Reason for writing: I didn't see a clear, unambiguous explanation of this concept when I did a search on LW.

Epistemic status: Confident, but this is a normative claim, not a descriptive one. Skip if you hate normies.

Eat plenty of vegetables. Drink plenty of water. Get enough protein.

For the average person in a developed nation, mixing these 3 heuristics into their current nutritional profile is likely to improve their health. I don't know if anyone has actually penned a distance function $p : Nutrition \to R$ that lets us rank how good or bad someone's nutrition *is*, but I'm pretty sure that if you aren't following these 3 heuristics, and then you start, $p ({you}_{2})$ will be greater than $p ({you}_{1})$ , even if you're inconsistent.

Past general heuristics, we run into a problem when we try to compare specific foods to find the best one in a given class. For example: is rotisserie chicken or canned tuna better for your meal plan? That's a question of moderate difficulty to answer; we would need information such as

The nutritional profiles of the chicken and the tuna,
Whether you prefer the taste of one or the other, and
How much you're willing to spend and the relative prices of both in your area. (Fun fact: A 1200 calorie $7 rotisserie chicken from a supermarket is actually less expensive, calorie-per-calorie, than a 120 calorie $1 can of tuna. Otherwise it would be >$10.)

But there are dozens of common kinds of common protein sources out there, and they all can all be prepared in so many different ways. When you factor in meal prep and individual variation and all the different kinds of recipes you can make, finding the "best" food for any given metric honestly starts to look intimidating.

The answer? Don't look for the best; look for good enough. Nutrition is satisficing. Take some time to learn some general heuristics about nutrition, don't be afraid to explore randomly a bit at the supermarket, but also don't stress out about your own nutrition once you've got a few solid principles in place.