tag:blogger.com,1999:blog-1564958057737541664.post32780872538978220..comments2014-12-01T09:33:59.475-08:00Comments on On Bicycles, and.... what else is there?: Raceweight: looking at rider mass data leading to San Bruno Hillclimbdjconnelhttp://www.blogger.com/profile/01484858820878605035noreply@blogger.comBlogger4125tag:blogger.com,1999:blog-1564958057737541664.post-19077855514082629652013-01-03T13:35:21.918-08:002013-01-03T13:35:21.918-08:00That's interesting. I have a body fat scale a...That's interesting. I have a body fat scale and at some point it started reporting my body fat as 23%, which I know is wrong. But I validated the weight using a giant bucket of kitty litter, which I independently measured with my bike scale.<br /><br />BTW, I didn't record R-squared. That is supposed to report the probability random chance could yield a similar result. I take Flannery and Press's "Chi By Eye" approach :). It looks like a good fit to my eye....djconnelhttp://www.blogger.com/profile/01484858820878605035noreply@blogger.comtag:blogger.com,1999:blog-1564958057737541664.post-32932410506939566922013-01-03T12:24:56.734-08:002013-01-03T12:24:56.734-08:00Did you calculate any r-squared values for this da...Did you calculate any r-squared values for this data?<br /><br />Interesting to see how meaningful the trends are, despite identical weight values happening 2 weeks apart. I've been tracking weight with a body fat and water scale, and it seems to be spot on if the water percentage readings are the same. However, I've seen overall weight differences on the order you're seeing here, hydration is definitely a factor.Fred Moyerhttp://www.blogger.com/profile/03821247646980679909noreply@blogger.comtag:blogger.com,1999:blog-1564958057737541664.post-52495246372909255022013-01-03T06:27:47.387-08:002013-01-03T06:27:47.387-08:00I assume you mean as a fraction of total body mass...I assume you mean as a fraction of total body mass. Yes, I could have done that, although the body mass is changing point-to-point so I could have normalized by an average or terminal mass. But their mass was quite close, indeed at one point C was lighter than B, so I didn't think given the precision involved this would have been significant.<br /><br />I was slightly surprised the RMS residuals were so close, since you might think one or the other would have taken more care in controlling variability, for example by eating a more consistent diet or practiced a more consistent hydration or been more consistent in weighing themselves before or after other morning activities. But the similarity, and the similarity of B's scatter before and after starting the calorie restriction, suggests to me this is a fairly typical result: day-to-day differences typically differ by around 1%, even if structural mass is unchanged.djconnelhttp://www.blogger.com/profile/01484858820878605035noreply@blogger.comtag:blogger.com,1999:blog-1564958057737541664.post-10094875058660918782013-01-02T22:26:27.526-08:002013-01-02T22:26:27.526-08:00Very interesting post, but I'm curious as to w...Very interesting post, but I'm curious as to why you plotted the residual graph in terms of kg as opposed to standard deviation? The riders are fairly close in weight, but wouldn't the standard deviation plot be a better comparison between the two in that second graph?Vitaly Gashparhttp://www.blogger.com/profile/08841211447943927157noreply@blogger.com