Mafiascum.net

If this game is re-run either it should stick to normal roles or double down on roles that work with the mechanic (King, Revisor, etc). I think the half measure made this game too swingy.

I would be interested in a rerun though fyi

Yeah

We can now permanently scrap this idea

If this game is ever re-run, I'd abolish the current proposal mechanism, make a normal lynch mechanism and probably attribute roles somewhat like Civilization with its priorities.

popsofctown wrote:There's like this weird edge case where scum lol-ignore the revisor and lose

But it wouldn't be a gratifying way to win really so calling it is o.k.

The Revisor is also scum (as is the Doublevoter, which is what gives scum the votes they need to win Day 4).

Titus accusing me of tactically replacing because I said your reads aren't very good is pretty unkind.

Nothing else in this game that I'd want to touch with a 50 foot pole but congratulations scum on the win!

In post 3279, callforjudgement wrote:

popsofctown wrote:There's like this weird edge case where scum lol-ignore the revisor and lose

But it wouldn't be a gratifying way to win really so calling it is o.k.

The Revisor is also scum (as is the Doublevoter, which is what gives scum the votes they need to win Day 4).

I could have sworn HH was Revisor.
Which speaks to the complexity of this setup :/

HH is scum

In post 3274, popsofctown wrote:There's like this weird edge case where scum lol-ignore the revisor and lose

But it wouldn't be a gratifying way to win really so calling it is o.k.

Nah. The trick was that with the double voter we could reject every proposal for the rest of the game. If a proposal is rejected, scum submits the list.

The role that could have caused issues was funnily enough, the chronomage but our night action was to rolestop and kill them tonight and then lynch apthet since she couldn’t possibly be the hero.

HH is scum.

Congratulations bingle boy

In post 3278, Creature wrote:If this game is ever re-run, I'd abolish the current proposal mechanism, make a normal lynch mechanism and probably attribute roles somewhat like Civilization with its priorities.

My suggestion would be snake draft at night for role assignments, actually.

Not everyone needs a PR a couple of PRs would just be fine

In post 3283, RadiantCowbells wrote:HH is scum

Oh Dunnstral claimed LUV as scum during d3.

I guess he was screwing with me?

Oh, btw, Dunn won us the game.

I know I lorded the shit out of the win, but Dunn caught the double voter gives us majority bit.

And I probably wouldn’t even have been willing to go for it if I didn’t think HH could get to endgame by themselves without us.

In post 3288, popsofctown wrote:

In post 3283, RadiantCowbells wrote:HH is scum

Oh Dunnstral claimed LUV as scum during d3.

I guess he was screwing with me?

Nah it was insurance.

The number one thing I disliked about the setup

Is that it created such a massive lack of consensus even when there was some consensus on the scumreads to point I just wanted to hammer anything to get d1 over.

d2 pacing turned out fine

But d1 just took so long.

like Bingle can be all proud and gloaty that he passed his proposal day 1 and all that the main thing is that whatever proposal passed felt like it should have happened sooner. It feels significantly worse than in like a mountainous game there might be two leading cantidates each with their own white knights and you get some stasis from that, that's much more mafiaish.

My thoughts on the setup:


I don't think this setup is unbalanced from a win/loss point of view. I do now think, however, that it was unbalanced from a swing point of view; both town and scum have possibilities to completely sew up the gamestate at a very early stage.

The fastest possible win for either town or scum is the end of D2: both are very unlikely (town need to lynch two scum and treestump the third, scum need to dodge three lynches/treestumps and control of all of King, Doublevoter, Inventor). Every member of the scumteam somehow managed to become almost universally townread (Hidden Happiness maybe wasn't as townread as the other two, but she wasn't on many people's scum radar, whereas Bingle (and previously Oversoul) was probably the top consensus townread and players were willing to trust Dunnstral with highly dangerous roles like Doublevoter), making it possible for them to manipulate a forced win by the end of D3, but I'm not sure I can blame that on the setup; when scum are being townread that much, they're more or less inevitably going to win and it's nice that the setup let the win come quckly.

The bigger swing problem, though, is simply that more living scum gives them more chance of gaining a valuable power role. That's something that I didn't really appreciate when designing the setup, and it makes early lynches disproportionately important, one of the classic swing problems. I think that the best fix for this would be some sort of "Vote for Town"-like mechanic in which instead of lynching players to eliminate, the Compulsive Quitter equivalent is good for the faction who gets it (the player stops participating in the game thread but the victory condition is based on how many people of your factiion can be placed into that slot). That gives factions who are losing more control over the voting and should help to normalize the swing somewhat.

In retrospect, some of the roles were also timed incorrectly; for example, Doublevoter should probably have been a D1 role in order to give town more of a chance to lynch them. (I was expecting town to give some of the more dangerous roles to players they planned to lynch the next day, but they mostly only focused on the positive roles). I think I underestimated the danger because I was expecting at least one scum to be lynched some time in the first three days. Likewise, I tried to reduce swing by placing more powerful roles later on (so that the earlier roles would become obsolete), but failed to realise that that would help out scum who were going for a quick win, before the powerful roles even turned up. That said, this sort of balance issue is almost inevitable the first time a mechanic is run, and we've learned enough to hopefully avoid that sort of issue in future.



The other big problem with this setup is that it really benefits from a playerlist who are in tune with the mechanics. Some of the players were and really enjoyed it, but some of the players were just making up the numbers and didn't really enjoy it much. This implies that the game should either have been smaller, or else that I should have pulled the game from signups as soon as it struggled to fill (I was tempted to, and really should have done so, but a player persuaded me not to); if we're struggling to fill the game from the start then replacements probably aren't going to enjoy it.

Games fill more easily at some times of year than others; most likely, if I'd tried to fill the game in the autumn, the players would have had more fun.



Mechanically weird games nearly always misjudge what part of the game will have the most need for discussion, and end up making the deadline too short, which I find very frustrating when I'm playing them. As such, this game I erred on the side of making sure that the deadlines would definitely be long enough, but of course deadlines as long as those that we saw on D1 are their own issue.

For future games, this would be easily fixed by making proposals reset the deadline to 72 hours (unless it's currently longer than that), rather than adding 72 hours to the deadline. (Actually, I think the deadlines were almost right; if I'd reduced the "proposal boost" to 48 hours they would have worked fine. With a lot of proposals D1, though, the extra days really added up.)

There was no consensus because that was the nature of the town
I think a different playerlist could have made this a very different game.

In post 3287, chennisden wrote:Not everyone needs a PR a couple of PRs would just be fine

Not every role was a PR.

In post 3293, RadiantCowbells wrote:There was no consensus because that was the nature of the town
I think a different playerlist could have made this a very different game.

Well 75% of the playerlist seems to be raging against the machine so I guess we'll never know

It's already fairly hard to reach consensus on one single assignment, imagine having to reach a consensus on the entire playerlist at once.

I prefer mechanically scumsided but dayplay townsided setups than mechanically townsided but dayplay scumsided setups tbh.

In post 3294, Bingle wrote:

In post 3287, chennisden wrote:Not everyone needs a PR a couple of PRs would just be fine

Not every role was a PR.

sorry meant to say "role" instead of PR

Welp.