Originally Posted by
fozzy
I wasn't doubting the validity of the research or their conclusions - more the reporting of them in the BBC article. From that article, I got the sense that they were advocating that you shouldn't drink anything and you will therefore run much faster, which is obviously not right.
Regarding my comment on the stats, if you look at Figure 1 in the paper, you see a lovely bell-shaped normal distribution. This is the same distribution you would see if you look at, for example, a statistical analysis for all the heights of the members on this forum, and possibly could be expected.
Figure 2 is the figure where I have the slight issue: the simple mantra of drink less, lose more body weight, run faster as advocated to a certain extent in the paper and strongly in the BBC article is not necessarily true. In fact if you look at the fastest runners in Figure 2, the top 5 were all in the range 0-5% loss, although one of these
could actually have gained weight. I'm also not 100% convinced on the validity of the fit in this figure. I could quite happily plot a straight line with negative correlation to that data and argue that it was valid, simply because of the clustering in the 0-5% BW loss. I would need to see more of the statistical methods they've used to come up with
the correlation to be completely convinced.