As a follow up to our previous post regarding the expected results of the 2012 Olympic swim competition, this post will present 1) the results of the prediction and 2) speculation as to why the results were what they were.
Prior to 2008, the prediction model successfully predicted 87% of the mean top-eight performances for all Olympic events between 1988 and 2004. In 2008, however, only 34% of the Olympic swimming performances were accurately predicted using the same model. We concluded that the Games were significantly biased due to the introduction of “new-tech” body suits. Given FINA’s 2010 suit policy limiting skin coverage and material type (presumably the cause of the bias), we hypothesized that the model would successfully predict the 2012 Olympic Games, as it had done prior to the suit bias of 2008. However, similar to the 2008 Olympic Games, the results of the 2012 Olympic swimming competition were much faster than expected.
For the 2012 Olympic swim competition, 31% (4/13) of the men’s events and 38% (5/13) of the women’s events were successfully predicted. That is, the model correctly predicted 35% (9/26) of the mean top-eight performances for all events. If we compare 2012 to 2008, about half of the men’s (6/13) and nearly all of the women’s (12/13) events were, on average, faster than the 2008 Games. Tables 1 and 2 show the results of the 2012 Olympic swim predictions.
Note: Shows a comparison of 2012 Olympic predictions to the 2012 and 2008 actual Olympic performances in each event. 8/13 women’s events were significantly faster than predicted. All events in 2012 but the 50 freestyle were faster than in 2008.
Note: Shows a comparison of 2012 Men’s Olympic predictions to the 2012 and 2008 actual Olympic performances in each event. 9/13 men’s events were significantly faster than predicted. Highlighted events in 2012 but the 50 freestyle were faster than in 2008.
The past two Olympics, 2008 and 2012, have acted to shift the performance curve toward faster rates of improvement. However, the improvement in performance seen in 2008 was much larger than in 2012, suggesting a slowing in performance progression (Figure 1, represented as distances A vs. B). This fact suggests that performances are returning to the expected performance curve (2012 is closer to the curve), or that the performance curve has shifted (Figure 1, represented as red line), and competitors are again approaching the limits to swim performance.