Creating a good cryptographic algorithm that will stand against all that the best cryptanalysis can throw at it, is hard. Very hard. So, this is why most people design algorithms by first designing the basic system, then refining it, and finally letting it lose for all to see.
Why, do this? Surely, if you let everyone see your code that turns a plain bit of text into garbled rubbish, then they will be able to reverse it! This assumption is unfortunately wrong. Now the algorithms that have been/ are being made are so strong, that just reversing the algorithm is not effective when trying to crack it. And when you let people look at your algorithm, they may spot a security flaw that nobody else could see. We talk more about this counter-intuitive idea in another chapter, Basic Design Principles#Kerckhoffs's principle.
AES, one of the newest and strongest (2010) algorithms in the world, was created by a team of two people, and was put forward into a sort of competition, where only the best algorithm would be examined and put forward to be selected for the title of the Advanced Encryption Standard. There were about 35 entrants, and although all of them appeared strong at first, it soon became clear that some of these apparently strong algorithms were in fact, very weak!
AES is a good example of open algorithms.