[calories eaten minus calories burnt] times 3500 = change in lbs of weightWhich is fine, and a useful tautology to start with, but people interpret this the wrong way.
Time magazine recently came down on the side that reducing calories eaten is all that matters, though others have argued that increasing calories burnt through exercise is all that matters.
The problem is that all of these things are missing a third factor, metabolism. which we don't directly control, but is probably far more important than the other two.
The typical American gains about one pound a year. That corresponds to 3500 calories.
That means over the course of a year where we typically eat around 1 million calories, the difference between calories eaten and calories burnt is only 3500, or or about 0.3%
I do believe that we probably have a fair amount of control over how many calories we eat (not complete, but fair). However, I find it hard to believe that somehow our conscious self manages our activity level to be within 100.3% of that calorie level. Clearly automatic mechanisms in the body are kicking in that regulates how fast we are burning calories.
Thus, as the Time magazine article points out, it is not at all clear that exercise increases calories burnt. It does while you are exercising, but if it causes those automatic mechanisms for metabolism to slow down for the rest of the day, then exercising would achieve nothing at all.
So that's the secret. Find a way to adjust that automatic mechanism and the path to weight loss is clear. Just don't ask me how to do that.