Being Too Clever
My friend and colleague +Jeff Tidball
and I often have conversations about RPGs, mostly in terms of the ones we're working on. He frequently takes the position that it's possible for a designer to be too clever with their rules design, such that too many obstacles are created to learning the game (you're trying to do this fiddly dice trick, AND the characters are non-standard, AND the setting is weird, AND you don't like GM screens Cam, what is wrong with you). This also applies, I think, to your intent as a designer for how the game is to be played.
We are learning with the new edition of Unknown Armies that it's really helpful to be up front with how the game is intended to be played, and provide some structure for that. Folks are going to ignore that if they like, or express their opinion that it'd be better if the game was intended to be played in a different way, but it's a great idea to at least be honest.
I wonder if when designing Marvel Heroic Roleplaying that I was not only trying to be too clever in places, but I also wasn't clear enough about how the game was supposed to be played. Because of our various mandates or expectations from Marvel, this was to primarily be a game where players would take existing characters and play out big Marvel events, making decisions for their characters that would maybe have the events turn out differently, etc.
To that end, I designed a game where XP was spent on unlocking story-level as well as character-level changes, such as a new power, a new base, a discovery of some new ally group, etc. XP wasn't supposed to be spent like it is in traditional campaign models. It was for influencing and building up the characters and story toward a big third act finish, determined by the players.
I'm sure many people wondered why character creation was, for the most part, "give your character the stats they should have." It was because you were expected to create a character who enters Act One as a fully-formed individual, with all of their powers and abilities at the level desired for the beginning of the story - powerful or not, experienced or not. When you were done, the character had been through a lot, and was very likely different.
However, for the next big event, just like in the comics, the creative team (the players and the GM/Watcher) might want new characters, or take the characters already played last time and discard story changes and elements that weren't as interesting or important. New power armor? OK, that's good, let's keep that, but let's toss out the awesome base and cast of thousands, because we want the character to go through a redemption arc, one that has them face the world more or less without their wealth and fame. New costume? Well, it was nice for dealing with last event's big bad villain, but now it doesn't really fit. Let's ditch it, and go back to the classics.
I'm not sure how much these design choices fully broadcast to folks how the game was supposed to work. People in some cases really didn't like how character creation worked. They didn't want to play existing characters. They didn't care about events. They wanted to make their own, new, often entirely random beginning level characters, and play week after week getting more powers or more abilities or more NPCs and more of a rogue's gallery. If there were story arcs, they were organic, created in the game by the decisions of players, or they were campaign sketches and outlines created by the GM.
Marvel Heroic wasn't that game. There are a LOT of super hero games that ARE that game. Perhaps, at the end of the day, I was being too clever in designing a game that challenged the traditional model not only of game rules but also the intended experience. I admit that it was mostly because I try so hard to fit the rules to the genre and the license, but I also think that I created a square peg for some gamers who really, really wanted it to plug a round hole.