Okay - looking for a little design direction.
Been working on a game with a focus of altruism vs. self-interest (this would be Horde of Corpses, the zombie game I posted about a while back). A friend suggested an interesting mechanic which I'm trying to incorporate...here's a breakdown, and the issue I'm dealing with.
At the beginning of the game, you'll generate a party of survivors - each player chooses a character, which is then their Active PC. All other unchosen characters are referred to as Inactive. At any time a player can switch to an Inactive PC, moving their character back into the Inactive pool (borrowing from Gregor Hutton's Remember Tomorrow, there). So - periodically, the group will be attacked by zombies or some other threat (i.e., other people - there's a mechanic to determine the frequency of these attacks). When attacked, one of the Active players is killed - unless they spend a Survival Token (a finite resource, say 3 per player at the start of the game) to shift death to another Active PC. That player can spend a Survival Token to shift it to an Inactive PC, while a third player can spend a third token to prevent the death from happening at all.
So - it's cheap to prevent your own death, but costly for the group as a whole to protect everyone. If an Active PC dies, they must switch to a new Inactive PC.
My issue regards the endgame - currently I envision the endgame being initiated when there are only as many characters as players. But what prevents players from hoarding Survival Tokens until the endgame and then spending them all to ensure nobody dies at that point? It seems pretty anti-climactic. I have some ideas for how to encourage players to spend tokens up front, to ensure a lack of them in the endgame...or to mitigate their use in the endgame, but I'm curious what you guys think. I've toyed with using a party tracking sheet, like Warhammer FRP 3rd Edition - something that tracks party tension and provides feedback for social breakdown (letting people die increases tension, etc).