I would also question whether investments that are unlikely (or less likely) to make a direct return are always "bad". The key points are that investment is an

I would also question whether investments that are unlikely (or less likely) to make a direct return are always "bad". The key points are that investment is an iterated game, deals are not necessarily independent and "networks" of people exist. Consider a successful serial entrepreneur; you might fund something you consider "not so great" right now because you value the relationship and want access to future deals with that person or their network. Or their uncle, whatever. Some sub-optimal deals might lead to more investments or opportunities due to "goodwill" and be net positives when the broader picture is taken into account. From this perspective it is not guaranteed that making "locally optimal" decisions at every turn would actually improve overall returns.