> Then, in a sense, I've lost 0.75 in an attempt to save 0.25. You have lost more than that, because it is not a one-off decision about a single drink - you ar

> Then, in a sense, I've lost 0.75 in an attempt to save 0.25. You have lost more than that, because it is not a one-off decision about a single drink - you are buying information for all your future choices. Suppose you drink 1 coke a day. The difference is $0.25/day or $91 a year. The gain from switching does not stop after a year, it goes on indefinitely, so at a fairly psychologically normal discount rate of 5%, the NPV of the gain is $1871. In order for your experiment to not cover $0.75 and not be profitable, you would have to assign a prior probability of ~0.013% to the generic being as good (or better!) and you switching and reaping a gain of $1871. Which would be crazy because you know that often the generic is fine and frequently is literally the same manufacturer as the brand name, either because they use the same contractors or because it's just price discrimination. Personally, I make a point of, whenever trying something new like food, to buy 1 of everything and compare them. I am no longer surprised when I find that the generic is as good or better at 1/3rd or less the cost, or that I prefer something I didn't expect to prefer. (Particularly in tea this has paid off in learning that I liked things I didn't remotely expect to like, like 'twig tea'.) It doesn't even take blinded experiments to show this, just trying stuff side by side will often reveal this. I think it's crazy how people will buy the same thing forever and overspend on brand names and buy small quantities. (And then they complain their monthly grocery bill is $400 and they wonder where all the money goes...)