Initial Thoughts on Mecanisburgo

So, after the first playing, here are my thoughts (mixed in a bit with some conversations I had with Matthew and Nathan). So first off - it does seem like there is a reasonably fun game in here with a decently integrated theme. The first problem is - the icon-ology. There are TONS of abilities and categories and classifications for things and they are all done by icon. This is great once you understand the icons, because the cards are pretty easy to decipher - ONCE YOU GET IT. However, it takes a while to get it - say playing through a full round (there are four rounds in a game) will get you enough that you will only occasionally go back and look at the reference card for an icon. Now that would be ok, except for the amount of effort it takes in that first playing to absorb it all. Trying to just grasp at the system means you aren't going to make much of a coherent strategy in the first game.
But, fear not - coherent or not, you will be in the game due to the instant win victory conditions. Now, I don't know how I feel about these. They are random, so you can't really plan to try and win with one. You might not see it or have the opportunity to use it, but as we saw in our game - being in last doesn't mean anything. Which also means being in first doesn't mean anything either. While I understand trying to keep people "in the game" in longer games, I dislike that I can play really well and still lose to something well out of my control. The instant victory conditions, if treated like a threat by everyone, add a little spice to the game, but otherwise just make it a bit random.
Speaking of random, we now come to the part of the game I really disliked. The mostly random resolution of the special actions. If you manage to win control of an area on the board, you get the opportunity card there. In addition, most of the zones offer a special action if you sent an agent that can perform the action. The actions are things like seduction (steal an agent from another player), sabotage (destroy another player's property), prosecute (force a player to lose a building or agent that is criminal or corrupt), hack (use a computer to steal funds from another player), etc. Most of that would be worth while, except to do the action, you have to play a card (from your limited supply numbered 1,2,3,4,4,5,6) and add to it a RANDOM card from 0-8 (only 3 each of 0 and 8 in 150, the rest evenly distributed). The other player does the same. Which means - you can try to influence the result with your card, but mostly, its random. AND you don't get to know the random until you've selected your card. Its not enough control to make it worth spending your cards on.
My last complaint - the rules. The rulebook was fine as a rule book, and terrible as a reference. Since there is a lot to try and remember, it'd have been better as a reference guide. I REALLY like the way GMT does their rules - very spartan and very straight forward. Rather than putting examples in blocks to the side of the rules, they simply publish a separate game example book which walks through a turn or two and references the rules appropriately. It also explains why you might do (or not take) a certain action. We spent a lot of time trying to figure out exactly what a card or threat or event does. A better reference would have shaved about 30 minutes out of our game.
Now, if it sounds like I really didn't like this game, you are wrong. I did like it, just not enough. I want to play again, but with a few house rules. Matthew and I talked about modifying the random part of the action resolution to be a bit more reasonable. That would help. So would a modified summary of events, threats, robots, characters, etc. I think with that, and not having to re-explain the rules would likely make this a 2.5 hour game. I'm still not sure the payoff is there, but the game still interested me. The idea of the game is cool, but the execution needs a little work. Maybe it just needs those couple tweaks and it'd be a decent game.