Weight simulation assuming normally-distributed weights around the slope.

That chart is a simulation: I made a line that had the same rough slope as my weight loss, and then made synthetic data points around that line. The points are normally distributed -- the famous bell curve beloved of statisticians -- and the variance is around 5 pounds. Notice how similar they look?

And that's the point. I am losing weight -- and yes, my glucose is also dropping still, and the tape measure shows I'm losing body fat. Day to day, though, it's hard to see -- perfectly normal randomness is hiding it. It's hard to see.

But then, I've only got 147 frames of this animation -- a little over 6 seconds of film. It might be a little early to be discouraged.