NYT Article - Overestimating your Calorie Counts

A New York Times article from earlier this month laughs about how people often wildly overestimate the number of calories they think they are burning when they workout.

The entire article is worth reading but here are some interesting bullet points:

• Holding onto the handle bar while doing cardio reduces your burn rate by 50% for some machines. I know that's certainly true when using a treadmill on a steep setting -- by holding on to the handle bar or grips you can make your body perpendicular to the tread, effectively making the surface flat again. D'oh!
  
• Machines drift in speed and grade. This is certainly true in my experience -- after becoming familiar with a set of elliptical trainers for example you may start subconsciously (or consciously!) choosing the one that seems "easier" than the others. I'm as guilty as the next person of this.
• Subtract the calories you would have burned sitting on your ass. For example, suppose I walk on a treadmill for 30 minutes and the counter says I burned, oh, 200 calories. Well if I'd been sitting around watching TV for 30 minutes I would have burned 40 calories. So how many more calories can I eat today? 160. Three Oreos, not four. As the Hacker's Diet says, it's better to just not eat them the in first place rather than try to burn them off!

If you are following the Hacker's Diet religiously and you are measuring every calorie that you are putting in your system, then you know exactly how many calories you are burning. Therefore you can tell how accurate your machines are because you can measure them against your spreadsheet.

But if you're like me and you have somewhat of a life, you have to estimate many of your restaurant meals and various "Goodies" that you might be treating yourself to and you can't really be sure how accurate your count is.

So I try to underestimate my calorie counts and mentally compare them to "was that workout really as hard as walking for one hour, which would have burned 300 calories?"

You know what's sad? I'm eating Doritos while I write this. But its a 100 calorie bag -- the little chips are like the size of a nickel. Not bad.

The Hacker's Diet - Forecasting

At the beginning of the Hacker's Diet, I started a tab in my excel spreadsheet called "Forecast". I plotted a line that assumed the loss of a single pound each week. At the time it seemed like kind of a crazy, optimistic idea that one might undertake with only blind optimism as one's guide.

And initially, the Forecast chart didn't seem to give me much, but it gave me something to look at while the hunger gnawed at my gut.

However, now that about 10 weeks have passed the graph looks like of cool. Check it out.


The blue line is the one-pound-a-week forecast line. Obviously its going to be a straight line, right?

The green line is the actual weight on the scale every Monday for the last 10 weeks. Taking a one-week measure kind of has the same effect of a moving average.

The orange line (in the middle) is the Hacker's Diet Trend Line (see my previous blog entry). I shouldn't be, but I'm surprised at how straight the trend line is.

So this graph allows us to compare two different types of moving averages -- the exponentially-weighted smoothing average with a period of 20 (orange) versus the crude green sampling every 7 days. The green line is more accurate but the orange line provides us with a better way to forecast. The green line appears to lead the orange line by about 10 days (or 1.5 lbs or so). Very interesting. Even better, the green line leads the blue "forecast" line by 2.5 lbs, which means I'm ahead of schedule.

We're going to Las Vegas again next month (don't try to break into my place, there will be someone there), and according to the forecast, I'll be 19 lbs lighter than when I started my diet. How cool is that? Then I might look better than this picture from earlier this year.

Yeah that's me the fatty on the left

The Hacker's Diet -- Examining the Trend line

To me, the most fascinating aspect of the Hacker's Diet is the Trend Line. The trend line is a simply a moving average, or more precisely, an exponentially-weighted smoothing average of the previous weights. The formula for its calculation is simple enough.

Let t be the trend and let t-zero be yesterday's trend.

The trend adjusts itself by taking one tenth of the difference between today's measurement and the existing trend. So if you are 2 lbs lighter on the scale then trend, the trend will adjust downward by 0.2lbs.Why divide by 10? Taking a tenth of the difference as your true signal provides a nice, smoothing effect. Let's take a look at how the smoothing effect works. Here's a contrived example: You weigh 177 at the start of the month and immediately lose two pounds so that your true weight is 175. Of course you vary a little from day to day (let's say alternating between 174 and 176). Here's what the trend line looks like:

The red line is the trend line. It appears to take the entire month to finally match your true weight of 175! Is this a good thing or a bad thing?

Its bad in the sense that for three weeks or more you are operating under the assumption that you are losing weight when you are not. But its good in the sense that you are carrying forward your credit so you can coast for a while and still appear to lose weight!

Using a larger multiplier, like say 0.3 (dividing by 3.33) gives a little bit of smoothing, but not much. As a result it becomes much more "accurate" in that it achieves you true weight in only 10 days or so. Here's what that looks like:


In the end, I must say it probably comes down to some level of personal preference. I find the 0.1 multiplier to be not just smoothing, but soothing.

The Hacker's Diet - November

I was on a ROLL during November. I was militant about the diet and I joined a new gym, this one much closer with all new equipment. I worked out 2 out of 3 days for a total of 21 days and 8,500 calories burned. Remember, a pound of fat is 3500 calories.

The weight just fell off during November. I remember waking up in the morning after having worked out the day before and skipping dinner and being beyond hunger; I should have been hungry but wasn't. At one point I felt like I had extra skin on my stomach -- maybe it was loose from losing weight so fast.
Each morning I would plug the latest weight number into my spreadsheet and watch it drag the trendline down a little bit and it would give me such satisfaction!

Who knew that seeing your weight go down by .2 lbs could be so thrilling?

I began to experience a feeling of victory-in-progress, of invincibility even. I know that sounds silly but it was working so well. Even the day after thanksgiving I woke up and weighed myself and STILL the trendline went down.

Needless to say, I am quite proud of November. I consider it no small feat to generate a deficit of almost 50% of the calories you need every single day for a month.

Lost in November: 6lbs

Total Loss Since Starting Diet: 10.7lbs

The Hacker's Diet - October

As you can see from the graph I had plateaued out at 187.5 and was kind of stuck there until I stumbled on to the Hacker's Diet. I started the diet on 10/16.
And from that day the trend started going DOWN. Yes, there was a scary part from the 25th - 28th when the scale went "backwards" for four consecutive days but the decreasing trendline kept me sane. I ended the month down a scant 3 lbs less than I started it, but the decrease was REAL.

Note the "Daily Deficit: 316 calories" at the bottom. The Hacker Diet spreadsheet computed that for me.

As you recall from my previous post, this is almost exactly the same amount of calories found in a white chocolate mocha. I was right. I am VINDICATED.

Total Lost in October: 3.3 lbs

Total Lost Since Starting Diet: 4.7lbs

The Culprit

For the last four years I had gained about a pound every month, leading to 40+ lbs over 4 years. This even though I was somewhat active; going to the gym, spending all day fly-fishing, etc. After reading The Hacker's Diet, I did some math and determined that I was consuming about 150 calories too much per day.

Well I know exactly where this surplus came from.

For years I have started each and every day with an evil concoction whose gluttony-enducing seduction is matched only by its massive amount of calories. The dreaded White Chocolate Mocha!

Yes, I have had one of those liquid bon-bons every day for the last four years. According to the Starbuck's nutritional site, a 12 ounce white chocolate mocha is 300+ calories. If I had instead been drinking just coffee and cream this whole time I would have lost 10 pounds a year instead of gaining 10 pounds a year.

I immediately gave up the white evil and as a result, if I don't do anything else, I'm going to lose all that weight -- it just might take four years. And you know what? That would be okay with me.

Discovering the Hacker Diet

Three weeks into my difficult diet, I stepped on the scale on a bad day and my weight was virtually the same as it was when I started.

I was simultaneously demoralized and re-motivated.

Demoralized in the sense that I felt that I was making no progress.

But re-motivated because here was another signal that I did not know what I was doing and that I needed some kind of help, some kind of plan if I was going to make this work.

And then, completely out of the blue, I read about the Hacker's Diet. I wasn't looking for Diet information - I was reading Jeff Atwood's daily software development blog (www.codinghorror.com) and he was talking about Geek Diets, of which the Hacker Diet is the grand-daddy.

I read the entire Hacker Diet online book over a few mornings -- I was fascinated and so motivated I was almost giddy. I was definitely receptive to the message. Here are the main points that I took away from it.

1. Small Changes over time Add up.
   
2. Reset your expectations about a successful diet.
3. Use Simple Techniques to track a real weight trend.

Small Changes over time Add up. Suppose that you eat just a tiny bit more each day than you burn off in your daily activity, say, two oreo's more than you should have (an extra 100 calories). Not a big deal for one day. But if you do this every day you will gain ten pounds a year. This is what I did for four years -- the equivalent of eating just a tiny bit more than I should have every day. In fact, I know the exact culprit, but I'll get to that in a later post.

Reset your expectations about a successful diet. Armed with a little bit of math I could see why losing more than a pound a week would be difficult. In fact, I now consider the loss of a pound a week to be just right. Two pounds a week would be like, crazy-talk! The reason for this is simple -- a pound of fat is 3500 calories. My metabolic rate is about 1800 calories a day. If I ate absolutely nothing I could lose a maximum of 3.5 pounds a week. Of course, I would probably collapse and be completely useless.

And look at it this way -- losing a pound a week is still fifty pounds a year. That's a lot. And years pass quickly.

Use Simple Techniques to track a real weight trend. This is key. Weight varies by several pounds a day due to random variations making it very difficult to get a feel for how one's diet is going. By applying the same techniques used to tease out trend data from financial market data one can get a true representation of their weight loss.

For example, check out the first two weeks of my diet. If I were just to look at the numbers on the scale I would be ecstatic in the first few days as I appeared to lose two pounds and then horrified two weeks later as I appeared to backslide.

But by looking at the trend of the loss you can see that its much more realistic representation and more importantly, its never going up, which is the all that really matters, right?

Deciding to Change

In mid-September of this year, I stepped on the scale and saw a number that I'd never seen before -- 192 lbs. Even though that day of reckoning had clearly been coming for a long time I was surprised and dismayed. I'd never weighed that much before. That same week, I saw a video of myself that I thought was downright disturbing. On the video, I looked so much larger (or should I say wider) than I felt.

I was determined to do something about it so I started dieting. I'd been successful dieting exactly once before when I went from about 180 to 145 shortly after Casey was born. And I'm not sure that time counts because I basically followed the Troy McClure diet, Smoke Yourself Thin. While that diet did work, I do not recommend it.

Anyway, since then I had had no luck losing weight -- in fact, it seemed that every year since then I had gained exactly 10 lbs and had gone from 145 to 190 in five years. And once I got past 155 or so, I wasn't particularly happy about it. This is the one thing in my life that I wasn't satisfied with. Everything else has been going so well that in some sense I was resigned to just being heavy as if it were the price I had to pay for success in everything else.

So when I saw the video, and saw that number on the scale I knew I had to do something but I wasn't particularly hopeful as to the outcome.

I dieted hard but after a month the scale was almost right back where it started.

It seemed I was doomed.