I’m not going to eat between meals and I’m not going to take seconds. Just today. And I’m going to try this for a year to see what happens.

I took my walk early this morning (2 miles) and while I was walking I thought about The Hacker’s Diet. And I think I understand the weighted moving averages thing (which is the author’s way of tracking weight more accurately than than what we happen to weigh on any given day)

Weighted Moving Averages for a Dummy (me)

He points out that when you weigh yourself your weight could be wildly up and down depending on how much liquid we’ve drunk or the bulk of how much we ate and peed or pooped recently. And those variations make it difficult to spot actual weight gain or loss.

So (he says) we can get a more accurate view if we plot a 14-day average of our daily weight. He says that taking an average for that period would smooth out the wild fluctuations of bulk and leave us with an estimate of what’s actually in The Rubber Bag.

But the days that are about to fall away from the average (days 12, 13, 14) shouldn’t be given as much value in the equation as the most recent days. In fact as the days pass they become less important. So he’s developed a formula for “weighting” the value of each day before calculating the average for the 14 days. And he gives the most recent day full value & the farthest days just a percentage of their full values.

(The author of The Hacker’s Diet, John Walker, has generously given me permission to discuss his diet in detail using some quotes, with links back to his site. (Earlier E4T articles on The Hacker’s Diet) )

The uncanny way a moving average ferrets the trend from a mass of confusing measurements can be seen by plotting the 10 day moving average along with the original daily weights, shown as small diamonds.

The moving averages we’ve used so far give equal significance to all the days in the average. This needn’t be so. If you think about it, it doesn’t make much sense, especially if you’re interested in using a longer-term moving average to smooth out random bumps in the trend. Assume you’re using a 20 day moving average. Why should your weight almost three weeks ago be considered equally relevant to the current trend as your weight this morning?(snip)Replacing the simple moving average with an exponentially smoothed one with a smoothing constant of 0.9 (roughly equivalent to a 20 day simple moving average in terms of lagging the trend), we get the following plot showing the trend line produced by the moving average versus the original daily weights from which it was calculated.

At last we’ve found a tool that accurately extracts the trend of actual body weight, the direct consequence of the balance between what goes in to the rubber bag and what gets burned, from all the distraction and confusion based on whatever’s in the bag at the moment you step on the scale. Now we’ll learn how to interpret this trend line.

You can download his tools including a spreadsheet (smooth.xls) — if you haven’t retrieved them yet this would be a good day to do it. We’re going to be relying more on his tools as we discuss the upcoming sections.

Share This!!