Replay, Episode 2: What's the greatest upset in US Open history?By Nov 10, 2020
Rajeev Ram, Joe Salisbury plan to continue partnership following second Slam titleBy Sep 16, 2021
US Open's return attracts 631,134 fans to groundsBy Sep 14, 2021
Emma Raducanu's US Open triumph garners blockbuster ratings on British TVBy Sep 14, 2021
Recognizing the value of a disarmingly honest Daniil Medvedev and his PlayStation-inspired celebrationBy Sep 13, 2021
Med Man: Daniil Medvedev makes history of his own in stunning US Open final defeat of Novak DjokovicBy Sep 13, 2021
Daniil Medvedev wins US Open, and ends Novak Djokovic's chance at a calendar-year Grand SlamBy Sep 12, 2021
The Rally: On the 20th anniversary of 9/11, our memories of that day and the 2001 US Open, and what this year’s Open has meant to the New York City and the sportBy Sep 12, 2021
Totally Rad: 150th-ranked Emma Raducanu won an all-Cinderella US Open final with clear, uncomplicated tennisBy Sep 12, 2021
Emma Raducanu, Leylah Fernandez cap a women's US Open tournament like no otherBy Sep 12, 2021
Replay, Episode 2: What's the greatest upset in US Open history?
In order to find out, we needed to create a formula that every match could be run through, to quantity their magnitudes.
Published Nov 10, 2020
In 2018, 50 years after tennis' Open Era began, Tennis Channel put together a list of the period's Top 10 upsets. That list featured six matches from Wimbledon, two from Roland Garros, and one apiece from the Australian Open and US Open. The fact that both hard-court upsets had taken place within just the previous three years—Roberta Vinci’s US Open victory over Serena Williams in 2015, and Denis Istomin’s upset of Novak Djokovic at the 2017 Australian Open—was surprising.
It got me thinking: what makes an upset truly shocking?
After focusing on how a new scoring format could have changed 2020 Australian Open in the first episode of Replay, I turned my attention to the US Open, which has a history of head-turning results. And with the aforementioned question in mind, I set out to find a way to more objectively compare Flushing Meadows' many upsets against one another.
In order to do that, we needed to create a formula that every match could be run through, to quantity their magnitudes.
It’s important to point out that ‘objectivity’ doesn’t exist in statistics. There is always a choice made in which variables to include, and how to weigh them. But we tried to be as balanced as possible so that no single value could unilaterally influence the final rankings.
Total number in draw (minus) the player’s tournament seed.
Unseeded players = ½ of total tournament draw
Qualifiers/LL = total of tournament draw
Wildcards = either ½ or total, depending on circumstances
32 (minus) player’s ranking going into tournament
If no rank during tournament, = 1,000.
Age coefficients calculated by historical averages.
Time on Tour
Number of years a “pro”, multiplied by 5
Each previous title = 1 point
Each previous grand slam = 5 points
Record on Surface
Grand slams won on surface = +10
Titles won on surface = +5
Record at Tournament
Previous wins at tournament = +10
Previous Finals = +5
Previous Semifinals = +3
Previous Quarterfinals = +2
Non-cumulative each year
First time competing at tournament = -50
First time making it this far at tournament = -25
Prior head-to-head record
Percentage of previous matches player has lost
If no prior matches, = 50
1st round = 10
2nd round = 20
3rd round = 30
4th round = 40
Quarterfinals = 50
Semifinals = 75
Final = 100
ATP = 10 x Sets
WTA = 10 x (Sets + 2)
Je Ne Sais Quoi
Non-quantifiable circumstances that add significance to match, +50
Upset Score = [(Losing Player’s Player Score) - (Winning Player’s Player Score)] * Matchup Score
Player Score = (Seeding Score + Ranking Score + Time on Tour Score + Career Record Score + Record on Surface Score + Record at Tournament Score) * Age Coefficient
Matchup Score = Prior Head-to-Head Score + Tournament Round Score + Sets Played Score + Je Ne Sais Quoi
I don’t blame you if your eyes glazed over all that, but I wanted to provide the detail for those interested.
Applying this formula to our matches still took weeks, and filled a very large spreadsheet. In the end, these were the Top 10 US Open upsets we found from both the ATP and WTA, along with their scores:
We didn’t count withdrawals or retirements, so don’t go looking for Djokovic vs. Pablo Carreno Busta. Those are huge moments, but upsets for reasons other than the play.
The biggest surprise to me was the top upset of them all: CiCi Bellis defeating Dominika Cibulkova in the first round of the 2014 US Open. I won’t rehash why this match scored so highly—you can watch the video for that (see above)—but I will explain why it made me so excited:
Upsets are largely defined by who lost: Serena Williams failing to clinch a calendar slam in 2015; Jimmy Connors looking for a Wimbledon repeat in 1975; Rafael Nadal either of the times he’s stumbled at Roland-Garros.
But it’s just as important, if not even more so, that they have an exciting player across the net from them.
Don’t think of upsets as moments when a great player should not have lost. Think of them as moments when the next great player proved they should win.
There’s still a discrepancy between how competitive the WTA and ATP are, and I hope to explore that in a future episode of Replay. For now, I just hope we can acknowledge the significance of the US Open in ushering in the next generation of talent. This year saw Naomi Osaka and Dominic Thiem cement their places in tennis history. I can’t wait to see what 2021 holds.