"Last One Standing..."

OK. I'm a rookie, so go easy on me. :)

I've had this idea about a Horror game rattling around in my head and felt compelled to share it. If it doesn't come to anything from my efforts, maybe it will generate some useful ideas for somebody else's efforts. Or maybe some collaboration can occur. (Read "No NDA, I don't care if anyone steals the concept, etc.")

So, the basic premise is, like many cheap horror flicks, one character survives. Every other character dies. The built-in end condition is that when the last character is left, the game is over.

I want to use a "Survivor" like model in which all players, including those who's characters have already perished, vote at certain critical crisis points on the play of all other players. They rate their fellow players on three factors: How heroic their character has behaved (the most important), how entertaining their play has been, and how good their ideas and problem solving have been (about equally as important, but less so than how heroic their character has been). The GM tallies the scores, assigns the qualities to the appropriate players, and then everyone makes a die roll with modifiers for being awarded the above qualities. The lowest roll snuffs it. "And so, as the zombies crash through the front door of the convenience mart, little Billy O'Connell is ripped limb from limb and devoured alive."

It wouldn't be the only way to bite it, of course. Just a set way to make sure the field is whittled down over time.

As each player's character snuffs it, they take over an aspect of the story...the big bad guy, the big bad guy's minions, the annoying NPCs, etc....so as the game progresses, you have more and more players collaborating on the conflict end of things and some mounting tension as the brain power shifts from the "Good guys" to the "Bad guys". You get more ideas about how to make things onerous but interesting for the survivors.

I guess I'm looking for a couple of things that I find lacking from many horror RPG sessions I've been involved in:

First, characters die. They should. It's horror, dammit. Not
"Barney and Friends".

Second, the tension mounts over the course of the game and the brain power working on making things deadly but interesting goes up as the game goes on.

Third, there's an aspect of competition. And yet, a tension with the benefits of being Heroic (adding to your survival roll). It seems that horror very often has this as a central theme: "I want to do the right thing, but in the end, I don't have to run faster than the bear, I only have to run faster than the others."

Is this too incoherent of a concept? It seems like I might be trying to mix a gamist approach with a narrative approach to ill effect. Obviously, it would never lend itself to a long running campaign and would quite often be a game that resolved a group of character's stories in a night or two. But is it remotely engaging as an idea?

Cheers,
Tom

This actually sounds like a cool idea to me :) I wouldn't make it a bonus to a roll or anything - I'd make it absolute - whoever the players vote out dies next. Now, given that, it seems to me you'll need some sort of mechanism by which the remaining players can influence the game. Maybe allow them
to play the role that the last 8 people on survivor do - like a jury of some
sort? Still, that's not very involved. Anyway - cool idea :)

Welcome to the Forge, Tom!

It's an interesting idea, for sure. I don't think it's a problem creating a hybrid game (i.e. coherent combos of gamism and narrativism are certainly possible, and maybe even enjoyable).

You've nailed what's the tough thing for most designer's already: "What is it like to play the game?" You can see (I hope, because your words made me see) how the game will be played by the real players sitting around the table, what the stakes are between the real people, and what the 'point' of the game is going to be.

I'd say it makes a great 'one-night' game (a weekender at most).

Design very quick (maybe partially done-in-play) character generation and settle on the simple, easy to use mechanics.

BTW - it's these mechanics (specifically reward mechanisms) that will determine to a large extent whether or not Nar play is more heavily supported or Gamist play is more heavily supported.

One final thought, FWIW, I disagree with Grover. You should definitely make the dice roll. Allowing the players to vote one-another bonuses for the roll is cool - but think of the motivation that the uber-roleplayer losing the dice roll to someone else brings to the table. The UR has real 'get even' Step On Up motivation to make the winner suffer...

I hope you keep going with this idea - could be a really neat final product - kind of the Titus Andronicus of the Horror RPG niche.

Cool.

OK, so brainstorming on character creation and task resolution...

An unstated rule of character design (or maybe it has been stated here with all of the excellent terminology and deep thought about RPG design), it seems to me, is that the mortality of the character should be directally proportional to how long and involved it is to create that character.

If the character has a high chance of snuffing it early, you don't want to have the player and GM put a lot of time into creating the character. I think one of the funniest examples of the counter to this rule is the classic version of Traveller, where you go through an elaborate die rolling process to determine your character's background during which time the character could actually croak, thus wasting your effort and time.

So, with the very short life expectancy of Last One Standing characters, character creation should be very straight forward and quick.

For task resolution, I want something highly narrative and collaborative, yet with a deadly twist to it. Maybe a twist on the brilliant and elegant octaNe resoltuion system? (High roll on dice: 1, 2 GM Narrates outcome 3 GM narrates outcome with player adding one detail 4 Player narrates outcome but GM adds one detail 5, 6 Player narrates outcome.)

How about the same resolution system except with a d20 mortality roll made in conjunction with dangerous tasks? The GM could set a mortality level based on what the character is trying to do and what danger they are facing, and set the low end of the d20 roll as a mortality range?

For example, a battle against a minor henchman or threat could include a mortality roll of 1. Even if you succeed and get to narrate the resolution of the task in its entirety, if you roll a 1 on the d20, your character dies.

Really risky direct confrontations against the Big Bad could have a mortality range of 1 to 5.

Why would you take such a risk if all it gets you is a big bonus to the survival roll at the critical junctions?

Hmmmm. Now I'm thinking that if the characters are passive and ineffective, maybe only one survives, but if they are aggressive and effective, more might survive...Last One Standing might not be such a literal title. Players could end the threat early before it gets down to the last survivor. The risk of heroic acts could be balanced by the possibility of avoiding death for all but one character. And the fact that players who's characters have already died are now part of the creative force behind the threat means that the surviving characters actions would have to be very convincing and effective. I doubt someone who's character snuffed it already would be willing to let the survivors "off the hook" easily.

The mortality and survival rolls could still drive a sense of urgency and tension. The players would know that if they don't confront the problem well, it will take care of the characters in short order.

What do you think? Any thoughts...

Cheers,
Tom

I am quiet a fan of players deciding among themselves who’s gonna die next through votes.
But I feel it’s a bit of “nice” game if everyone has to vote for best actor. Furthermore, there will be difficulties in sorting a dead body when the number of living characters will decrease. My suggestion is:
1. GM decides who dies first, preferably through a totally fortune-based mechanism (i.e. everyone rolls a dice; the lower is called a meal for the zombies).
2. After that, on every climax*, the players whose character is dead choose the next casualty.
Thus, you avoid the possibility of players carefully counting votes and deciding at the last minute to vote to save themselves instead of bringing all their guts in the role-playing phase. Furthermore, players of dead characters will be as an audience, advising zombies tactics and finally deciding who’s next.

This design clearly entails narrative play over gamism.

* I would also advise that decision making is best made when every character is in potential danger: Butch in the basement, his light fading away, Sally under the shower, unaware of the strange shadow in her back, Larry struggling to his feet in a dark and too-peaceful lawn…

First: I like it.

arete66 wrote: How about the same resolution system except with a d20 mortality roll made in conjunction with dangerous tasks? The GM could set a mortality level based on what the character is trying to do and what danger they are facing, and set the low end of the d20 roll as a mortality range?

For example, a battle against a minor henchman or threat could include a mortality roll of 1. Even if you succeed and get to narrate the resolution of the task in its entirety, if you roll a 1 on the d20, your character dies.

Really risky direct confrontations against the Big Bad could have a mortality range of 1 to 5.

Why would you take such a risk if all it gets you is a big bonus to the survival roll at the critical junctions?

Hmmmm. Now I'm thinking that if the characters are passive and ineffective, maybe only one survives, but if they are aggressive and effective, more might survive...Last One Standing might not be such a literal title. Players could end the threat early before it gets down to the last survivor. The risk of heroic acts could be balanced by the possibility of avoiding death for all but one character.

This is an idea that just came to me... Why not eliminate the role of the GM? One person starts out leading the opposing forces, and as the other players die, they join in. Maybe they get to add forces to the opposition under their control, or maybe they just rise again as zombies out for brains. It's kinda like that game Zombie Tag (or whatever it's called), where everyone who gets tagged turns into a zombie who can tag other people, and at the end there's a room full of zombies and one person trying to escape.

You realise that, to a Gamist Gamer, this might be enough reason?

Say, you're free to choose (as player) the mortality rating for the next 'scene', and if you live through your choice, that's the amount of d10 you roll on the next 'critical junction' check. Lowest sum dies.

(One problem is that there might be an optimal mortality rating (probability-wise) that people will then always pick. So there needs to be some reward for making people choose different figures some times.)

If you're going to approach it from a movie-standpoint (or at least a 'plot'), there's no need to have 'strength', 'dexterity', etc. (you could, but there's no need, really).

You could have 'heroism', 'ruthlessness', 'luck' and 'appetizing' as scores, for instance.

Just to help you keep thinking outside of the box....

Hi Tom, and welcome! That's a great game concept.

One quick thought about how to allow heroic actions to influence the "vote". Each time a character performs an heroic action, he gets "votes" which he must cast there and then against the other characters. The more heroic the action, the more votes won.

At the end of a predetermined period (or as a result of a random event), tally up the votes, and whoever has the most "votes" cast against them is the Red Shirt, and dies horribly.

This means that the more behind you are in the vote, the more heroic you have to be to survive, which means you are more likely to die anyway.

"Ganging up" on one guy only should be discouraged; remember that the character in last place could die from heroic attempts to catch up, which means that the next-to-last player is going to bite it.

Hope this is useful (and evil) enough for you!

Regards,

Doug

I just had a thought. So - the threat level needs to escalate as you go, but since characters will be getting killed off, you want the remaining characters to gain in effectiveness, so the escalated threat doesn't overwhelm them. How about this:
There is a shared pool of 'hero points' (probably want a better name than that) Anyone can use hero points at any time - however, the pool is only refreshed when you get to the vote and someone dies. If you want to, you can grab all the hero points at once to do something really heroic, and the other players can't stop you. However, they can all vote for you to die in the next vote.
So this way, as people die, they can join the other side, and increase the threat level without making the game inevitably doomed for the heros.
I like this idea - I'll try to write it up more clearly next week, when I have more time.

Steve

An ideas that popped into my head that might fit the game you're working on (cool idea, btw):

Instead of character generation, how about a set of pre-generated archtypes that anyone can take control of and run. They are fodder for the monster until they successfully become heroic. I'd see this as risking something and succeeding. Perhaps there should be a player resource that they amass and spend. It seems to me that the characters don't compete--the players do.

For instance, in John Carpenter's The Thing (movie), Kurt Russell's character becomes the hero when he subdues a character who has gone off his rocker and is shooting at people. Up until that point in game terms, he was expendable and had no player ownership.

Just a thought about what to do with players whose heroes are killed through misadventure.

An interesting possibility is to design the game as a sort of "survivor-style" game. Have the game progress, one scene at a time, with each new scene setting the scene for one or more dead people, voted on by the living (secret vote probably works best). This'll require some players who aren't going to get pissed when they get "voted out" of cause. Dead characters' players take on some aspect of the "evil", like the ghost hunting the living people.

That means that there's no mechanics to "judge" a character but his fellows. The sense of paranoia that builds between the players could well support the mood.

Note that Chris' idea on pre-made templates also would be very fitting on making it an 'introductory' game for new roleplayers.

'How to host a Murder Mystery', Horror-style.

One thing I'm seriously considering, given Chris's input, is to just have a set cast of generic characters. After all, if I'm basing it on standard horror movies, there usually is a set cast of character types.

Jock
Nerd
Floozie
Good Girl
Annoying Kid
Crazy Guy

Keeping to the mortality theme, I figure the archetypes should be fairly normal. No Kung Fu master or Sharpshooter or anything.

Any other normal horror movie archetype ideas?

By the way, some great ideas out there on how to run with this. I like the heroic die pool idea and the characters as something player's exchange and "inhabit" throughout the game. Maybe there's a couple of variations cooking out there that might work out...

Cheers,
Tom

I really like the idea of voting players off in a horror game, but I've been busy lately, and haven't had a chance to get my ideas down. As a result, my approach is pretty different from what you have come up with, but now that I've finally had a chance to sit down and write it out, I thought I'd share it.

Last One Standing (bit of a misnomer - the game can finish with more than 1 person still alive, and also with nobody alive) [I haven't come up with precise system mechanics yet - square brackets indicate a place where actual game rules are needed]

Core mechanic - the vote.
Every player is a hero, a monster threat, or an environment threat.
Every hero has a hero point, which gives [Insert mechanical benefit here].
When at least half the heroes have spent their hero point, a vote occurs in which one hero will be voted off.
Ties are broken by previous votes - if it's still a tie, vote again, if it's still a tie vote again, but the tied players have immunity (if this isn't possible because the number of players remaining is too small, then bid [Insert mechanical benefit for the voted off player here])
The player which is voted off becomes a monster threat - they get to narrate the next scene, in which they must describe their hero ceasing to be a hero. In the process of this scene, they should describe a new aspect of the supernatural threat facing the heroes. They will be playing that new aspect for the remainder of the game. Note that their hero doesn't need to die, but if they don't die one of 2 things must happen:
1) They join the side of the monsters.
2) They become incapacitated - from now on, one of the heroes must take [Insert mechanical penalty here] or the character will be abandoned to die a horrible death.
After the vote, the remaining heroes have their hero points reset to 1 (unless there is only 1 hero left - in which case he doesn't get a hero point).

Related advantages and disadvantages:
Innocent - character takes [Insert mechanical penalty here], but in each vote, the number of votes against them are reduced by 1.
Badass - character gets [Insert mechanical bonus here], but in each vote, they get an automatic vote against them.
(There's room for more voting-related advantages and disadvantages here, but these are the two that I see most strongly reflected in the genre)

Example of (well, not play, but what I think play could be like)
Zombie movie
Opening - Truck carrying radioactive materials goes off the road near a small town - the body of the trucker is mysteriously missing.
1) The zombie trucker attacks and kills one of the heroes (established fact - zombies)
2) The jock hero organizes a posse for revenge, but when they track him down, they discover a horde of zombies - there's a big fight, everyone flees, but the jock hero doesn't make it out alive. (established fact - there's a horde of them)
3) The cheerleader heroine realizes that she left her little yipping dog outside, when the town is under zombie attack - she foolishly leaves the group to save her dog and disappears. (established fact - zombies can see in the dark) (Note - this is a dynamic that I'm really proud of - in horror movies you see characters doing _stupid_ things - wandering into the basement, leaving the group, and otherwise behaving in a manner so that the entire audience can predict their death - this mechanic encourages that behavior)
4) The hunter hero starts hunting down zombies - he incapacitates them with high caliber shot, and then dismembers them with a chainsaw. The camera zooms in when he accidently nicks himself with the chainsaw - he's infected, and gradually turns into a zombie over the course of the next day. (established fact - zombiism is an infectious disease).
5) The scientist hero decides that zombiism is actually advantageous - look how hard zombies are to kill. He fiddles with technobabble, and eventually turns himself into an intelligent zombie. (established fact - scientist zombie who doesn't just try to kill people, he's still intelligent, and goes around trying to turn everyone into a zombie too).

Bystander points.
Depending on the scenario, there may be a number of bystander points available to heroes.
Each bystander point represents 1 person or group of people.
Heroes can get [Insert mechanical advantage here] by spending a bystander point.
When a bystander point is spent, that bystander is killed.
Threats can get [Insert mechanical advantage here] by describing how the bystander (or group of bystanders) dies.

Hero agendas.
Each hero may have an agenda. Having an agenda gives a benefit to the hero on actions relating to that agenda, but all agendas give [Insert mechanical penalty here] when directly confronting a threat (this represents the hero doing stupid things like trying to take samples while under attack). Each hero can have at most 1 agenda. Agendas can have different levels, for larger bonuses and penalties.
Agenda list:
We must study them! - gives benefit to science related actions
We must profit from them! - gives extra resources
I will display my strength by destroying them! - gives benefit to fighting (penalty goes away if character is the only one left)
We must use them! - gives benefit to actions related to intended use. (industrial - technology related actions, medical - first aid related actions, weapon - fighting related actions)

Scene framing.
Scene framing rotates around the table starting with the first threat (the only player who didn't start as a hero). Scenes framed by heroes come in 3 different types:
1) Unrelated - the scene has nothing to do with the threat, and is there primarily to show off an aspect of the hero. Framing a scene like this gives [Insert mechanical bonus here] to the hero for that scene.
2) Seeking Immunity - This scene represents a hero trying to guarantee their own survival, without regard to other people. [Insert method of generating Immunity level here]. The player with the highest immunity level cannot be voted out. Immunity goes away after each vote.
3) Confronting the Threat - This scene represents a hero trying to go destroy the monsters. [Insert mechanism for confronting the threat here]. Succeeding will create a benefit for all heroes who confront the threat in the future. [Insert mechanical benefit here] (Note that just because a particular hero frames the scene doesn't mean that's the only hero confronting the threat - all the heroes can confront the threat)

Scenes framed by a threat come in one type. Trying to kill a hero. [Insert mechanism for killing heroes here - should be related to the mechanism for confronting a threat. Should get more difficult as more heroes are voted off]. If the threat is successful in killing a hero, the player of the hero becomes an environmental threat.

When a player is voted off, they frame the next scene, and scene framing rotates from there.

Types of threats.
Threats come in 2 types - supernatural threats, which are an aspect of the horror that the heroes must deal with, and environmental threats, which are some problem in the environment which wouldn't normally be a big problem for the heroes, but makes confronting the supernatural threat more difficult. An example of an environmental threat for the zombie game would be the power plant being destroyed, so now there's no electricity in the town.
Sample supernatural threats:
- the monster is ...
- the monster can ...
- there are more monsters ...
- the monsters have found location ... (think Tremors)
- the monster has negated one of it's weaknesses (Dracula tricked the maid into inviting him into your house)
Sample environmental threats:
- Infrastructure destroyed (no power, no lifesupport, no water)
- Nature gone bad (too cold, too hot, blizzard, storm, solar flare)
- Meddling humans - must be from outside the scenario (i.e. not bystanders) (CDC has cordoned off the town and is shooting people trying to leave, MIBs come in and try to kill everyone who knows about the threat, Shortsighted officials come in and try to negotiate the threat) All of these threats involve some effort which is doomed to failure, but will make fighting the threat harder.

Endgame.
The game ends when all the heroes are killed, or when there are at least as many supernatural threats as heros and every remaining hero frames a scene in which they successfully confront the threat.

Notes about mechanics.
Some general effects I'm looking for:
When a hero is voted out, confronting the threat becomes easier, and it becomes more difficult for threats to kill heroes. When the threat successfully frames a scene killing a hero, there is no effect on the difficulty of those actions. At the start of the game (1 threat, n-1 heroes) it should be very difficult to successfully confront the threat, and killing heroes should be very easy but not inevitable (greater than 50% chance, drops to less than 25% chance if the hero spends a hero point). If there is one hero left, and the threat has not succeeded in killing any heroes (1 hero, n-1 supernatural threats), confronting the threat should be a virtual certainty, and it should be practically impossible to kill the hero. If there is one hero left, and no heros have been voted off (1 hero, n-2 environmental threats, 1 supernatural threat), then the odds should be worse than they were at the beginning of the game (because every environmental threat gives the heroes a penalty).
I'm kinda thinking about a dual-level mechanic here. There should be one level of mechanism for interactions which don't involve the supernatural threat, and another level of mechanism for actions involving the supernatural threat. I'd like to incorporate some level of inter-hero competition, and I think a hero-hero mechanic would be useful, but it should be completely trumped by the hero-threat mechanic.

Steve

I love this idea - everything, the survivor-style, the hero point idea, reducing down to one person. I'm just wondering though - do you think it will be scary? If everyone knows the premise, they know that most likely their characters aren't going to last long, so they won't get too attached. The competition aspect between players also takes focus away from the scary threats: the zombies or whatever. Instead, the players would focus on the REAL threat, namely, the other live players. And when a player gets voted off, the other players feel relieved instead of frightened.

What if the "dead" players instead voted who gets killed next? Here's my suggestion: at any point, the "dead" players pick one of the survivors to target. That hero must do something foolish and in character (like the cheerleader going after the yipping dog) that puts their character in danger. At this point, if the rest of the group can attempt to save the vulnerable character, possibly using their hero points in the process. Perhaps you could forbid the targeted character from using hero points in their own favor - that sets up an interesting dependency dynamic amongst the good guys, with a party that cooperates lasting longer. Also, it creates a more in-game way of letting the survivors "vote" by deciding or refusing to spend a hero point. If the character dies, replenish hero points etc. If the character lives, rinse and repeat without replenishing points.

shaheddy wrote: I love this idea - everything, the survivor-style, the hero point idea, reducing down to one person. I'm just wondering though - do you think it will be scary? If everyone knows the premise, they know that most likely their characters aren't going to last long, so they won't get too attached. The competition aspect between players also takes focus away from the scary threats: the zombies or whatever. Instead, the players would focus on the REAL threat, namely, the other live players. And when a player gets voted off, the other players feel relieved instead of frightened.

That's not quite what I had in mind for the endgame -
The game is over when 1 of 2 things happen.
1 - everybody dies - pretty obvious :)
2 - every surviving hero succeeds in confronting the threat, and there are at least as many supernatural threats as heros.

So theoretically, only half the initial players are doomed to die, and it's possible that they all could die.

As far as fear goes - I dunno if I'm aiming for that - If you want the players (as opposed to the characters) to fear the monsters, you need them not to have a sense of control over the monsters - which would tend to indicate a more GM-centered game, like Call of Cthulhu. Note that players do not have a total control over who dies - in addition to people being voted out, the monsters can also kill players.

Steve

...that's why I thought there should be a voting out mechanic *and* possible lethality in between those voting points. Otherwise, there's really no threat in between the voting points and I think that would be rather lackluster.

And I guess if fear is the aim, I've played in horror games where the GM had total control of the monsters. It still didn't necessarily produce tension and fear. (I'd say tension is a better aim, anyway.)

The way I see the overall tone of the game is one of tension and urgency...as I said in a previous post, the players should get the sense that if they don't do something risky, timely, and smart, they'll all die. A mechanic that supports that, say an adventure that has as many voting junctures as players, assures that exactly that will happen. Unless they derail the "natural" course of events, they're doomed.

I've played horror games that support a sanity/fright mechanic adequately, but I was trying to come up with one that includes mechanics that create a sense of threat and urgency. Anyway, that's my 2 cents for now. :)

Cheers,
Tom

Here's a thought for balancing the options of how many people survive, allowing for team spirit, and causing paranoia as well.

How about having a mechanic, either at the critical junction points or in between, where players can decide to divide their character's points (or whatever) between two things: defeating the evil, and ensuring their survival. The allocation is secret until the critical junction. Whoever has allocated the least points to own survival has the highest chance of dying. As soon as the players have spent a certain amount of points on defeating the evil, the threat is gone (but they don't know the amount of points). So each time the player has to make this decision, he or she has to decide whether it's more advantageous to try and end the game early, or to ensure that one's own character makes it farther in the game.

- Christian

Now that's cool. I love it. It's simple, clean, and gets rid of the need to vote "someone off the island", as it were. That could get cumbersome and have the vagaries of favoritism, etc.

Characters could gain and lose overall points depending on their success and failure in between critical junctures.

Maybe with a gambling element...the more you risk, the higher the danger and the higher the potential reward in points. Which would then both help you defeat the overall evil *and* survive at the critical junctures since you have more points to divide between the two.

Many checks and balances. How much to risk in between critical junctures? If you don't risk anything, maybe you end up with less points to spend on survival and defeating the evil, thereby leaving you short when the critical junctures occur.

Wouldn't you then just assign zero points to defeating the evil and put it all into surviving if you didn't have enough points to compete with the others?

Maybe you make it so two people have to roll off at each critical juncture...the one who assigned the least points to survival and the one who assigned the least points to defeating evil! Wouldn't that be twisted? Then you'd really have to think about how you assigned your points.

Cheers,
Tom

I really like your idea of gambling for more points... you take a lower risk in between junctions (but still a risk that others might not take) in order to have a better shot at not being the one killed at the junction. And that way, you have RP scenes in between that actually matter.

I don't think taking the lowest for survival and the lowest for defeating evil would work well, since then everyone will land in the mid-field, and it would take away import from the selfish survival expenditure. Instead, I would create a repercussion mechanism where, if players realize one character is out only for him/herself, they can disadvantage him/her. For example, a player can cancel someone else's points with their own, so three players can lose one point each and cost the traitor three points. Imagine friends ganging up on one another to get one specific guy killed! Maybe it's just my personality that that sounds like a lot of fun. Especially when that guy beats the odds and is not the one getting sliced&diced, and now he can gamble for major payback :)

And maybe some people would assign everything to survival and not gamble at all, but to prevent that from being a sure-fire way, you have to make the gambling rewards simply big enough. No gamble=3 points. Gamble=5 to 10 points, or something like that.

So, you end up with four phases: Hero Phase (gambling for points), Decision Phase (splitting up points secretly), Backstab Phase (cancelling other people's points once revealed), and Time To Die Phase (if evil not defeated, two lowest survival score characters square off). All phases involve role-playing on how the outcome is achieved.

That's just one way, I am sure you can come up with many others.

All in all, I think you have a great idea on your hands here, and I'd love to play it once it's in a playable stage.