The purpose of using a constant is to eliminate the use of magic numbers, also known as unnamed numerical constants. The problems with using a magic number in your code are manifold:

  1. Future developers reading the code (or yourself after a few weeks) will have difficulty identifying what that number represents and why you chose it. Introducing a constant gives you a chance to name the concept that number represents and also add a comment to explain your reasoning if necessary.
  2. Magic numbers increase the opportunity for subtle errors and inconsistencies. For example, consider these two Ruby methods:

      def circumference(r)
        2 * 3.1415 * r
      end
    
      def area(r)
        3.1416 * r * r
      end
    

    In the above code, the mathematical constant pi is used twice, but is rounded differently each time.

  3. When the number is used in multiple places, it is difficult to change all instances of it correctly. While a simple find and replace can sometimes be sufficient, this is not possible when a single number represents multiple concepts. Digging through multiple files to verify if each instance of a number should be changed or not is not a scalable coding practice.

Despite the name “magic number,” the problems described above are not limited only to numeric types. These problems affect any hardcoded value in a codebase.

The above problems may seem minor at first, but each of these mistakes compounds the others. As the size of the development team and the size of the codebase grows, the problems that these poor practices present become increasingly acute. Understanding these mistakes is the first towards avoiding them in your code.