Wield: a fun game that I need to get better at GMing

I ran my first game of Wield last night, for four players. You play vatcha (intelligent magic items like The One Ring or the Luggage from Discworld), and the poor saps who own them.

One of the players described it like so:

"Tonight, Matt and I had a duel of wits as adversarial artifacts and their bearers. I played a magical finger and a scholar bearing a magical gauntlet. He played the magical gauntlet and the owner of the magical finger.

Yes."

Some observations, in no particular order:

• The central tensions between the vatcha (the magic items) and the heroes worked well.
  
• It wasn't quite as easy to let the players generate their own stories as I thought. I may need to steal the idea of player-generated Bangs from Sorceror and my own 'take note of highly charged situations and play them immediately' technique from Monsterhearts.
  
• The web of social relationships in the game challenged my savannah-optimised primate brain. With four players, it was just on the edge of what I was able to handle. Most team-based games (eg. D&D, Shadowrun, Leverage) have an easy-to-track 'us versus them' dynamic. In terms of games with (fictional) social complexity, I think it goes: Monsterhearts --> Wield --> The full version of Left Coast
   
• I definitely needed to create my own play aids, in order to understand the combat rules. I've broken it down into a 5-step process, which is easier to follow and adjudicate.
   
• Combat and overcoming obstacles are lethal. They're great situations to force vatcha players to decide whether to give the heros more power (and therefore more control).
   
• After consulting with some Wield GM-gurus, I can see the following player arrangements working:
• A 3-player game where everyone plays artefacts and heros
• A 4-player game where everyone plays the hero for the vatcha on their right or (for a far more focused game, you have two vatcha players and two hero players) 
• For 6+ players, have half the players playing artefacts and half playing heroes.

• I found it much easier to invent destinies for all the heroes. Those destinies also gave me tonnes of NPCs and locations to track (which is a good thing). ... The rules say to randomly assign heroes to the vatchas. When I created heroes, I riffed off what the vatcha players were coming up with. That made me suspect I could be better for the GM to assign heroes to players (at least at first).
  
• As well as creating a relationship map, I probably need to track the history that the vatcha players create. I definitely need to track each vatcha-hero combination: I was getting really lost in doing that.
  
• There's nothing that really points the vatchas at each other. As a result, the stories can drift apart a bit. That's probably fine in a multi-session game. I'll need to think about how you'd optimise it for a convention game.

So, I'd run this again ... after I refine some of my player handouts and streamline some of my 'how to introduce and frame the game' notes.

I'd also think about providing some pre-generated vatchas (to speed up a convention game). I've created 'playbooks' for the vatchas but I think it probably needs to be even faster to hit a climax in a three-hour slot.