In the days before the US Open every year, tennis fans wait for the two things that will shape their favorite players’ experiences there: the draw and the first Order of Play. The draw tells us who might get in the way of our favorite players making the later stages of the event. And the first Order of Play shows us which half of the draw plays on which cycle. With players getting a day off between matches at Grand Slams, the people who play on the first day will be the same people playing on days 3, 5, 7, and 9, while the people playing on day 2 will also play on days 4, 6, 8, and 10 (weather permitting). At the US Open this year, the bottom half of both draws played on the odd days, while the top half played on the even days. The halves of the draw come back together after the quarterfinals, and with those ending yesterday, we can take a complete look at how each side progressed.
This is not about who made it through the draw – semifinal previews will be posted on this site, but this is not that. It is about testing the sentiment that we observed, where the odd days (bottom halves) were universally considered far more entertaining than the even days (top halves). Using our Excitement Index, we quantify the level of excitement in each match, so we can plot the distributions Excitement Index (total drama in a match) by half of the draw to reveal the reality behind this sentiment.
These graphs are of the Excitement Index by match for each half of the draw. Bottom half in red, top half in blue. Men on the left, women on the right.
On the men’s side of the draw, we can see that the bottom half does not have nearly as high an initial peak as the blue distribution, that it has a clearly defined second peak, and that it is much higher as the values come out towards the higher Excitement Index values on the right side of the graph. This indicates that the matches on the bottom half of the draw were longer and more competitive. That second peak indicates that the bottom half of the draw had more long matches. And the numbers bear this out, too – there were 12 5-set matches on the bottom half of the draw and just 6 on the top half, while each half had 20 4-setters.
On the women’s side of the draw, the distributions each have one main peak. This is because the lengths of best of 3 matches are more consistent than the lengths of best of 5 matches. On the bottom half, that peak is farther to the right than the top half. This indicates that the average match on the bottom half of the draw was more competitive than the average match on the top half. That bump in the blue graph towards the right shows that a few of the most competitive matches did come from the top half of the draw, but overall, the bottom half was more competitive.
This data backs up the sentiment that the days on which the bottom half played were more entertaining than the days on which the top half played. But what could the tournament organizers do about this? Honestly, not much. There is no logical way to switch the days on which one of the draws plays in the middle of the tournament. There is, however, a way to have a day on which some of both halves of the draw are playing on the same day. They could start the tournament on the Sunday instead of the Monday, spreading the first round over 3 days. Roland Garros and the Australian Open already do this to clear the logistics of the first round days, in which over half of the main draw singles matches for the entire 2-week event are played. But it would also have matches from both halves of the draw played on the first Monday. So, while the trend would still exist over the following four rounds, there would not be a feeling of having two separate tournaments that are entirely different in quality.
Comments