Here is a post by a man. A man like you or I. A man who loves his games - and leaves them. A Don Juan of games.
Actually, the essay isn't the important bit. It's the comments. They reveal that many people feel the same, and that many people have the same basic idea.
Pattern adaptation is death. When your player knows your pattern, your player has no reason to play. So they stop.
Sure, some players prefer different kinds of patterns. Some players prefer innovative play, other prefer efficient play, others prefer social play, others live through the narrative. And, sure, some players will play to the end of even the most godawful games, but I don't think you want that to be your sole audience.
Therefore, every game uses pattern adaptation control. PAC. They know - semi-instinctively - "this part is kind of boring. Let's punch it up a notch." They can tell based on experience and trial-and-error which parts of the game are 'fun' and which aren't, and how to make things run.
Some people are more in tune with this than others. For example, people who make successful casual games are very in tune with it. They design their games around replayability - a pattern which is never fully mappable. Thus the appeal of Freecell, Spades, and Apples to Apples. Their rules are designed such that the pattern is never the same twice.
My methodology simply makes that explicit. You can logically determine the best ways to make your patterns adapt. After all, having infinite replayability doesn't help a bunch if your game is a forty-hour long RPG... and having an involving storyline is useless in a game of Spades. But knowing when you need which, and where your sticking points are, is priceless.