Well, people shared your intuition for a while but then new results came up. This literature started with the supposed rationality of compelling axioms of decis

Well, people shared your intuition for a while but then new results came up. This literature started with the supposed rationality of compelling axioms of decision making by Ramsey, von Neumann, Savage, etc. These are in the end based on measurement theory. People noticed early that humans seem to remain rational even if they violate intuitively acceptable rationality postulate. Take Luce's coffee cup example as an illustration. You prefer black coffee to sweet coffee. Suppose you compare coffee with no sugar to coffee with one grain of sugar added. You're indifferent: a~b. Then add another grain, and so on. You will get comparisons a~b, b~c, c~d, e~f, ..., j~k, and then suddenly a>k, a violation of the supposed transitivity of equipreference (aka indifference, equally good). But that seems to be rational. So people relaxed rationality requirements and now there is the problem what 'rationality' actually means. Fast forward a few years and empirical studies found the following strange behavior: If you mention a high number before asking people for some fictional charity contribution, then people tend to be be willing to pay more than if you mention a low number before, and it does not matter in which way you mention the numbers. (Actual experiments were made by making people roll a rigged lottery wheel before doing some completely different task, for example.) You can even tell participants about the observed effect before, it will still be observed. I see no way how this "anchoring effect" could be described as being rational. But many people nowadays share your opinion, and there is a whole field called "ecological rationality" in which scholars try to re-interpret supposedly irrational biases as good and rational heuristics increasing e.g. evolutionary fitness. I don't think they're right in general, though. Some of the biases are just flaws. If I flash a number before your eyes and this affects your subsequent decision making, then that's not a useful heuristics, it's a flaw in your brain processing. My 2 cents, others disagree with me.