Okay, introductions: This is the first play test of a game that in retrospect might not have been ready for it. I would call what we did more of a run through and though it was very brief, but it highlighted a lot of things that needed to be addressed, less procedural and more statistical (that is, balancing the numbers), which was awesome. Two out of the three players seemed to genuinely enjoy the experience, and the other was a little too tired for full participation but gave it a good go. They are all long time friends, one's my younger brother, and barring the last few years we've gamed together a lot. The players were Clark, Garrett and Matt.
The game is called Shadows (which I know, might be taken, and Google comes up with a few things by the same name but nothing published). My design process is purposeful in that I knew I wanted a game, but it didn't come with any specific agenda or goal in mind (more about it can be found on my blog/design diary which should be linked below), so part of play testing is, for me, feeling out what the game should be doing, what the themes and tone are and all that.
The setting is a fantasy one which tries to be a little less black and white than traditional D&D or even LotR, but still maintain a sense of us v. them. The players play the members of a rebellion composed up of the remnants of a displaced knighthood, rebels from the common people and sort of magi-tech alchemists, but ones I'm trying to really hard to not make them seem or feel “steam punk.”
The rebellion began after a contagion killed off a large chunk of the population, creating enough chaos for open revolt to be possible, but it fell to pieces until a Common Sense/Constitution hybrid came out of nowhere and began to be amended and adopted by disparate rebel groups, who then had something around which to rally.
The major bad guys in the setting are the Merchant's Guild who usurped power from the nobility that the knights used to be a part of and the foreign mercenary armies that they pay to protect them and enforce their laws.

Game mechanics wise, I'm going to go through various mechanics and their intended/anticipated implementation and try to contrast it with what happened during the game.

Character Creation: Characters are created with some pretty straightforward point allocation stuff, divided across different types of attributes. This worked out as expected, with characters having just enough points to fill a niche. The character playing the magi-tech types grumbled a little about the expense, which was desirable. Each player found a rebel type they were comfortable with and then went with it.

Quests: Quests are group goals that are suppose to supply structure and group cohesion. They're simply agreed upon in advance before the game starts. Completing Quests increases the character's access to the magical world. Deciding on the Quest was pretty easy, different ideas were combined together and everyone seemed to be on the same page regard what counted as a Quest for starting characters (Quests are suppose to escalate in difficulty).

Expositions/Walk-On: The idea of the walk on is that play starts with one player setting up his character's placement in the narrative and then the other players have their characters “walk on” scene as they see fit. In theory this is all done very low key, in the manner of exposition, but in play it got a little out of hand... one player's walk on had him displaying some real martial prowess and heavily impacting the little bits of established setting that had been worked up by the first two players and I. It wasn't a deal breaker, but it certainly wasn't the planned upon out come. The feeling that these actions were kind of walking on the system made me want to get the Conflict mechanics into play, which in retrospect probably a bad decision.

Conflicts: Conflicts are a little involved mechanically. First things first: The GM can't just initiate Conflicts or ask for dice rolls, he must expend a type of point, called Stance, in order to do so. Each character has a different Stance. So when Matt's character, the last one to enter the game with all that beating people up, didn't have enough Stance to start a conflict with his character, but since when one character is in a Conflict, all characters in a Conflict, I started one with another character and got things swinging.
Conflicts are clunky and took some getting used to, and there were lots of little numbers gaffs, but for physical combat they worked very well in terms of creating fun and drama, or at least they didn't stop us from having fun and creating drama.
The problems came with pacing. I started Conflicts to often and for no real reason and then resolved the Quest, which should have been stretched out to last the night, in a single conflict. This is partly because I was a little overwhelmed by my own pacing mechanics and working them out, being afraid that I would never be able to initiate Conflicts, I flubbed things so I could.

Investments: The game tries to be allegorical, using the character's values as plug ins to the game's Conflict resolution mechanics. In something like the combat Conflicts that we play tested, the value part of the trait pales and the mechanical plug ins shined. Example: Honor translates to how many dice the player rolls when defending his character. So the players liked rolling their defense dice and determining the effects, but they never tied the characteristics to the behavior that they should have felt intensely.

The reason for this, in part, was the complete lack of implementation of the growth/change mechanic, where the player invests his character's characteristics into various elements in the game along with a statement about his desired relationship with the element, he then goes about trying to achieve it, and then based on the way the relationship turns out, he gains or loses and redistributes the invested points.

In theory, Conflicts should occur only when Investments or the Quest is at stake, but I failed to stick to this at all.

Anyway, that's an overview, any thoughts or questions. This is my first real play test post and I'm not positive on how best to go about talking about the experience. Feedback would be great, even if only to point out ways to better handle play testing and talking about play testing.