Statisticians reveal the source of your frustration.
It's no secret that diet fads and trends—low-carb, low-fat, or gluten-free, anyone?—come and go, and often don't result in significant weight loss. Now, a pair of mathematicians have devised a formula that explains why these trendy eating plans are usually ineffective. In Alogrithms to Live By, a new book by staticians Brian Christian and Tom Griffiths, the writers apply common computer algorithms to analyze different aspects of human behavior.
The part of Christian and Griffiths' text that will likely be of most interest to potential dieters focusses on a key principle in statistics that could sabotage people's ability to make healthy decisions and lead to destructive food choices. "Overfitting" is the computer science term the authors use to describe the tendancy to over-simplify an ongoing issue—such as healthful eating—into something that becomes limited or mandated. This represents a form of "over-optimizing," which tends to turn a potentially positive behavior, such as adjusting your diet, into something destructive.
"There's a tendency to over-optimize for the thing that you can measure," Christian tells Tech Insider. "There's often a gap between what you can measure and what really matters." In this case, the thing that can be measured is the kinds of food eaten—whether they be carb-free or high-protein. By fixating on eating a limited variety of food or sticking to a rigid regimen, people lose sight of the fact that the motivation behind their action should be being healthier overall.
"You see these extremely violent swings in popular taste," Christian says, using the popularity of soy and almond milk as the prime examples of this sharp change in consumer trends. "Adopting an extreme diet to lower body fat and taking steroids to build muscle... can make you a picture of good health, but only the picture," they write of a classic example of overfitting.
Christian suggests that the healthiest course is one that avoids diet fads altogether. "Just because there's one study that suggests X, Y, or Z, that can be really tempting to overfit the most recent piece of information," he says. However, as these mathematicians point out, in a few months that study will be old news and another will take its place. Instead, it's always a good idea to get a sizeable fix of fruits and vegetables—healthy foods that never go out of style.