In software development we are inundated every year with advice on best practices, new technologies, principles and techniques. Picking and choosing which ones to follow can be a tough call. Being able to step back and determine how quantified a best practice is before becoming dogmatic once engaging in one, can be an even tougher call.
In our industry it pays to be on the front foot with new ideas and practices. When REST gained popularity the internet became abuzz with advice on how to best implement and design a REST based architecture. Today, as Microservices gain popularity the industry is abuzz again on how to best design and deliver a system based on this architectural model.
With the pace that the software industry moves at, new practices and principles are often drawn from observations and experiences. Casual experiences, anecdotal evidence and observations that compliment well formed and understood principles are how most new practices and principles are proven to be beneficial.
For example most of the advice and principles today on Microservices are not quantified by unbiased, controlled studies, they are typically quantified by experience on a team that tried to implement it (casual experience), information cited by other individuals who either observed teams or spoke to teams doing it (anecdotal evidence). These items are subsequently mapped to an established best practice or principle, for example it is often stated that Microservices embodies the Unix philosophy of having small composable utilities that you can pipe together to perform powerful operations. The Unix philosophy is a well formed and understand principle, the benefits of Microservices are being quantified by drawing parallels with this similar practice.
There is nothing wrong with this, it is perfectly reasonable to conclude that if a practice is similar to another best practice, than it too is most likely a best practice. Anecdotal evidence and casual experiences are also the only forms of evidence a new innovation has. The most important take away is to not be dogmatic when following any principle, always carry a bag of salt and use your best judgment (which coincidentally is predominately based on unquantified evidence).