[mike_organon]

Taubes is referring to the 20 grams above maintenance of current weight (difference between red and green lines in your graph), not the grams above what ideal weight would require. The 7 strips of bacon are needed to maintain the excess weight, not to put on more (according to conventional wisdom). His point is that weight control can't be explained by counting calories so precisely for decades - there must be something else going on. He's not being deceptive.

[yummyfajitas]

The only reason it's precisely 20 cals above maintenance for 20 years is because he asserted it to be this.

The mainstream theory predicts similar weight gain with 20 cals above maintenance for 10 years, or a random walk with mu=20 cals and sigma=250. It also doesn't predict that a 50 year old should be precisely 40 lbs overweight - the people who only hit a mean of 15 cals above maintenance were only 30 lbs overweight and are simply excluded from his hypothetical by the premise.

The precision you seem to be describing doesn't exist in the real world and no theory needs to explain it. It's purely an artifact of his assumptions.

[mike_organon]

Wait, I thought your point was that the 20 cals (10, whatever, a small number) was a straw man and it's 7 strips of bacon that causes overweight.

My point is you're misinterpreting Taubes, because he is saying 1) the mainstream believes such as small number of calories matters, and 2) that is absurd.

Do you agree with #1 now?

When I said precision, I don't mean 20 cals leads to 40 pounds. I mean the precision of someone eating an average calorie count with a margin of error so small. Consider the other side of that - for me to have maintained my own weight for the last 20 years, I would have to have been extremely precise in my eating, which I'm sure I wasn't.

[yummyfajitas]

My point is you're misinterpreting Taubes, because he is saying 1) the mainstream believes such as small number of calories matters, and 2) that is absurd.

The mainstream says that 7 strips of bacon will make you 37.5 lbs overweight and tacking on an extra half strip will add 2.5lbs more. In order to make this seem absurd, Taubes never mentions the first 7 strips of bacon.

Whether or not I agree with (1) depends on how low we set the threshold for what matters. If we care about the difference between "six pack" and "thin but no six pack", we care about the first half strip of bacon. (Fun fact: bodybuilders and boxers do stress about 20 cals/day.) For the purposes of Taubes example (thin 50 year old vs obese 50 year old), mainstream nutrition and calorie counting says 20 extra calories doesn't matter.

When I said precision, I don't mean 20 cals leads to 40 pounds. I mean the precision of someone eating an average calorie count with a margin of error so small. Consider the other side of that - for me to have maintained my own weight for the last 20 years, I would have to have been extremely precise in my eating, which I'm sure I wasn't.

It's exceedingly unlikely that you maintained your weight with a margin of error of 2lbs - my daily variation can be as large as 6-7lbs. I'm also fairly "stable" in my weight - it never leaves [215,230] (unfortunately I'm a hard gainer).

Further, if you want to stay roughly stable in weight, the target is not your maintenance needs. If you instead target the maintenance needs of a person of ideal weight (rather than a person of your weight), you don't need much precision at all to avoid becoming overweight. If you are off by 100 cals, you might be 5lbs too heavy, but you won't get anywhere near 40lbs - your maintenance needs will increase and consume the excess.