Welcome Guest [Log In] [Register]
The Mafia Waiting List
While we're on the subject, I'd like to maybe open a discussion about stages of readiness vs. sign-ups. I think a potential pitfall of games is getting the game perfectly set and then failing to get the right number of players, leading to either a. emergency last second balance changes or b. an extended sign-up phase potentially leading to less-than-enthusiastic players signing up just so the game can launch.

I in no way intend to say being ready right away is bad or anything, just that perhaps we should be open to starting sign-ups also if the runner is pretty sure they can be ready by game launch time, rather than bumping them to the back/delaying them. It'd be a case of just trusting the judgement of mods on whether they'd be ready in time.

This feels more scattered than it is in my head, but I'm cooking stuff and wanna post before I forget.

Super Dangan Ronpa 2 Mafia Fallout Thread
Okie dokie, my thoughts:

The Good:

Basically, this game did a few really fresh and interesting things that I liked a lot, and tackled a few common problems in very creative ways.
  • Mass claim protection: I was pretty delighted that Yugi had some really good countermeasures in place to prevent the day one cheese from town's end. It's something I don't see in some games, and having it how it was here made for very interesting discussions and gambits. I think it needs noting that there are actually balance concerns based on the way things are handled in preventing mass claims, and one way or another can easily favor town or scum.

    A common trick is to give scum safe claims (I give them X-1 in normal setups, where X = the number of scum). Whether a safe claim comprises fluff, a role, or both can vary--obviously, the more scum has, the more powerful this makes them re: passing undetected. Safe claims with both fluff and roles are really powerful unless there's some sort of tell/the roles are hard to fake.

    The other option is, as in this game, having no safe claims but lots of town members sharing a similar role. Having that role be flavorless, and having it not necessarily track to PR/lack thereof, worked really well in dissuading the strategy (though there were some pretty big issues with one part of the handling--more on that below).
  • Innovation: I love games that are willing to try new things--and I don't mean just in terms of weird power roles. This game was generally speaking really hard to figure out, and I think that, handled carefully, that's a really cool experience. It's especially that way for people who've played a ton of mafia. This game was one where any number of weird paranoid theories came up--an the actual result was weirder than all of them. And that's without everyone having a power or coming back from the dead or being a cop or whatever (all of which can be fun in their own rights). It really captured the paranoid experience that's at the heart of mafia for me, and that was great.
  • Modkills: This could have been better for sure, but you know what? It takes a lot to modkill people when doing so screws the game up, and I think it's 100% the right decision--perhaps not for the present game, but for the state of SOTF mafia as a whole. Inactivity is something that shuts down games and has been such a huge and strange part of our meta for a long time--I can't exactly point t where it started, but it's gotten to be a big deal. Seeing action from the mod side and players realizing when they were unable to continue and bowing out was really heartening to me, and while I think it could've done with a bit more of each in some cases it was a huge step in the right direction.


The Not-So-Good:

Now, other stuff didn't go so well. Here's a breakdown of what I think didn't work and what might help improve future games:
  • Mod activity: This is a thing that seems small but my observations have suggested it's actually pretty much make or break for game activity. If you're modding mafia, I really believe you should post a vote count and deadline countdown every single day. I get why it can seem like a lot when a game balloons twenty pages overnight, but there's a really simple trick I've found: skim.

    The whole rationale behind votes being in bright bold red is so that they're easy to pick out. That means you don't have to bother reading the posts as mod--you're not gonna be voting and know the roles anyways--just skim for votes and update things once a day. I use the zetaboards notepad function and just have a template that I then update every single time a vote changes. I post updates even in phases where there are no vote changes. It usually takes about five minutes if there was a ton that happened. If you miss something because someone formatted wrong, hey, that's their problem for formatting wrong. That's why you ask for the red.

    The reason this is such a big deal is that a constant stream of updates and votes adds some pressure and sense of momentum to the game. It lets people who are less involved at least have a hint of how the game is trending, and it lets everyone know super clearly when the deadline is. It does encourage sheeping votes, but I'd personally take sheeped lynches over inactivity any day of the week.
  • Clarity: I really favor games that overexplain if in doubt. Basically, there are a lot of fiddly details that don't seem to matter but actually do. What time zone is the game's clock on? What order do actions resolve in? What sort of results will a cop get? What happens if a roleblocker is roleblocked while blocking someone else? Do you get a message if roleblocked, and if so, what is it?

    The thing is, it's super possible to answer these questions on the fly, but doing so can screw players up by deviating from the norm (and what "the norm" is in mafia depends massively on your home meta, so assumptions are never safe). For instance, in this game, Ricky's cop results showing as Traitor threw him and Aura for a huge loop because it functioned as effectively a role and alignment scan. There are potential issues here (if Aura knew at first that his results would be TOWN, SCUM, or TRAITOR then he would know there was a traitor in play since the start of the game, while if it was a unified THIRD-PARTY then the information gained by the scan would be less useful). In general, though, if I think anyone could even possibly be confused I like to have things spelled out, and I think that could have helped this game a good deal just in terms of deadlines/vote-counts/whatever.

    Also, as a rule, if you're making anything nonstandard, I think it's really good to think about why and what effects if may have. If you have a double-voter, know how you'll deal with vote counts (and definitely don't just skip them--you'll have to be perfect so your players know that any incongruities are due to funny business and not mistakes, or you'll have to include a placeholder "ANONYMOUS" name or something).
  • Balance: This is the big one, and factors into the next point. Basically, as I've reconstructed things, the town had the following power roles: a cop, a bodyguard, a medium, a one-shot bulletproof (who didn't realize it), and possibly a bomb (I spent all game assuming I was one but was in fact told I was a townie). Scum, meanwhile, had a traitor and three power roles, each of which could at will detect as town.

    Right off the bat, that presents a game skewed massively in favor of scum, because town's sole tool to catch scum out from a power standpoint was compromised by every single member of the scum team. The medium and the bodyguard are both roles that serve, effectively, as power multipliers. A bodyguard dies so a more important role won't, and a medium is a safety net to get info from dead important players. Both offer town almost nothing on their own, and in this setup seemed to be there to give a hand to the cop--but the cop was basically useless because only Ricky could reliably be caught out and scum could appear as town and make town appear as scum. It would honestly have been skewed in favor of scum even without the inclusion of a traitor, just because town had a single power role capable of catching out of mitigating scum's actions in a meaningful way and that role was massively compromised due to scum's powers.

    This also ties to Un's and my roles. Secret roles are really iffy for balance in most cases because they keep players from playing optimally. Bulletproof townies, to have value to town, need to try to draw nightkills. Bombs need to work hard to draw only scum kills while not setting off vigs/JOATS. And again, for all I know I was a miller and/or actually a vanilla townie--I basically gambled on my understanding of the fluff carrying me this game. Secret roles can certainly be thematically cool, and can absolutely be done well (I favor secret millers myself), but the suboptimal play they force needs to be factored into the game balance, because they will just flat out not be as useful to their teams as players who know what they're doing (and can, in fact, be active hindrances--see: Un figuring he didn't matter and playing as townie, then blocking a lynch and tripping town up in the lategame). In general, having extra confusing roles (flavorless is townie except when it's bulletproof (Un) or when possibly-townie has flavor (Me) favors scum more than town because scum at least knows the score with their own powers.
  • Counterplay: Stealing this term from video game discussions. Basically, it refers to the ability to counteract strategies. A great example is follow-the-cop.

    Follow-the-cop can be problematic if poorly balanced because it forces scum to chase the doctor or play whack-a-mole with confirmed townies without ever actually getting rid of the cop (and therefore likely losing unless they get lucky and hit the doctor before the number of confirmed townies is insurmountable). This situation is absolutely problematic (though I take exception to the idea that cops as a role are in any way problematic on their own in any sort of well-balanced game--in fact, mafia as a game mathematically breaks down in some pretty stupid ways without power roles of various sorts and the cop is the best choice for forcing town to not play the boring yet safe game.

    The typical counterplay to follow the cop is to give scum: a. a role that detects as town (often the godfather), b. a role that can shut down the cop even if no kill occurs (often a roleblocker), c. a role that can shut down the cop despite protection (usually a limited-shot strongman), or d. some option to render the cop unreliable (a framer, a miller who has no clue they are one, etc.). Redundancy here is a good thing because it means the cop can't go "Oh, roleblocker's dead, now town wins!" For Star Wars mafia, for example, I gave scum a godfather, a roleblocker, and a strongman kill, which meant that for the cop to be totally safe and accurate, the game would effectively have to be over (or scum would have had to allocate their resources poorly). Ideally, a cop should be always be worried about full-claiming because doing so runs a good risk of leading to them being rendered impotent.

    The issue in this game was that scum had a. members that could detect as town (three of them, in fact) and b. the ability to make members of town detect as scum. This effectively neutralized the cop, who was, again, town's sole proactive power role. In fact, it allowed them to use the cop to their advantage following a claim.

    This isn't necessarily bad! I actually don't think the scum PR spread was in any way conceptually overpowered, and thematically it made for a really interesting team. If I had to make one change, I'd say letting the framer only frame people would have made the team a lot more balanced by preventing the turtling strat vs. detection (the same reason doctors can't target themselves in most metas).

    The big problem was that the cop was town's only way of dealing with what scum could do. See, town needs ways to handle scum just as scum needs ways to handle town. In a typical game, the most common town roles I see are a doctor, a cop, a roleblocker, and a vigilante. All of these roles in some fashion counter scum. The doctor, by picking well, can stop scum's nightkill from going through (giving town more room to make the right choices and potentially saving key roles). The cop can detect scum and confirm town. The roleblocker, by choosing well, can prevent scum's kill from going through and can shut down other active scum roles (their own roleblockers depending on action priority, role cops, etc.). The vigilante can overrule the lynch process and go straight to a kill (potentially, if targeted well, catching out scum and tilting the balance back in favor of town). Each of these roles have their own counters as well, both from scum kills knocking them out of play/scum roleblockers locking them down, and from their own actions requiring careful aiming to accomplish anything (and, in the case of the roleblocker/vig/sometimes even doctor, to not screw over town).

    The issue in this game was that town had no meaningful way to prevent any action scum took. The bodyguard could redirect a kill, but doing so still ticks the clock down towards scum victory. The only role that could absorb a kill and waste a phase for scum had no idea he had that role, and ended up wasting a phase for town instead, again unknowingly. If I was a bomb, that was an option to take out a member of scum guaranteed if I got hit, but again because I didn't know it it wasn't really something working in town's favor (and actively worked against them if I was wrong in my guess).

    If town had a watcher or tracker, they could have checked actions. If they had a roleblocker, they could have locked down some portion of scum's ability to juke the cop. If they had a doctor, they could have kept voices alive and kept LyLo from being quite so looming a threat. If they had a vig, they could've knocked out members of scum without wasting a phase and possibly counteracted some of the inactivity. Without any of that, it's no surprise the game devolved into follow-the-cop because there was literally no other option built into the town's PR spread, and because the scum team was perfectly tuned to completely counter and punish follow-the-cop, town was in a really awful state from the get-go. Scum could have just turtled and focused on detecting as town and there would have been nothing the cop could do to get around it and nothing anyone else on the town team could have contributed from a PR perspective.
  • Dead Chat: I really, really think that this was a game that needed to not have a dead chat. Normally I love dead chats, but in this case there was a role who could talk with dead players, and people in dead chat casually spoil stuff all the time. If, say, a scum member got lynched first day and told the nightkilled target who everyone on scum was and then Bik asked them and claimed, that would have instantly ruined scum's chances for really silly reasons. If there's any chance of players coming back, I feel like they need to be isolated from each other for the game's duration.

    This sits awkwardly with replacements, but in general I feel that replacing back into the game and having a dead chat are mutually exclusive, because a dead chat is a section of game discussion that nobody alive can see, and that dead players can still check if they replace back in. It makes it really hard for living players to know where returnees are getting their ideas from. Did a bunch of dead PRs compare notes? Did one player talk everyone else into a crazy theory for kicks and giggles? It's totally impossible to say, which is really unfair on everyone who does not have access to that information.

    This actually has pretty big game impacts, too. On the first day, I tossed Doc under the bus despite actually being ~75% sure he was town, for various reasons. In doing so, I made some very shaky plays that I knew certain people would or would not notice. When he came back, it was really tough for me to tell if he was pointing fingers my way due to the earlier situation or due to being scum. The thing is, had I recognized the possibility of returning players, I would have played the first phase entirely differently so as to not compromise potential reads later on. I assumed due to the balance issues above that the issue wasn't going to come up, and when it did it really threw me for a loop--and all of the balance problems were still super valid concerns.





    That got long. Despite my nitpicking/theory crafting, I at least actually had a lot of fun with the unusual setup of this game and thought there was a lot of cool stuff going on in theory in the construction. Part of why I've been so thorough in breaking it down is that I think if Yugi ever runs it again elsewhere it has a ton of potential to be amazing, and the other part is that a lot of this is stuff I've been kicking around for a few games and think can be generally useful to mods in the future. This game, due to its recent finish and providing some examples of places where things didn't go smoothly, makes a great launching point for discussion. I haven't read page two of discussion yet due to this taking a while and manifesting over multiple sittings, so apologies for any redundancies. I'll follow stuff here up maybe if I have the time/inclination.

Super Dangan Ronpa 2 Mafia Fallout Thread
DuckyB
Jul 28 2015, 11:16 AM
The use of the Traitor role was also really problematic from a game design perspective. From what Kilmarnock has described of his role conditions, it has no idea who the Mafia is, it has no kill, it is instantly outed in both alignment and role to the cop, and it actually counts for the town rather than the Mafia in numbers win conditions despite winning with the Mafia.
I'm 95% sure that the role was either incorrectly explained, oddly implemented, or a misdirection by Ricky. Typical traitor is just a scum member who's detached from the normal crew, which is a bit of a power damper but has pretty clear play and win conditions.

My thoughts to follow when I'm not in a rush! There were some pieces of game design that were really good IMO, some that had issues, and some that I'll need the role setup to know for sure.

Scrubs Mafia Sign-Ups
I'll play!

The First Mistake
"Yeah, I should be going too."

BB had finished her lunch and then sat quietly watching the others, mostly just following the conversation and judging everyone else at the table a bit—not in a mean way or anything, more sort of a pitying one—and wondering whether it was worth heading anywhere else. And then, before she knew it, enough of the period was over that it really wasn't but it was also close enough to the start of class that it wouldn't be so bad to get there early.

So she hopped onto the bandwagon and followed after Jerry, heading into the halls of Cochise. She gave a little wave at the others as she left.

"See you around."

((BB Gunnerson continued in Smoke On Her Face))

Super Dangan Mafia 2 Game Thread
Also, we're in a very nice position if we don't instantly lose after this lynch.

If Aura dies over the night, Vyse is obviously sketch and we lynch him and win. If Aura doesn't, we get a read on RC and can act on that. The very small number of remaining players means we're narrowing down the possibilities very quickly and will soon have an overall picture of the game.

We should still chat, though, while we wait for votes and such.

New V5 Reduced Activity Notices
I have family in town for the next 8 days (and the past few), hence my extended absence.

Super Dangan Mafia 2 Game Thread
Current remaining crew, by my count, is:

Vyse (Claimed protective role, pinged town)
Me (Claimed townie, univestigated)
Aura (Claimed cop)
RC (No claim)
Deamon (No claim, pinged town)
Un (Claimed oneshot bulletproof, pinged scum)

That means we need four to lynch. In light of the investigation and my greater suspicions yesterday, I'm very willing to believe that Un is scum. This is especially true given the way flips have come--thus far, I think every no-flavor player has flipped town. I find it pretty unlikely that pattern would be randomly broken in a single case in the opposite way that I know it already has been broken--my own status as a townie with flavor.

Credit where it's due, I think Ricky was right in prodding in that direction, and I'm willing to go with the cop on this one since I doubly doubt there's a miller in the mix.

Oh and also my further research into fluff let me know one of the main villains of DR dies and then returns as some sort of AI or something. So there is, shall we say, another fitting fluff option for this one.

Super Dangan Mafia 2 Game Thread
Also, quick note on the modkills:

I PMed Yugi about the activity situation at the start of last phase, noting that I didn't want to lose the game to modkills but also couldn't really in good conscience sit while ignoring things. He let me know that when he prodded them about it, Goose volunteered for the modkill prior to the end of the phase, but RC didn't respond--that's why only Goose was killed.

I only mention this at all so that we're all clear that RC's not getting modkilled is no indication he's scum. I mean, I still think he's scum, but I think everyone should be on the same page in this situation.

Also, VOTE: Un

Super Dangan Mafia 2 Game Thread
I'm inclined to say yes, we lynch Un. Very much so.

See, if my theories/math is right and Vyse is telling the truth, then our scum are Un and RC. That's what the numbers suggested to me last phase; I didn't say anything because I didn't want to provide any more cover for Vyse because he's still my #1 pick if we hit a point where everyone's flipping town.

Vyse, I see no reason to full claim yet. Maybe next phase if RC doesn't ping as town during the night.

Super Dangan Mafia 2 Game Thread
VysePresident
Jul 24 2015, 11:50 PM
Actually, I'm kinda wondering if Imehal wasn't this game's equivalent of a Godfather. I mean, she claimed the Seth kill & green kill flavor, and nobody counterclaimed her. (I can't remember these names very well, sorry.)

Some kind of limited Investigation Immunity? :-/
I actually think that may be possible and it's part of why I'm going with you on this. It is one of the big outstanding mysteries for me.

Also, Aura, I think Vyse may well be semi on the level here. I mean, I stand by almost all my other stuff, but it should be quickly resolved next phase.

Also also, maybe investigate Un instead if you get the chance? Since we've had a lot of questions about his role and so on.

Super Dangan Mafia 2 Game Thread
I mean, I'm convinced enough. And I don't think a full claim will change anything. Either we trust Vyse or we don't. I'm very iffy on doing so, but in this case I read the risk for misfire as greater if we lynch him.

Just, like, remember to do it if Aura dies/nobody turns up as scum in his sweeps.

I don't think we're far from LyLo, though. I think we're sitting basically right on it--that's why I'm easing off on Vyse, actually.

Super Dangan Mafia 2 Game Thread
VysePresident
Jul 24 2015, 07:50 PM
(Yugi's mentioned being used to having about 25% of the game being scum. That means two left, in all reasonable likelihood.) You have guaranteed scum at hand, unless you really, really believe ScumMK's insistence that his role is irrelevant to the game & in fact, is a worse role than SK.
To be fair, we've only gotten one scum member out, so it's probably three counting Ricky.

Super Dangan Mafia 2 Game Thread
Oh no, I do accept your claim and believe almost every part of it. It's just that I think Yugi explained your win condition vaguely or it's not quite what you're saying, because in every game I've ever played/read/encountered a Traitor does count towards the numbers to win (and even if it doesn't, can tie a vote in scum's favor if we misfire today and an NK goes through, resulting in a scum win). A no-lynch today is super bad news, and a town lynch today is super bad news, so you're basically a safe lynch that doesn't tick the clock.

Super Dangan Mafia 2 Game Thread
MK Kilmarnock
Jul 24 2015, 03:24 AM
Toben's changed his vote for the time being, but I'm feeling confused as to why. My brother's not stupid. I KNOW he's not stupid. That's why I find his last couple of posts so strange. The only math that possibly 'checks out' is mine because I threw the more logical argument. So, to this, I kinda have to ask for... elaboration on your elaboration? What you've said thus far, Toben, makes no sense.
Judging based on the balance, I'd be shocked if there are more than two scum left if we accept you as third party (especially with such a power-light town). With the numbers left, that means we actually have a phase to work with. If Vyse is telling the truth, then he kicks it to pull a save. If he's not, he gets lynched tomorrow. I'm basically all set to lynch him tomorrow unless someone flips scum.

A few other people's actions have also made me a little more willing to consider the possibility he's telling the truth. But really, it's about getting a cop read here so we have a bit more to work with. If RC and/or Un flip town, we have Vyse as a backup.

Super Dangan Mafia 2 Game Thread
A few thoughts:

If Aura dies tonight, we should lynch Vyse tomorrow.

If I die tonight, we should lynch everyone Aura turns up as guilty, then Vyse if there's no guilty result.

I'm fine with Vyse full-claiming, but I'd rather have a Ricky lynch than no lynch no matter what.

Super Dangan Mafia 2 Game Thread
To elaborate: Vyse's math pretty much checks out to my count. My push against him was predicated on the idea of his having some sort of special power that would benefit from an extra night--a oneshot nightkill, say--and trying to buy time for that. And that still may be the case. However, there's enough potential here for him to be telling the truth that I think we can take the risk. And if it shoots us in the foot... eh, I'm comfortable choosing to lose in this fashion.

Super Dangan Mafia 2 Game Thread
Here's the thing: I really, really don't trust Vyse here. I was basically at the point where I decided I'd rather lose the game due to being too suspicious than lose the game due to being mislead. I'm really unsure here.

What's changing my mind on this one is that Vyse's arguments seem... oddly desperate. And honestly, having seen Vyse on scum, it came off really differently. But I'm really not sure here. I'm almost talking myself back into things. I'm gonna swap my vote back if the phase end is approaching without overall movement. I don't know.

Super Dangan Mafia 2 Game Thread
UNVOTE
VOTE: Ricky

Thoughts to come.

Super Dangan Mafia 2 Game Thread
I get what you're saying, Vyse. The thing is, after a very lengthy period of consideration, I don't buy it. It just doesn't sync up with how things have gone so far, including under the past players assigned to your role, and this whole situation and your actions in it leave me highly unsettled. I'm almost positive that you're scum of the particularly bad news variety, both due to how I've run through things logically and on an instinctual level based on your play.

As such, my vote stands, and I'd really encourage others to assist. We can't afford a standstill or no-lynch here, and I really think we need to get Vyse out of the picture this phase.

I count two votes on him at present (myself and Ricky). That means we need three more. I'm looking especially to Aura, Deamon, Un, and RC here--Goose's assistance would be great too, but I'm not so sure he's checking the thread.