I ran some dummy tests last night for my Nagademon 2013 game. The good news is that it was very successful. I found something wrong with what I had.
It's an issue of what makes a good story. I like to think there are three key elements to a story for any protagonist.
1. What does the character want?
2. What's stopping the character?
3. Why should the audience care?
The first two are standard fare for most story games. The last one is about getting player involvement in the game. Players have an emotional connection to their characters and need to share the motivation so that they make the character do something. I've been at game tables with players who've been confronted with The Big Bad Threat and responded with, "Why should I do anything about that? I'm going to find a monster, kill it, and take its stuff."
In my game the first two are clear: get the kids back from the cannibals, but your place in the raiding party is not assured. The third one is where I found the problem. Mechanically it's fine, but emotionally it's not strong enough. There's a good set of obstacles in that part of the game, but I want them to have a stronger connection with the rest of the game.
I consider the dummy testing successful because it found this problem. Now to fix it!