A couple of weeks ago I was trying to noodle out an idea with a friend of mine. I had described what I was going for, trying to draw out some ideas or at least a direction. After I laid out what I thought was a decent rough draft, my friend says, "that isn't innovative."
My first words back were unprintable, but I ask this? Does a game have to innovate to be good or successful? Lots of games use a well-loved mechanic in a new or interesting way and they do fine, but is there any really awesome out-of-the-blocks game that went completely off the beaten path and didn't use something that had been a previous mechanic?
You have two questions and there are two different answers...
To be successful, a game must stand out in comparison to its competition. Having a new mechanic helps but is not strictly necessary. However, the successful game must do something to distinguish itself... Either mechanically, thematically or with components or technology.
To be good, a game does not need to be innovative in Any category, just above average in most all of them.
And yes, somewhere around once every other year or two we add some new mechanic... Deckbuilding being the last one.
You touch on a vast and complex subject: does a design need to innovate to be "good or successful." About a half dozen points leap to mind to discuss -- and that barely scratches the surface.
1) If a design doesn't innovate in *any* way, it's by definition a copy of something else. So, from a player's perspective, why wouldn't I play the game that already exists? As a designer, I may have fun creating a product, but that's a different kind of fun than others get from playing the product. As much as I enjoyed making my "run from zombies" game, if brings nothing new to that genre, why would others play it?
2) There are many ways to innovate outside of mechanics. This is easier to see in video games, but it applies to board games as well. Improving a game's interface, changing its audience, and even re-balancing can all affect success and "good-ness." Take World of Warcraft, as an example. Mechanically, the original game brought very little new to the MMORPG table, but it revolutionized the genre by streamlining the interface, reducing the system requirements, and balancing difficulty toward the "easy" scale. That made it much more accessible and correspondingly drew a larger audience. No one has been able to compete since -- even games that introduce "innovative" mechanics.
3) Finally, innovation won't guarantee success -- or even "good-ness". All media obey a cyclical process that dictates popularity (this is not my idea -- I think it's a standard psychological concept, but I heard about it informally, so I don't know the name or proper references). Think of it as a clock. At 12:00, someone introduces a "fresh" idea, and the audience responds favorably. At 3:00, a second wave of copy-cats arrive, saturating the market. By 6:00, the public has tired of the idea, and even good products, launched into this window, will likely fail. By 9:00, most people have forgotten the idea, leaving room for its reintroduction at the next stroke of 12:00. The actual amount of time this cycle takes varies by media type (I guess), but it's easy enough to see in the world of film and TV -- or games for that matter (the CCG/TCG market, for example).
So, to summarize, I would say that you don't need to innovate mechanics to make a successful game, but you'd better offer *some* improvement (as Matthew Rodgers said), or make sure you're releasing at precisely 12:00 o'clock. :)