfaces. For example, Medvedev ranks first in hard court Elo, but eighth on clay since that is a weaker surface for his flat strokes. Now with the actual draw, we can simulate the tournament thousands of times using Elo ratings in order to see the different potential outcomes. We used this same approach to simulate last year’s Grand Slams. This will ultimately give us a probability of each player making a certain round or even winning it all. For instance, if Medvedev wins the simulation 3,000 out of 10,000 times, he has an approximately 30% chance of winning. While there is seeding in tennis (No. 1-No. 32) the draw is more random than in other professional playoffs, which rely purely on seeding. In a Grand Slam, a No. 1 seed could play a No. 17 seed or a No. 32 seed in the third round. This creates many interesting matchups throughout the tournament with potential upset opportunities of the star players. Djokovic’s late scratch from the draw also creates an opening in his half of the bracket since the seeds were not reshuffled due to the timing. Especially as the first Grand Slam of the year with few warm-up tournaments, players’ rankings and ratings might not be as reliable, as some might have taken a big leap forward during the off-season while others regressed.
Nadal returns for his first Slam since falling at last year’s French Open, where he has only lost twice in his career. Following that defeat, Nadal withdrew from Wimbledon and the Olympics with a foot injury, tried to return for the U.S. Open by playing the Washington warm-up (lost to Lloyd Harris in the second round), and then pulled out for the rest of the 2021 campaign. While Nadal has only played one warm-up tournament, he took the title in Melbourne, where he faced middling competition (average opponent ranking was 87 in three matches). What works against Nadal is priors and his history at the tournament: the Australian Open is his second-weakest Slam behind Wimbledon, with an 82% winning percentage (which is still remarkably impressive) and just one title (2009). Compared to the French Open, in which he wins 97% of his matches, Nadal’s game style, with heavy lefty spin and longer rallies, is not as well-suited for the hard courts. Even if Rafa doesn’t win the Australian Open, he will be in good position to take the all-time Slam lead in the Big 3 race at Roland Garros. He enters the field as the No. 5 seed but has the third-highest title odds at 7%.
Medvedev is now the favorite to win the Australian Open and is the highest seed at a Slam for the first time in his career. Fans might be surprised that even before Djokovic’s withdrawal, Medvedev was actually slightly favored over him (27% vs. 23% chance at the title). This was due to his higher hard court Elo ranking. Djokovic is rated higher overall, due to his versatility on clay and grass, but of late Medvedev has had better results on hard courts. One could argue that he plays more tournaments and his ranking might be inflated from those results, but nevertheless he is the favorite (31% chance) to win in Melbourne and claim his second Slam in a row. Starting from his U.S. Open championship run, Medvedev is 21-4 in his last 25 matches. As Djokovic will not be able to defend his championship, Medvedev could be World No. 1 soon as he trails by only 2,000 points.
Like Medvedev, Zverev also had a career campaign in 2021, winning six titles including the Masters 1000 in Madrid and Cincinnati, the Tokyo Olympics and the Nitto ATP Finals. Nobody in the draw benefits more from Djokovic’s withdrawal than Zverev. Now that he doesn’t have to face Djokovic in the semis, his odds have greatly improved. He went from a 30% to 40% chance of making the finals and a 25% chance at the title, up from 19%. He still has a tough path, as he would likely have to face Nadal in the quarters (3-6 in career). But he’s now the favorite to emerge out of the top half of the draw, and fans will be hoping for a very competitive Medvedev-Zverev final (6-6 in career).
21 and Unders
When looking at the traditional rankings, there are not many players 21 or younger near the top of the list. There are only four in the Top 50. So why are some high up on the probability list when their ranking or seeding would suggest otherwise? These younger players have not been on the tour as long and have not built up as many points as some of the veterans. After all, it takes a certain number of points to qualify for a Grand Slam in the first place. Take Jenson Brooksby, for example: He started out 2021 playing Challengers against players ranked outside the Top 200. He didn’t even receive entry into Australian Open qualifying last year, had to win three rounds of qualifying at the French before losing in the first round, didn’t make it to Wimbledon, and then made a run to the fourth round of the U.S. Open before losing to Djokovic. Brooksby is ranked No. 56 in the world right now, a career high, but is 11th in hard court Elo. The eighth-ranked player by Elo, meanwhile, is 18-year-old Carlos Alcaraz, who had impressive runs to the third round at the French, the quarters at the U.S. Open, and was the ATP Next Gen Champion. These up-and-comers join the ranks of more established youngsters, like Felix Auger-Alliasime and Jannik Sinner, who are both seeded in the Top 10. With so many high-level youngsters in the draw, it’s likely that one will make a deep run.
Ever since reaching his first Slam final at Roland Garros last year, pushing Djokovic to five sets, Tsitsipas has struggled. He was ousted in the first round of Wimbledon and the third round of the U.S. Open. He also suffered early-round exits at the Vienna Open and Paris Masters towards the end of the season. The Australian Open is still the Greek’s second best career Slam (73% winning percentage) after the French. He has reached the semis twice, losing to Nadal in 2019 and Medvedev in 2021. In Medvedev’s bottom half draw, he’ll be hoping that the Russian somehow gets upset early as he’d be a considerable underdog in that match (27% chance of winning).