It is very rare for a game that consists only of patterns and rules from other games to be a hit. (Although it does happen, as you'll see a me-too clone like Warcraft sell just as well as Dune 2.)
A game must have something that sets it apart.
So what's the point of writing down rules and patterns? It's a good idea for software engineering, where someone who comes up with a clever new way of doing the same old thing deserves to get bashed over the head with a stick, but it may not be what you want for anything creative. But I'll argue that formalism is still useful:
First: you can't break the rules until you know what they are. If Ron Gilbert hadn't recognized that punishment was a pattern in adventure games of old, it may not have occurred to him to violate that pattern. (Actually an AntiPattern)
Second: if you want to ship a game on time, you cannot afford to innovate in very many areas. The innovation in your game should probably be its focus, and everything else should fall back on tried-and-true.
Or, from Better_By_Design?, June 2003: "A foolish consistency is the hobgoblin of little minds, adored by little statesmen and philosophers and divines." - Ralph Waldo Emerson