A New Distance Metric: Evanswich Mean Distance

In keeping with the statistical tradition of the Morton Salt Report, and the author's goal of promoting anything statistical, including fledgling statistician Bill Ireland's analytical ideas, an analysis has been conducted. In this analysis, we (the royal we of course) compared swimming times for the 27 crazy swimmers who participated in both the Santa Monica Breakwater Two Mile Swim on 6/23/01 and the Santa Monica One Mile Swim on 8/12/01. Turns out they didn't need to swim both as the first predicted the second with alarming accuracy (slope = 1.2, intercept = -8.1 (min)), indicating the races must have been close in actual distance. That or all SCAQers are really slowing down as the summer wears on. However, upon recollection that both races were set according to Evanswich Mean Distance, which is related to the ESP expounded upon last summer, we were not that surprised.

The attached graph shows the swimming times in minutes for the first swim on the X (horizontal) axis and the times for the second swim on the Y (vertical) axis. The blue dots show our crazy swimmers. A 45 degree line indicating that part of the plotting space for which the two times would be equal is shown as a solid line in red. The least-squares regression line, corresponding to a correlation of 0.92, is shown as a dotted line in blue. Most of our swimmers are slightly below the 45 degree line, indicating that the first two mile swim was indeed slightly longer.

In keeping with last year's Catalina Challenge, our two favorite (well, my favorite anyway) swimmers are identified. As usual, Morton is in green for "Go Morton!" and Kostich is in red for ... well, as usual, you can figure it out. Morton is not quite keeping up with Kostich but the summer isn't over yet so my money is still on her. In keeping with her statistical prowess, she is right on the regression line. Kostich isn't. Enough said.

Now I must return to the work they actually pay me to do here at the old thinktank.