ResearchProject.
A standard test to analyze contol systems:

The author joined a men's weight-loss group and decided to set a very different goal from the official one. He had just read a book on weight loss that recommended clamping The book says to simply not eat if you don't have a 'good appetite' when the rest of the family sits down. Have a cup of tea... down on his habits until " growlies" appeared.
He recognized This also resembles the way a control system with a thermostat behaves. No growlies; don't stoke the furnace. its strategy as biofeedback and wanted to see what his body Possibilities:

The body could gently lower its BMI or

drop too fast (0vershoot) and bounce back up.

would naturally do
if he changed his habits suddenly.

Measuring The gentle drop possiblity would allow a spreadsheet to measure this characteristic parameter of biological control. Time Constant

DeeJay Weight Loss Graph He plotted the graph to the right and tinkered with the Rate parameter until the green line went sideways. Spreadsheets have a "slope" function, which gave the rate measurement above (also called a Time Constant measurement). Sure enough, it turned out to be four seasons.

Looking for The overshoot possibility was also a candidate, because the majority of people who 'diet' will meet their goal and then their weight will rebound strongly.Yo-Yos

DeeJay Weight Loss Graph Secondly, he wanted to see if Overshoot in a control system usually is caused by a too-fast response of something in a closed signal path. "yo-yo" dieting In yo-yo dieting, it seems to be will-power that is both strong and short-term. was caused by "ringing" in the Step Response that one gets from suddenly producing growlies. The answer was very clear - the curve was very smooth for the first few months and it followed the Standard Decay Curve (decaying exponential) very well. Furthermore, the other men were running into their own "yo-yos" and were not doing as well overall.

The fact that using growlies is a BioFeedback method; (with no In the previous link, BioFeedback reveals something to a wired-up patient.

In our case the signal is just scale weight plus filtering.

instrumentation)
suggested that the Green Line for slope measurement could also provide feedback. Just multiply slope by the measured Time Constant, one year, and try to hold it on your goal.

Horiz Feedback. That gives a horizontal line, which is what the app feeds back to you. Responding to one number seems a lot less tricky than responding to growlies. What remained was to add a layer of strong filtering to remove To focus on jitter hides real but slow progress.

It is discouraging when daily fluctuations give and then destroy false hopes.

scale fluctuations
and Trusting the projection lets your willpower recover while you coast a while.

That reduces the tendency to diet and collapse afterward.

give a reliable
one-year projection.

His Body's Horizontal Line Became DeeJay's Target.

That sent 40 lb down the river and, over a decade later, it has not come back.
Footnotes:
  1. See Wikipedia on Set Point Theories, which says they have largely been discredited. There is good reason - too fast losses. This method avoids "set points".

  2. Mitochondria: It is not yet known how to manage mitochondria for weight loss, so our best bet is to do it gradually and let them adapt naturally.

  3. More detail: calculating projections.
    Testing with artificial data:

  4. Mayo Clinic seems to advocate a brute force approach to plateaus. They are far from alone in that - protocols seem to miss dealing with the body's characteristic time constant after "set point theories" failed to prove out.