Rational Inattention

Decision making, choices, outcomes. In my last two posts, I've covered some interesting ideas, facts and theories so far from various areas like chess, Bayesian filters, philosophy, The Way of the Samurai, Picasso and the Art of War in order to create a picture of the vast world of decision making and given this vastness I'm feeling that I've only started to scratch the surface of it.

There is no doubt that is wise to practice diligence and careful analysis before going head on into action but how much data do we need to collect and know in order to make correct decisions?

Usually, when it comes to decision making, the more data we have the better and when you need a resource in limited supply like data you need to consider the hidden costs of data acquisition. No data comes free and without an expenditure in time, energy, attention or money so it's necessary to have a good estimate of this hidden cost and sometimes it's more rational to go ahead and make decisions based on incomplete data rather than get stuck with acquiring expensive data and sunk cost. This type of rationality is called the **theory of Rational Inattention**.

In economics, the theory of rational inattention deals with the effects of the cost of information acquisition on decision making. For example, when the information required for a decision is costly to acquire, the decision makers may rationally take decisions based on incomplete information, rather than incurring the cost to get the complete information. - Wikipedia

Unfortunately, Wikipedia doesn't offer more information than this about Rational Inattention. I've done a search and came upon an article by Christopher A. Sims, the author of this theory and the article is laden with equations, information theory, Gaussian-Linear-Quadratic equations and so on, in other words, unless your advanced math is up to date, the article is quite cryptic. Probably the guy that edited the Wikipedia article gave up too trying to understand the finer details of the theory and wrote just the stuff that I quoted. I don't think that there are many people that make decisions after getting a pen and paper and solving Gaussian-Linear-Quadratic equations. Speaking of data acquisitions costs, trying to understand all of the mathematical models of the theory and writing about it would require a high cost in time, energy and attention, a cost that it wouldn't be rational for me to incur given the small viewership that my blog has. Lesson in that.