Is The Olympic Hangover Real?

With the world shutting down, we’re reaching into our archives and pulling some of our favorite stories from the SwimSwam print edition to share online. If you’d like to read more of this kind of story, you can subscribe to get a print (and digital) version of SwimSwam Magazine here. This story was originally published in the 2017 Summer Preview edition of SwimSwam Magazine.

One of the main narratives entering any post-Olympic world championships is a sportwide reset. Many of the old stars have retired (Michael Phelps), and new athletes have to step up to take their place at the top.

Retirements combined with some elite swimmers’ training breaks appear to have a slight negative effect on times the year after an Olympics relative to what we would expect given the overall progression of the sport.

To show this effect, I grabbed the results of every world championship and Olympics from 1996 to 2016 and compared one year with the next, excluding 2008, 2009, and 2011 because of the super suits.

Individual-event winners in post-Olympic world championships were 0.2 percent slower on average than event winners at the Olympics the year before for the men and women (1998, 2001, 2005, and 2013 world championships).

Olympic years were noticeably better than the world championship years before them, with men’s event winners dropping an average of 0.6 percent from the event winner before them, and women’s event winners dropping an average of 0.8 percent* (2000, 2004, 2012, and 2016 Olympics).

This trend held across the spectrum of places. At worlds after an Olympics third place, men and women were on average 0.2 percent and 0.3 percent slower than third place at the Olympics, and third-place men and women at the Olympics were 0.7 percent faster than the previous worlds.

The eighth-place time to make finals was 0.3 percent slower at worlds than a previous Olympics for men and women, and 0.7 percent faster at Olympics versus a previous world championship for men and women.

The difference was even bigger further down the rankings. Sixteenth place in prelims was 0.4 percent and 0.6 percent slower for men and women, respectively, at worlds versus a previous Olympics, and 0.7 percent and 0.9 percent faster for men and women, respectively, at an Olympics versus the previous worlds.

Despite a possible post-Olympics drop-off, the 2017 world championships are still likely to be the second-fastest world championships of all time behind Rome 2009. Times at worlds and Olympic meets have steadily improved over the last 20 years. Since 1996, men’s winning times have improved about 0.18 percent per year, and women’s winning times have improved about 0.20 percent per year. (A similar trend exists for the lower places as well.)

Using 1996 as a baseline**, women’s winning times at the 2016 Olympics were on average faster than the those at the 2009 Rome world championships at the peak of the super-suit era. 2009 women’s winning times were an average of 3.90 percent faster than 1996 Olympic times, and 2016 Olympic times were 3.94 percent faster.

The lower places were usually slightly faster in 2009. Second place in 2009 was 4 percent faster than in 1996, and 3.8 percent faster in 2016 than in 1996; third place was 4 percent faster in 2009 and 3.8 percent faster in 2016; and eighth place, to make finals, was 4.1 percent faster in 2009 and 4.2 percent faster in 2016.

But elite women’s swimming appears to have nearly caught up to the times of the suit era. The men are still a bit behind 2009. Winning times in 2009 were 4 percent faster than those in 1996 on average, and winning times in 2016 were 3.5 percent faster.

The gaps between 2009 and 2016 times were consistently 0.5 to 0.6 percent in all high places. This is much closer than a few years ago. In 2011, men’s swimmers were 1 to 1.2 percent slower than in 2009.

The women’s improvement in first-place times since 2009 isn’t explained away by Katie Ledecky boosting the 2016 numbers. Dropping the 400 and 800 free lowers the 2016 average change versus 1996 to only 3.85 percent — still in line with the best of the suit era.

The biggest drops in Olympic events since 1996 weren’t in Ledecky’s events. The largest drop by percentage was in the women’s 100 fly, which went from a winning time in 1996 of 59.13 to a winning time in 2016 of 55.48, a change of 6.2 percent. The 800 free was fourth (8:27.89 to 8:04.79, 4.5 percent faster) behind the 200 IM (2:13.93 to 2:06.58, 5.5 percent), and 400 IM (4:39.18 to 4:26.13, 4.7 percent).

The biggest time change in a men’s event was the 100 breast, which went from a winning time of 1:00.65 in 1996 to 57.13 in 2016, a difference of 5.8 percent.

The most similar winning times between 2016 and 1996 were still significantly faster than the winning times in 1996. The closest was the women’s 200 back, which was won in 2:07.83 in 1996 and 2:05.99 in 2016, a difference of 1.4 percent. The closest men’s event was the 100 free. The winning time in 1996 was 48.74, and in 2016 it was 47.58, a difference of 2.4 percent.

As a sport matures, we would expect to see diminishing levels of improvement of the best swimmers. As athletes approach the upper limits of human performance, the overall improvement of the sport should slow. Elite swimming does not appear to have hit that point of diminishing returns yet.

As seen in the plot, improvements at the top of the sport have been steady for the last 20 years, except for the outliers of the super-suit years, and have shown no signs of a major slowdown.

This improvement will eventually flatten out. It seems unlikely that a human is capable of breaking 40 seconds in the long course 100 free. However, until the slowdown comes, we should be in for lots of meets with big times and records. Enjoy it while it lasts.

The trend lines do not include 2008 and 2009 because of the super suits.

*The differences in mean are statistically significant, with p<.001.

**This doesn’t account for 1996 not being a neutral baseline. Some winning times were weak, and others were strong. In the aggregate stats, this should average out, but view close-margin single-event stats with some skepticism.