Here’s a screenshot of my Calorie, Body Fat, Exercise and Weight tracker:

Screenshot of my Calorie, Body Fat, Exercise and Weight Tracker

Among the recent changes, I switched to the visualization suggestion from The Hacker’s Diet. A running average with an exponential weighting factor of 0.9 is compared to the raw values. It not only lags the inputs, any single measurement under its current (lagging) value will drive it still lower, while a measurement above it will drive it higher. In other words, trend reversals are easy to see with the naked eye, while meaningless fluctuations are de-emphasized.

Weight Chart

This makes it very easy to see that you’re still on the right track, despite the ups and downs of your daily scale readings.

Body Fat Chart

Very motivating, no?

Body Water Chart

The dashed line is the model’s prediction based on food and exercise logs, subjected to the same moving average.

Calorie Deficit from Food and Exercise Log vs. Weight Loss, in Calories

Speaking of the model, you can see above that the agreement between the projected weight loss and the logged calorie deficit remains very good. Aside from the actual weight loss itself, I must say this is the most satisfying and motivating aspect of the whole exercise.

This is a plot of the Body Mass Index vs. the Body Fat percentage. It includes the actual measurements over the 11 weeks of my diet so far (the little red crosses), as well as some guesses about where this is going next.

Body Mass Index vs. Body Fat

Obviously, I started in the upper right-hand corner (with a BMI around 32.5 and a Body Fat percentage of over 32%!) and as the diet progresses, I’m moving down and left. You can see the measurements just hit the transition between Obese and Over-Weight as per BMI.

The most important parameter impacting the shape of the curve is the Fat Loss Ratio (FLR), i.e. the percentage of weight loss that is actual fat loss (as opposed to muscle loss).

The thick brown line is the model’s best guess as to where BMI and Body Fat are going next, based on my historical average FLR value (i.e. 63.06% as of today), while the dashed lines show where the curve would lie for other values of FLR.

For the record, I expect FLR to change as I continue to lose weight and/or make an impact on my muscle loss, but it’s interesting to see how wide a range of possible Body Fat ratios could potentially correspond to the various BMI threshold values.

Lastly, these BMI and Body Fat values are valid for me, right now. I doubt they have any predictive value for anyone else, especially if they are not a 35-year old male.

Graph of the difference between the theoretical and the actual values for total body water percentage. In other words, anything positive on this chart is water retention, and anything negative is dehydration.

Water Retention Graph

Note the biweekly cycle, corresponding exactly to my business travel schedule over this period… Correlation does not imply causation, I know, but there’s definitely something there!

Two months into the calorie-counting experiment, and not only is the weight coming off, but the agreement between the theoretical weight predicted by the calorie logging and the scale’s verdict remains very good.

Interim Report on Calorie Counting Model

Motivation is strong, and a Body Mass Index of 30 is just around the bend, now. I will no longer be officially obese, just overweight.

Five weeks of comparison between the calorie deficit as calculated from food intake and exercise, and the weight loss as weighed.

Calorie Deficit vs. Fat Loss

Weight loss is converted to calories of fat (3,500 per pound) for comparison purposes and both data series are smoothed with a 9-day moving average function to remove noise.

The agreement is uncanny, no?