Another problem with it is more subtle.
A real danger in developing models based on past data (input data) for the purpose of providing useful predictions is that because the model is built to fit the input data (if it does not fit it, it will be tweaked till it does fit it) models have a way of fitting input data really, really well.
But this does not necessarily indicate that they will fit other data well. Just the data they were built on, perhaps.
So yes, the article gives “impressive” statistics on how well the model’s “predictions” fit some famous natural bodybuilders. But this shows nothing of the predictive ability: it shows only tha the model fit the data it was created to fit.
Now, I am not natural anymore, but someone with better genetics, someone more of a natural mesomorph, most certainly can match or exceed naturally what I can achieve with the juioe. So I thought it fair enough to see what the calculator would do with my figures.
It pretty much nailed the predicted weight – only 2 lb off my actual weight at the bodyfat percentage given. So far so good. (And note: contrary to at least one above post, the calculated weight is NOT intended to be LBM, but total weight at the provided percent bodyfat.)
However, the calculated measurements were not at all in agreement, IMO, to the calculated or actual weight… Most of them were much greater girths than actual in my case – and of course it does make complete sense that a gifted natural bb’er of my height, wrist, and ankle measurements could have much greater girths in those regards – but the greater girths would have to give greater weight. Yet somehow it does not, in the calculated figures.
It’s not that I’m disproportionate and there is vast weight elsewhere making up for being in fact considerably lighter in the arms, chest, and calves.
The girths given would have to yield a considerably higher bodyweight at the bodyfat percentage given than what the calculator actually says, or at least most surely would in my case. I would have to add a great deal of muscle to have those measurements, yet I am a touch over the weight that supposedly is associated with those figures at my percent bodyfat.
So the thing is not, IMO, even consistent within itself.
