# Better Decisions at the Supermarket

(I just went to a supermarket and wanted to share some thoughts. The decision theory stuff will be elementary. Here is a LW FAQ on decision theory if you actually want to learn the basics.)

Today I had a class on repeated games, and went shopping by myself.

Usually when I grocery shop with someone, there is a high probability we will end up having an argument.

Behind the debate on what kind of yogurt to buy, there is a real issue: misaligned values.

I believe understanding what we each value can help to better communicate on what matters. More importantly, it might bring better decisions.

I used to spend hours in supermarkets, thinking about how to optimize my cart because I did not have an accurate model of my own utility function.

Today, I ended up listing the factors which mattered the most to me:

• Money: I do not have an infinite budget for food. I am concerned about overspending on foods what will become stomach mush.

• Time: Time shopping and time spent cooking.

• Health: Food quality might impact my digestion, short-term mood but also long-term health. High quality food might prevent diseases and improve my longevity.

• Biological Needs: Our body evolutionarily needs a certain quantity of protein, lipid, etc. Therefore, if I already have carbohydrates at home, “buying protein” is more important than “buying carbohydrates”.

• Quantity: I want to buy the maximum amount of food I can so that I don’t need to come back, yet I can’t carry too much weight

Some useful ideas about the above criteria:

• The law of diminishing returns explains how, all things being equal, optimizing one parameter will less and less impact the utility function.

• The time criterion is highly correlated with every other criterion. Indeed, if I try to optimize my protein intake, I will spend some time thinking about it.

• People tend to value low-risk instant utility, rather than some reward which may arrive later. It is possible to introduce a discount to take into account that I prefer \$99 now than \$100 tomorrow (depends on the inflation rate, etc.)

Take time into account in your decisions. A simple heuristic can go a long way:“if I save less than a dollar, I should spend less than 3 minutes on optimisation (Assuming a wage of \$20/​hr for the value of your time)”.

Second, you might want to consider the whole process as a coalition formation inside your cart.

Third, write out your factors. You know what matters to you. Knowing how it matters is called instrumental rationality.

This is the 10th post of a series of daily LessWrong posts I started on April 28th.

Next post: Applied Coalition Formation

Previous post: Beliefs: A Structural Change

EDIT: incorporated Elo’s suggestions and changed underline to bold