once a fat guy wrote:
Right, it is only potentially linear within narrow bounds, which makes it nonsense as applied by the poster who is trying to "call BS" on my original post.
I think his point was valid that extrapolating down using the 2 second rule shows that your time, although not exceptional for a light guy, is very fast for a fat guy and therefore unlikely. He probably took it too far by assuming a low normal weight.
Your El G example shows that his mile time is really exceptional for his weight, which we would all agree on.
btw, here is a plot of the clyde results vs weight for the 2014 MCM marathon. A linear regression, although there's obviously a lot of spread, gives a minute per lb for the full marathon or 2.3 seconds per mile per lb, pretty close to the 2 second rule of thumb.
http://i.imgur.com/fIyWcXm.png