As it turns out, human behavior is more complicated than this.
''[W]e do everything (...) because it brings us the most happiness out of all the options'' is the basis of a century long discussion that led to the development of contemporary microeconomics. The barest, least objectionable version of your intuition is the consumer model. Roughly speaking, you can imagine that individuals have a list of options, all of which are ranked in decreasing order of desirability. Everyone faces limitations and cannot pick just any option they want, so you can picture this as crossing off some of those options from the list. The choice you observe, in this light, is the highest available option on that ordered list.
This model captures the essence of your intuition, but it avoids the need to quantify pleasures and pains. It merely requires ranking whereas a more literal application of your intuition would not only require that we be able to rank options, but also that we compare magnitudes of pleasures and pains (at least across options, but it might also imply a need to do it across people as well).
Now, why do I say that this is not a completely accurate picture of human behavior? Well, the ordering minimally requires completeness (all pairs of options must be compared) and transitivity (your ordering must be internally consistent). You can always conceptually work around the first requirement by specifying a suitably well defined set of options that makes any problem of comparison irrelevant. However, the second requirement is more annoying. Human choices in laboratory experiments fail to be transitive in many circumstances. For instance, you would prefer A to B, B to C, yet prefer C to A. Hence, either transitivity is violated or you observed that A, B and C are equally good options which would make the model completely useless.
As you can imagine, the problems do not stop there: when you add assumptions, you can start to talk about demand functions, supply functions, attitudes toward risk, attitudes toward uncertainty, etc. and there are problems in some of those cases too. The bottom line is that it's a powerful idea and it has some very compelling applications, but it's not perfect.
Economics PhD student