Tue Aug 16 11:17am EDT
When Novak Djokovic begins play at the 2011 U.S. Open, he could theoretically draw one of the best unseeded players in the tournament, like Marcos Baghdatis or Ernests Gulbis, in the first round. Yet a recent study by ESPN shows it's far more likely that Djokovic and No. 2 seed Rafael Nadal will defy statistical probabilities and instead draw some of the worst players in the 128-man field. Same with Caroline Wozniacki and Vera Zvonareva, too.
An "Outside the Lines" report finds that the top two seeds at the U.S. Open have been receiving easier first-round opponents than expected for the past 10 years. In the blind draw of the 96 unseeded players in the tournament, a top seed would expect to see a mix of players near the top and bottom of the 96-player pool. Because the draw is random and all unseeded players are thrown into the same draw, there should be an even distribution of first-round matches.
ESPN finds that this isn't the case. In 1,000 simulations of the past 10 U.S. Open draws, only three simulations created an easier draw for the top two men than they've actually had in the past 10 years. None of the simulations created an easier draw for the top two women. A statistician said the odds of that happening on both sides of the draw are about 1 in 300,000, the same as flipping heads on a coin 18 times in a row.
Paula Lavigne and Alok Pattani write:
The top two seeds in each draw could have a first-round matchup with any unseeded player whose tournament rank is 33 through 128. Over the last 10 years, the average rank of opponents in the women's draw has been 98.5, and 97.2 for the men. A random draw should produce an average closer to 80.5.
It's a mathematical improbability for the first-round opponents of the top four players in the tournament to be so easy. What ESPN is hinting at (but not actually saying) is that there's something shady going on with the U.S. Open draw, whether it's intentional or not.
The USTA has numbers of its own, though. While zero percent of the ESPN simulations provided an easier draw for the top two women in the U.S. Open, the results for the French Open were on the other side of the spectrum. Over 99 percent of the simulations for the clay court major made an easier draw than the real ones. In essence, the numbers say the U.S. Open draw has been too easy for top women and the French Open draw has been too hard. Yet in those tournaments, the top two seeds are 20-0 in first-round matches. Top seeds in America lost two sets while top seeds in Paris lost a total of four. The decks may not be even but it's having no effect on results, the USTA claims.
ESPN takes careful pains not to accuse its television partner, the USTA, of anything nefarious. The only time the word "fix" is mentioned in either piece available on ESPN.com is when USTA director Brian Earley dismisses the notion as comical. It doesn't take a statistician to crunch these numbers, though. If both draws are statistical anomalies and the USTA commissions the draw, someone or something is to blame. The gist of these pieces aren't "look how weird this is," it's "isn't this a little too weird?"
But what, though? Earley is right when he says that there's no reason for the USTA to mess with the draw. (If anything, throwing a compelling first-round matchup could benefit what usually is a slow first three days of the tournament.) If the Djokovics and Nadals and Federers are going to win anyway, why give them another boost? At the other three majors with statistically viable first-round draws, the top two seeds haven't lost since 2003. Is the risk, whether it is taken by a rogue programmer or the USTA, worth the reward? It's almost certainly not. Still, not having a good reason to do something has never stopped anyone before. Nixon didn't need to spy on the DNC to beat McGovern in 1972 either.
Is there anything here or will this mini-controversy fade away without much notice? My guess is the latter unless there's a smoking gun nobody has found yet. In the meantime, Novak Djokovic better be ready for a first-round fight against a top-40 opponent in two weeks, just in case someone feels the need to even out the numbers.