Imagine a sci-fi dome city in an open ocean. For some reason, all computers go crazy, monsters spawn everywhere, and the city begins to sink. There is a hundred or so survivors, crowded together in a safe place. The goal: escape. To achieve it, people have to study surroundings and reach several locations through a series of "rooms".

So, players have a pool of 100 undefined characters. A character becomes more defined the more he or she gets into action - in other words, when he is picked to explore a room. In each room, each player picks a character or two. Think cinematic: at first, a character is just "a mid-aged man in a lab suit". When he solves a conflict, he does this by his skills and traits and thus gets more defined (and effectively powerful). We learn that his name is Sam Berkovski, he is a biologist with the degree in gene engineering, he has a wife and two kids back home...

So far so good. But then the guy dies. It isn't a survival game if there is no way to die. The player then picks another character and goes on - there is a hundred of them, after all. Thanks to this natural selection, in the end we'll have a group of well-defined urban heroes and a crowd of grateful survivors.
In reality, though, I'm sure that character deaths will leave a bad feeling in their respective players, especially if the character was well-defined ("leveled"). When you read a book, you can appreciate such ending to a life story, but in a game, when you get into character, you will probably feel disappointed.

Is there a known way to deal with this? The ideas I got:

• Like in Spirit of the Century, the player gets something for a negative event. Probably a conflict is solved by the cost of this sacrifice, or the subsequent characters get more defined or powerful from the start (less people - each gets more spotlight)... Although I want neither "deaths are win" nor "deaths are fail" attitude.
• GM should describe a defined character's death in such way that it's dramatic or fulfilling for the story. Think Gaiman. On the other hand, this still looks like a weak solution.

Ahey, :)

Yeay, The Poseidon Adventure! :) Cool.

Anyway, you're floundering a bit, methinks...

EvilCat wrote: It isn't a survival game if there is no way to die.

[...]

I want neither "deaths are win" nor "deaths are fail" attitude.

Thanks for an opinion. How about "death is both good and bad" or "death makes things interesting"?

I think J strikes the Gordian knot but doesn't quite finish the cut:

In a game with survival as the goal, death must necessarily be failure.

Failure however, is not the same as being 'bad' in terms of gameplay. Death can be made an entertaining and constructive part of play - it provides dramatic emphasis, emotional involvement and an immersive sense of danger. Done right, death can be it's own reward: the narrative-thrust and limelight generated can go someway to reimbursing a player for his involvement with the character.

If this, you feel, is not reward enough perhaps the transference of a few skills, or powers of a comparable level, to the next character generated? Group 'fate'/survival points returned to the players? 'Monologue of Defeat' allowing players to author something new within the setting: 'My character dies, but he fixes the reactor/holds the monsters off/throws you the child before falling.'

Thanks, that's probably good: all-out effort at the death's door.
I also imagine players sending undefined characters to unexplored rooms, to minimize risk for their faves. This shouldn't invoke the death monologue feature, the exploration is trade enough, I guess.

The question is; He died doing what?

What was so important to him he'd risk death for?

OR if he had no choice but to risk death over it, this is a bitter end where a world snaps up a lively soul. In otherwords you don't think of it as a dangerous world - you think of it as the world that killed Jake, or Fiona. You remember how those characters died. How is remembering them a bad ending for them - or are we like stoneage man in relation to RP characters and the instant a character dies we forget them?

If a player would get a bad feeling then one of the following is happening;

A: The player isn't figuring what's important to his character and is just 'doing stuff' - this is the player failing to forfil his authorship role. Perhaps because the texts of the game didn't inform him to do so,
B: or perhaps because he just like to float around and 'be' in the world and thought he could do that in any old game. This is a player who basically ignored the text telling him what job he has during play. This is essentially a dishonest player. I don't mean that in a huge drama way - I just mean that that dishonesty results in disfunctional play where he has a 'bad feeling'.
C: This is the same as A, but the player just didn't see it in time to formulate anything - he walks around a corner and a boobytrap kills him...uh, wha? The real life time needed to formulate what's important about the character or be prepared to remember his death, wasn't there. This is a bit of a pacing issue. Note: I think it's possible to provide the time after the death, even. It doesn't have to be before. But some time is needed.

I also imagine players sending undefined characters...

Thanks for all the good points.

After some thought, I made following decisions:

• The crowd should be defined from the start at least vaguely, such as "few guys from security, some civilians, both men and women, with children, a bunch of science staff". They are people, not resource.
• The room sequence (adventure) should be solvable with basic character skills and player thinking, emphasis on the latter.
• The goal, the escape, is reached by mastery (learning the environment and how to get by), resources (obtaining useful equipment and control over the computers) and improvisation (clever thinking in unexpected situations).

This way the game should seem fair, I think. First, player "root" for the crowd as whole; second, it all depends on you, the player.

I think you're approaching this the wrong way.  Death isn't the punishment... Living is the reward.  You should structure your game so that in every conflict, your character is presumably already a dead man.  It's only by his wits and cunning does he manage to ever survive... and your reward is living.  For example, in D&D, your character is presumably going to live through every encounter, but sometimes a dragon gets a lucky blow on you.  In your game, every conflict should be "You're trapped in a flooding room, you've got 10 minutes to get out or you're dead."  The next conflict, "Horrible beasts block your only exit from this room, there's no way you can survive..."

Just preserve the atmosphere and attitude, and players won't (or shouldn't) be irritated their characters died at the hands of the GM, but instead proud of how long they did manage to survive.

Also, you should track how many survivors are left.  Not because they're a resource, but so that they can know how many they did manage to save... if any.

Hiya,

I suggest that you are imagining yourself into a problem which isn't a real problem. The sequence goes like this:

1. I have a game idea in which character death is totally fun. (This is true. You do.)

2. Oh, noes! What if a player (or ten, or a hundred, or gasp, all of them) hates the idea of character death no matter what?

3. I must now concoct twice as much text for my game in order to explain at every step that no, it's really fun, and convince these people ...

I'm saying that #2 is nothing more than your fevered imagination making yourself crazy. You have a game idea in which character death is totally fun. There are many, many people out there who will like it exactly as you've described it. There may be a few who can't handle it, and your proper response to them is "get fucked."

I recommend checking out some early Forge threads about a game called Dead Meat, a zombie-horror game which is based on almost exactly the same idea. It was fan-tastic. Run searches in Actual Play, and in fact, the old Destroy All Games forum is still in the Archives section, too.

Best, Ron