Not signed in (Sign In)

Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.

      CommentAuthorTheForumIsDead314 (Advanced Member)
    • CommentTimeAug 24th 2016 edited
    So we all have seen eight (!) variants on the "good guys try to find the evil and destroy him".
    I thought of an idea that takes this to the next level.
    We could either do this as a forum game, OR, I could make it in the CIE with the multiplayer feature.

    First off, the large game consists of several mini-games. Let me explain how the mini-games work.

    Mini-Game Rules
    • Minimum players: 3 players + 1 GM
    Each player is assigned an alignment of Good or Evil, with the condition that there is at least one of each alignment.
    At any point during the game, you can attack another player (post a PUBLIC message). If they are of the opposite alignment, you win. Otherwise you lose.
    There might also be a spy who knows one of their enemies, but cannot attack.
    (I also thought of a Guard, but that complicates matters too much.)
    <ins>Also, for the forum game, you cannot attack until all players have indicated they are active, or 24 hours have passed.</ins>

    When the game ends, score:
    • 1 point if your team wins (whether you attack an enemy. a team-mate actively wins, or an enemy kills another enemy)
    • -2 points if you actively lose (by killing someone on your own team)
    Whether you were attacked or not is irrelevant.

    If there is a tie, divide each game score by the number of people who won it. (So if you are Good, and Good wins a minigame, you fare better if you were the only Good than if you were one of three Goods.) Then add the altered scores. (I prefer playing multiple rounds, but even with that it could end up in a tie.)

    The Meta-Game
    • Minimum players: 4 + 1 meta-GM

    In the meta-game, you are playing in n-1 games at once (n = number of players).
    You are also the GM in a game with everyone else as the players. (Your other games are moderated by your opponents, one for each opponent.)

    But here is the thing: The mini-GMs are biased. Now, when I originally thought of this, I was expecting 4 players, so for 5 (or any odd number) there is a special rule.
    To find your GM alignment, take the majority of your alignments in the other games. (You will not know this when picking roles.)
    You are allowed to give advice, but you cannot win actively, only passively if your team wins. (So if you are good, you want to either trick an evil person into killing an evil, OR get someone good to kill an evil person.)

    Oh yeah, the special rule. If your roles in the other games split 2-2, then you go by which side you are the spy more times on. If that is still a tie, it is decided randomly by the meta-GM.

    Order of Events
    Players: A,B,C,D. GM: X.
    1. A,B,C,D first tell X (not the opponents) what they want the roles to be. So A might say "B is good, and C and D are evil. C is the spy."
    2. X then relays the roles to everyone after everyone has submitted theirs. (This prevents someone from assigning roles based on their own biased alignment.)
    3. All four games are played, with the mini-GM giving information to try to get a certain side to win.
    4. Scores are tallied at the end. Highest score wins. You can also play this with multiple rounds.

    Game labels are based on the user moderating it, so refer to "JohnDoe42's game", unless JohnDoe42 prefers it to be called something else (approved by the meta-GM).

    I think that about covers it. What do you think? Should we try to start a forum game, or make this in the CIE? Or put it on the back burner?

    Players: A,B,C,D
    GM: X
    Italics means an internal thought or other things not posted.
    [spoiler2=A\'s perspective]
    2016/8/26 16:00 -- We have started our game at last! I now need to set the roles for my game. Let\\'s see... I think I will do it randomly.
    Whisper A >>> X: B is good. C and D are evil. The spy is C.

    16:10 -- Ooh, quick response time! Looks like I was the last one. He just sent me roles.
    Whisper X >>> A:
    B\'s game: you are good
    C\'s game: you are good
    D\'s game: you are evil, you are the spy

    The only game I can act in is D\\'s game. So I can decide if I want to spy, or attack. However, I have little information, so I will spy.
    Whisper A >>> X: I would like to use Spy in D\'s game.
    A goes offline until 19:45

    19:45 -- OK, so I got a response. This means I can no longer attack. Also, C just gave some information.
    C: Hey all. Just so you know, in my game, B is a legit spy.
    D: Remember that a spy could be good or evil.
    X to A: In D\'s game, B is on the good team (your enemy).

    2016/8/27 09:00 -- Ooh, B has not signed on! I should see if I can persuade him.
    Whisper A >>> B: {Game A} Hey, both C and D are evil. Pick either one to attack, you will win.
    I also have a message from D, saying that he wants us to convince C to attack B, since I cannot.
    Whisper A >>> C: {Game D} Please attack B. I used spy and he is your enemy! (Unless D lied to me?)

    17:20 -- Attack phase began 80 minutes ago. I will attack C in B\\'s game, since he leaked himself on the chat!
    Whisper A >>> B: {Game B} I am attacking C.
    Whisper A >>> X: {Game B} I am attacking C.

    17:22 -- Quick response!
    X: Game B has been resolved. Player C was attacked by A! The actual roles will be revealed after the metagame ends.
    X: Game D has been resolved. Player A was attacked by C! The actual roles will be revealed after the metagame ends.
    Note: Player A attacked C and gains 1 point. B was also on the same team as A, and scores 1 point as well. This is not revealed until the end.

    Two games left. They could be over at any time.

    22:11 -- I think I should try to mislead C to delay his vote so B can score the attack.
    Whisper A >>> D: {Game A} Your opponents split on teams, so be careful. I would recommend chatting with B or C to find their true identity.

    2016/8/28 12:55 -- Darn, he apparently attacked immediately after seeing that.
    Whisper D >>> A: {Game A} I see your trap. It happened to me 3 times in the past. I will attack B immediately.
    X: Game A has been resolved. Player B was attacked by D! The actual roles will be revealed later.

    17:10 -- This should be it. I just chatted with C about his game and he leaked important details. If this fails... I will probably lose.
    Whisper A >>> X: {Game C} Attack D, please!

    18:55 -- X: Game C is resolved! Player D was attacked by A! This concludes the game.

    19:02 -- X:
    (Roles and Scores) (hide)
    Game A:
    A = Good GM, B = Good, C = Good Spy, D = Evil. D attacked B.
    Game B:
    A = Good, B = Good GM, C = Evil Spy, D = Evil. A attacked C.
    Game C:
    A = Good, B = Good, C = Good GM, D = Evil Spy. A attacked D.
    Game D:
    A = Evil Spy, B = Evil, C = Good, D = Evil GM. C attacked A.

    Final Scores
    Player A:
    A=0, B=1, C=1, D=0. Final Score = 2.
    Player B:
    A=0, B=1, C=1, D=0. Final Score = 2.
    Player C:
    A=0, B=0, C=1, D=1. Final Score = 2.
    Player D:
    A=1, B=0, C=0, D=0. Final Score = 1.
    Tie Breaks:
    Divide each game score by the number of players who won that game. So game A is worth 1, B worth 1/2, C 1/3, and D 1. Then A scores 1/2 + 1/2 = 1, B scores the same, C scores 1.5, and D scores 1. C wins.

    Interestingly enough, no one scored a -2... all attacks were accurate.

    3.14159265358979323846264338327950288... Nothing Here
    Well, now I mentioned the special rule for a 2-2 split in the minigames. I mean, we can have 7 players, but that is not likely and the minigames are designed for 3-4.-----------------
    3.14159265358979323846264338327950288... Nothing Here
    • CommentAuthorXyuzhg (Moderator)
    • CommentTimeAug 27th 2016 edited
    I'm interpreting the minigame to end the moment anyone is attacked.
    I haven't put too much thought into this, but what prevents this strategy in the non-GM minigames (3 players + 1 GM) from ruining all the fun?
    As soon as the minigame starts, before anyone else, pick one of the other three players randomly and attack them.

    As far as I can tell from a quick look, you have a 2/3 chance of winning net +2 over your two opponents and +1 over your ally, and 1/3 chance of scoring net -2 under your opponents and -1 under your ally. If we compute the expected point lead over any other player, that works out to 2/3*5/3 - 1/3*5/3 = +5/9.

    The fact that this zero-interaction strategy appears to net you a lead is a little concerning. Perhaps your own GM game might complicate matters, but the fact that the players know that you're the GM could make any whispers to your allies more convincing.-----------------
    Hopefully PA is inconsistent.
    You do play all games at the same time, which could be confusing and hard to manage. On the other hand, each minigame is very simple and ends quickly.

    Hmmm, interesting analysis. If I modify it to be -2 for an active loss (still 0 for a passive loss), it might work out better.

    In some minigames you might have zero allies. For instance, you are the only good player in a game with an evil GM. Which messes up your probabilistic calculation. Or you are one of two good players with a good GM. Which means you have 2 allies.

    Roles should be revealed after the end of the large game, or else it might get too easy to deduce the GM alignments.

    Interesting stuff.

    I wonder if an active loss could be -3 points, even? Or scale everything up and tweak, like +4, +3, and -5.-----------------
    3.14159265358979323846264338327950288... Nothing Here
    New Update
    • Changed the scoring rule. You score +1 if your team wins, regardless of who did it, and -2 if you ACTIVELY lose. (I might later on double these, and then set a passive win to 1 point.)
    • Sample game complete!
    • If you are only doing one round, a new tie break system is implemented. Divide each game score by the number of players who won in that game. (This also reduces the effect of a -2...) Then sum the altered scores. (Though I would rather just play 3-5 rounds.)-----------------
    3.14159265358979323846264338327950288... Nothing Here