I've been thinking of MArson. Nvm carry on.
Type: Posts; User: notblackorwhite
I've been thinking of MArson. Nvm carry on.
I think the lack of information the night kill brings offsets a lot of the potential benefits in most circumstances. Like sure some games you're gonna have a villager who is both trusted and correct that'll win town a game they wouldn't otherwise by forcing through an additional kill, but the situation where it's a benefit are pretty rare I imagine. Night kills generate a lot of information (more from just how players respond to them in various ways rather than the direct inferences drawn from them), and not having that is typically a pretty big hit to the village's ability to advance the game state.think there is maybe a marginal benefit to town just because there's a trustworthy voice there throughout the day phase before they dieNot to mention the issue with Mountainous wasn't what happened on N1, and I don't even think it's necessarily beneficial to villagers to skip it. It also feels like a slower and more boring version of what can often be the slowest and most boring setup so. lol.Mountainous sucked for town last year, why do it back to back?Not sure whether it's been proposed yet, but I think the poison mountainous setup that's currently being used for Arapocalypse's mentor/mentee game should be in the running. I particularly like how there's no N1, giving a maximum number of players a chance to play at least two days.
Since there's an ongoing game with it, discussion has to wait till that one's finished, but I know I'm going to forget if I don't post this now
specifically in champs i think the benefits of allowing players to perform for longer, especially under the conditions of being a known villager where they're put in a position where they can effectively lead, outweighs the negatives of making speed slower. I don't really think the impacts it'd have on speed of the game is a huge issue given it's only a delay of a day phase, and even then in a sense it's not really a delay since everyone knows who will be dying at the same time as normal
though i do agree with you that mountainous is boring and should never be used as the champs setup imo, but it's highly likely mountainous will continue to be used in the future so shrug. I do think the poison variation is a bit better than the normal variation if mountainous is still going to be played in champs
I also don't think the extra time is going to be all that beneficial most of the time because my hypothesis is a delay n1 kill makes day 2 more like day 1.5 with the day 1 yeet alone generally doing very little for game progression. Again, won't always be the case yadda yadda. Juice just isn't worth the squeeze in my mind. If you're gonna play Mountainous then play Mountainous and skip the bells and whistles.
Not to mention the issue with Mountainous wasn't what happened on N1, and I don't even think it's necessarily beneficial to villagers to skip it. It also feels like a slower and more boring version of what can often be the slowest and most boring setup so. lol.Mountainous sucked for town last year, why do it back to back?Not sure whether it's been proposed yet, but I think the poison mountainous setup that's currently being used for Arapocalypse's mentor/mentee game should be in the running. I particularly like how there's no N1, giving a maximum number of players a chance to play at least two days.
Since there's an ongoing game with it, discussion has to wait till that one's finished, but I know I'm going to forget if I don't post this now
I agree with most of this post with the exceptions being things I've already talked about, and won't rehash here. The bold sentence I do want to talk about because I think it's reductive, and a bit too harsh generally. I think I know the point you're trying to make in that adding in a role that perfectly counters another, neutralizing both and thereby making it balanced is, well, lazy. Most of the time anyway. However, a setup may need a Fruit Vendor to be balanced for reasons that are not lazy. Context is queen, and everything is relative, and this is especially true for how powerful and interesting a given role is or is not. Like in Sunbae's setup with a wolf firefighter with town arsonist. It's a naively strange inclusion that just functions as a benign visiting role, and yes, it affected the balance, but more importantly it made the game more interesting. The setups I dislike the most are the ones where none of the pieces have any meaningful interactions, and it's just power level paint by numbers: "X PRs and the total power of PRs is this line." That's also why I dislike Matrix setups because you inevitably have variations of power on either side that sort of, maybe works together maybe, but I digress.This entire post is apropos of nothing except me casually skimming along and trying to avoid being negative again and again. I’ll get it all out in one go instead.
I think there are some decent setup designs floated and I don’t want to disparage any because some do look fun and interesting. But also, I really don’t think it’s worth twisting ourselves into knots trying to find the perfectly ‘balanced’ Champs setup. The goal of the tournament is to bring people together and produce some entertaining games for the rest of us to watch and just have some fun. Trying to find the exact perfect mechanic to, like, prevent mass claiming or whatever just feels like we’re overthinking it.
From experience, I’m confident that players and jurors are able to largely recognise ‘skill’ within town/wolf/PR play regardless of the outcome of the game. So all a Champs setup really needs is roles/mechanics that are easy to understand, and an overall design that is accessible (i.e. doesn’t produce situations that favour MU-centric players who are familiar with them).
Where I’m going is that excessively complicated guess-the-PRs-get-a-prize mechanics are unnecessary. Setups that need to include a fruit vendor to be ‘balanced’ are…lazy? Is that too harsh? A setup based entirely around fruit vendors would be one thing, but where they are included just because you want something to counter a certain role makes me just think the setup isn’t right for Champs. And trying to find a semi-open design where all setups are equally balanced and players can choose and whatever is probably impossible and impractical and also just…doesn’t matter.
All we need this year is something with PRs but not too role-madnessy. I don’t see anything wrong with a Matrix tbh – yeah, not every row/column is equal but also so what? They don’t need to be. As long as there’s nothing overtly and awfully x-sided it shouldn’t matter for the purposes of a social tournament. Deserving players will still be deserving and fun will still be had.
Okay, I think I’m done being a grumpy old man now.
I agree with this. Good to be aware of, but no need to twist yourself into a knot preventing.The Bucket wasn't too fun but every other Champs season has been alright so I don't think it's, like, the singular thing we need to build a Champs setup around. Just something to keep in the back of our minds to avoid future Bucket-like situations.
It's a 1-shot doctor and 1-shot vigi. The potential outcomes here are pretty narrow like town can only gain a kill if the 1-shot doctor gets a save and the vigi shoots. Vigi shooting and doctor getting a save are both neutral KP changes and the doctor getting a save is a confirmed which is good but relatively rare even with a full doctor. Delta of variance is extremely low even if wolves don't have a roleblocker at all. The universal backup matters if the roleblocker dies, the doctor would get a save with their action, the vigi will shoot before dying, and the wolves would've accurately blocked a PR without killing them earlier. It's just not gonna happen, and wolves getting punished for letting their one PR die is... fine? Good actually? Universal backup is just excessive.yeah i agree with role cop -> vanilla cop for 3rd setup, can make the 1-shot Doctor a Mafia Goon with that as well as no longer needed without role cop. I'm guessing your suggested changes for setup 2 are a result of it being fairly wolfsided compared to setup 3 for example, so yeah those seem reasonable.Setup 3 I would probably change the role cop to a vanilla cop, and setup 2 I would get rid of the strongman, and change the role cop to a tracker/watcher. Setup 1 I have no idea why wolves need a Universal Backup? It just seems outrageous. Setup 4 has a full Roleblocker and Role Cop against... not enough to warrant those. Just seems like the wolves have enough power to completely overwhelm the town PRs and pretty quickly.
Your anti-mass claim mechanic seems like a non-starter for champs. Even if it was the sort of thing the organizers would be interested in (I suspect it's not), it would need to be automated which is gonna be a big ask. Did I miss Ara or someone else saying this was going to be considered or something?
universal backup is there in setup 1 to reduce swing involved should mafia roleblocker die early, not entirely necessary but I prefer it with than without. but i understand it may be making it too wolf-sided. to be honest it might be best to just make it so that all 3 mafias are goons
agree 4th setup wolf PRs are too strong. JOAT => 1-shot Strongman and making the Roleblocker and Role Cop both 2-shot would probably work
additional mechanic is put in there as an optional thing, realistically i don't see it being used because it looks outlandish, but I believe something along those lines is the best feasible solution to avoiding early massclaims from town while minimising potential issues. The only other thing that might be effective that isn't outlandish is having the setup be one where Town inherently wouldn't ever want to massclaim, and the only open setups where I have ever seen that being the case are setups where there is 1 town pr (particularly where the PR is high power), but those setups tend to have issues of being too close to mountainous / wolfsided (if the 1 town pr is low power) or high swing due to heavy dependence on the lifespan of a single strong town pr (if the 1 town pr is high power).
automating the mechanic seems like it definitely wouldn't be a big ask. the host maybe has to parse a list of guesses from mafia on 1 night, and from that maybe has to take an additional kill action from mafia or change a setting on 1 night
Not every powered 13er has a mass claim problem, and I think people are typically mass claim averse by default so as long as you're not adding too much pressure to mass claim, they're not going to. It was a major issue with Gold Rush because it didn't have role flips, pretty weak town PRs, and the consensus was the village was more likely to lose equity from wolf fake claim potential than they were the PRs. With role flips, mass claims probably don't become meta with the same set of roles.
Yeah there's some merit to this line of thinking. I feel like 2 PR setups tend to be pretty uninteresting mechanics wise with a relatively low skill ceiling, but I won't pretend there isn't merit.
Setup 1 I just noticed it was a Weak Vigi so I'd add the 2-shot RBer back into it (the backup makes more sense now but still seems unnecessary). Setup 2 is fine, and I think my favorite of the four. Setup 3 I may swap the parity of the Tracker and Cop so the Cop is on Even, but this setup really benefits from a 3rd PR potential interest wise so there are more Cop hits and equal probability at rand. Setup 4 is... fine. Again Universal Backup seems excessive and I don't think wolves need the help really, but it's whatever. I would maybe do full wolf RBer instead of 3-shot, and cut the universal backup. I think making the RBer less expendable and not give them a backup creates more room for interest.
Also I am hesitant to make the setup choice preference based for both logistical reasons, and competitive integrity reasons (this may or may not be a weak reason, but it's something to think about). I would still advocate for 3-4 similar powered setups that are chosen, and not announced which setup is being used for a particular game until rand.
Setup 3 I would probably change the role cop to a vanilla cop, and setup 2 I would get rid of the strongman, and change the role cop to a tracker/watcher. Setup 1 I have no idea why wolves need a Universal Backup? It just seems outrageous. Setup 4 has a full Roleblocker and Role Cop against... not enough to warrant those. Just seems like the wolves have enough power to completely overwhelm the town PRs and pretty quickly.
Your anti-mass claim mechanic seems like a non-starter for champs. Even if it was the sort of thing the organizers would be interested in (I suspect it's not), it would need to be automated which is gonna be a big ask. Did I miss Ara or someone else saying this was going to be considered or something?
Yeah which was disappointing. I've tried to mitigate that in the ones I've pitched although it might still happen which is why having multiple setups is worth trying since I think that'll prevent it from happening as much.If you want to design a 17er setup feel free to, but you're right in that the logistical issues make it so that it's unlikely to be picked over the multiple smaller setups which have been suggested thus far!!
I've skimmed the other setups and will get into them in a bit more detail (probably) later, though one thing that does stand out is the focus on mafia hunting for PRs... that doesn't happen much when town PRs make it easy for them by claiming, which I recall was the issue with your last one as well!!!
I wouldn't argue strongly against any of them, but I would argue about the details. All of have upsides and downsides.I’m not reading all the setup specific details because it’s not my cup of tea, but I think we’re reaching the point where we have:
-people arguing strongly that open setups are bad for Champs;
-people arguing strongly that closed setups are bad for Champs; and
-people arguing strongly that semi-open setups are bad for Champs.
It’s almost as if we all have opinions and there is no universal Mafia Setup Truth.
Semi-open tends to mean matrix setup and I would prefer to see Semi-open as one of many open setups or a setup composed with a known set of constraints and small pool of roles. I would agree with Koba that matrix semi-open setups have a pretty bad variance problem, but I'm advocating for semi-open that means something other than a matrix.
On the other hand, Koba's argument against Open setups seems to be that PRs will just claim, and that it's damaging to wolf equity which is not wrong per se, but the implication is the damage it does to wolf equity is devastating. That's a position that feels close to objectively wrong given wolves could still win in Season 8 with Gold Rush bucket claiming, and I've personally been a wolf in three or four 3 PR D6 13er setups, and never lost (it being semi-open was never taken advantage of by any of those teams). Mass claiming I believe is a consequence of design constraints, and one that I think can be addressed as not all open 13ers have developed a mass claiming meta game.
Closed setups are fine in theory, but I would have logistical concerns. Creating like 14 balanced and diverse 13/15er setups using a small role pool, and other design constraints is not trivial. They couldn't be tested (at least not publicly), and requires designers, balance reviews, and organizer reviews. I'm not even sure if you could make 14 diverse setups with the same number of players, and faction balance with all the constraints of Champs, but idk it's worth a try. However, they'd need to commit to it relatively soon, and start working on setups relatively soon.
Thinking further, I think having a 3 or 4 setup spread, and picking randomly from them should just be all good setups and not worry about overlapping them. It's fine if the wolves and PRs know the setup they're in. As long as the village has higher equity with PRs unclaimed, I think it's a net positive. PRs can make informed decisions, and PR hunting can be influenced more via TMI of PRs or lack thereof from VTs.
I take it back you're only a weenie if you have a problem with a 1-shot watcher against 1-shot wolf ninja. 2-shot watcher might be too much in a 15er after further consideration.
Kind of? It's partially a cognitive load issue where upfront complexity is more likely to make someone shut their brain off and go with whatever comes first, and partially that you'd have to ensure the matrix doesn't have conflicting strategies. Like if wolves know the setup based on their own role set and there are two variants with a doctor and in one wolves have a strongarm and in the other they don't, there isn't a strategic choice to be made. It's a coin flip.
Vigis are fine. Maybe slap an Even modifier on there (which I think makes it not necessary to be 2-shot so we can stay at 1 modifier) to reduce variance some. Thinking about it further it should probably be a 15er rather than a 13er but that might go for most of these setups just to marginally increase wolf wiggle room for misyeet candidates since it's very tight in a 3 PR 13er setup. It's fine though, and wolves need to be afraid of it being unchecked.
See the BPV addresses the first thing. It allows them to use a single strongarm shot relatively freely to force a kill through, but if they're afraid of the doctor and blow the second strongarm to force a kill through, the BPV punishes them significantly. BPV also has anti-claim pressure because it passively has high value as long as wolves don't know who the tracker is. A town BPV just plays really well with wolf strongarm.
I agree that n2+ is still a good modifier to slap on the ninja/strongarm pair regardless to prevent the N1 "free" SPK probably.
It's an open setup. They can stop being weenies. It's fine. If you have a problem with a 2-shot watcher in an open setup especially when wolves have a ninja modifier you're a weenie.
Didn't read everything so apologies if I've said something that's already been discussed. Sunbae's setup with the Firefighter was my favorite of the ones I saw. It's very much in line with what I think Good setups are.
Setup matrices are fundamentally flawed. They succeed at producing variable setups from a small pool of roles that are sometimes not guessable after the first or second PR flip, and they can be tuned to have a well-understood balance range. However, they have a pretty big problem in that they rarely produce good setups, and that usually delta of balance variance is bigger than you want it to be (which is a particularly bad problem for Champs where you want things more uniform game to game) e.g. both D6 and Matrix 12 can produce some pretty extremely town-sided setups. Good PR setups usually require some thought to ensure there's a proper tension between roles, and a matrix (it's worse for closed setups) can make PR play more naïve or basic until the setup is known because of how much information a player lacks to make decisions.
In lieu of a matrix, I think just having multiple open setups across the season is fine, and has some nice bonuses like they're less likely to become solved quickly (e.g. bucket claiming didn't happen right away with Gold Rush and it took time for that to become the meta strategy). You can wait until just before rand to post the setup even. If you really want "semi-open" then idk handcraft like 3 setups with role overlap and rand between them. You'll get higher quality setups with nearly all the benefits of matrix with less downside.
Anyways, in general a Good PR (mostly open or semi-open) setup for something like champs is going to be low power town PRs with more frequent utility than being Named VT, and ideally make wolves make kills that aren't naïve SPKs. Wolf PRs should create tension and potential counterplay with Town PRs. Referring to Sunbae's setup, the Wolf Firefighter both added more potential positive results (really important with a town MD), and made its results require inferences to be useful. Win-win.
$%#!ing around with setup ideas (roles are non-consec, no self-targeting unless otherwise noted):
(Semi-)Open 13(15)er #1
Village
---
Compulsive Fruit Vendor
Compulsive Motion Detector (Tracker?)
N2+ 2-shot Vigi
7-9x VT
Wolves
---
Fruit Vendor
2-shot Tracker
2-shot Roleblocker
Dueling Fruit Vendors looks dumb, but they muddy the waters for both invests while avoiding it being just a Named VT for the village. The 2-shot Vigi needs to be a 2-shot Vigi because Fruit Vendor + MD/Tracker is pitifully weak role set for the village and it needs the help. The power of the vigi, compulsive modifier on town invest, and the dueling Fruit Vendors are all there to reduce the value of bucket claiming (wolves can more easily find the vigi n1 since it would be the only one to not act).
This setup is pretty weird, and warped around the vigi. It has some interesting elements, and is gonna be fine balance wise as either a 13er or a 15er, but probably should be a 15er. I really would not reduce the vigi to a 1-shot. Town MD should maybe be a Tracker to make it more potent, but I'm not sure how much it matters and think it might make some of the dynamics less interesting.
(Semi-)Open 13(15)er #2
Village
---
1-shot Watcher
2-shot Doctor
Tracker
7-9x VT
Wolves
---
1-shot Ninja Modifier
1-shot Strongarm Modifier
2-shot Tracker (Watcher?)
There's intended to be tension between the Town Watcher/Doctor pair and Wolf Ninja/Strongarm pair. Individually the Watcher/Doctor have internal pressure to act quickly before they die to use their charges, but don't want to waste it by walking into the wolf counter. The Ninja/Strongarm similarly don't want to pull the trigger too quickly, but can use it force kills through when absolutely necessary. I would consider adding a Night 2+ modifier to the Strongarm as well, and I go back and forth on making it a 1 or 2-shot, but I lean 1-shot especially if 13er. Wolves are pressured to hunt the PRs for both early game kill freedom, and avoid a late game Tracker->Cop upgrade, and the Wolf invest is intended to facilitate. Wolf invest I'm the least sure on, but I don't think it's correct to make it a Role Cop. Ninja existing has the slight benefit of making a Tracker not a cop if they're both alive in an endgame which is a plus.
(Semi-)Open 13(15)er #2 Variant
Village
---
2-shot Watcher
2-shot Doctor
1x BPV Tracker
7-9x VT
Wolves
---
1-shot Ninja Modifier
2-shot Strongarm Modifier
2-shot Tracker (Watcher? 1-shot Role Cop??)
Same general idea as the previous one, but relatively more town power and higher pressure on wolves to PR hunt well. BPV makes the Strongarm very important, and the increased charges for the Watcher mean wolves can't perfectly counter both the doc and watch with modifiers, and will at least occasionally need to think about kills to avoid getting hosed by a PR. I would consider going to 3-shot Doc as well just to reduce sleep chances some and boost town power a bit more. Also think this variant has higher anti-mass claim pressure which is a plus. I'm a bit concerned about the wolves PR hunting tools, and I'm not sure what the best answer is. 2-shot Role Cop is the strongest I would even consider, but I really want wolves finding PRs by playing werewolf and have the button pressing solve only part of the puzzle.
Also why can't we have 17ers again? I know there are some logistical concerns, but it's much easier to design good low power 17er setups than it is good low power 13er or 15er setups. There's just so much more design space with the 4th PR on both sides.
Too late I get credit for all the fun roles and everything else that was good you all saw itwhoa, i never said all the credit...
u can have like, 1 credit, and ill take the rest of the credits
Jokes aside, Katze did the vast majority of design work and I was there as a sounding board for some of the details and sanity checking. I'm not sure I made even a single role that wasn't prompted by Kat one way or the other. I mostly was suggesting tweaks, and spitballing concepts. She's a very good designer and I like working with her.
my contributions were largely the result of your ideas or jokes, but if you're willing to give me all the credit for the fun roles i'll take it ig 😔
but np bby glad i could help out some
This is your fault >: (
@Conq gg. You were my favorite player to read, and I was rooting for you regardless of alignment. I very much enjoy your style/process even if it didn't quite work out this game.
My reads were dog and Katze bullied me in DMs until I came around on Lumi wolf
I'm "friend" and I've been told I lost a bet to SPF over illario's alignment
I didn't read the game, not going to.
For starters: cope, seethe, mald
For second: cutting others down because you don't think your skill was properly recognized is bad form.
For third: even from your own description I don't see what was supposed to be all that special about it. You may have been right on some things, but didn't push your correct reads through and you made some mistakes. You also seem to be leaning heavily on results oriented thinking like "this was a demonstration of skill because we see I was right in hindsight!" That's all fine and good, but me making a reads list by pulling names at random out of a hat and being 80% accurate doesn't make it skillful. Process matters. If you think your process is sound, but others don't it would follow that they don't value it as highly as a display of skill.
Anyways congrats Anne and everyone else who advanced to finals.
Deliverer? Delivery Person?
Also people know it as Caroler and there is value to just doing what people know. Most of the alternate names suggested in this thread are good names, but so good they should be reserved for better roles.
Caroler is the conventional name. I've seen it used elsewhere (not EM) with the same name. Soothsayer is fine. Could go with any divination-based name (Diviner even!) and it would be fine. Something stupid like Etymancer (etymology + *mancer) which is would be a literal Diviner of Names could be fine. It doesn't really matter in this case because literally no name you pick is going to make someone go "ah yeah I bet they give someone 3 names at random".
Edit: also Soothsayer is a much cooler and more flexible name, and the caroler mechanic is a very niche role that should only be used in like role madness or games with strange balance considerations so a random, niche name makes sense. Reserve Soothsayer and Psychic for roles that might be used for commonly.
Can a Disabled Goon carry a factional, and is it blocked automatically? Seems like it could lead to some weird game states if the only the goon left is Disabled and the village can Sleep. If an Empowerer was alive it would mech lock the game even.
Okay so A blocks B, B blocks A. A tries to shoot C but is jailed. C shoots A but it fails because A is jailed. B tries to shoot D but fails because A is blocking B. I don't think this makes sense. Let's say A is tracked and watched. The tracker sees A visit B but not C, and the watcher sees B sees visit A. So B claims JK, the tracker claims seeing A visit B and the watcher claims seeing B visit A. Well that doesn't make any sense because if B was really telling the truth, how did A still visit them? B might get yeeted for having a seemingly impossible claim. (Ignore that like C could claim to have tried to shoot A and had it fail, pretend I set up the same scenario without C and you'd have the same predicament.)
Like it's a way to resolve it, sure, but it also opens the door to some nonsense that's hard for the average player to sus out actually how a complex interaction will play out. In the world with mutual failure, it's still weird, but at least it sounds less like random bull$%#! when explained. There's not an intuitive, clean answer that'll work in all cases here, but I think deferring to ones that feel the least contrived is a good rule of thumb.
It depends. A roleblocker and jailkeeper cross-targeting create a dependency resolution cycle. They both depend on each other. You would need to establish some arbitrary priority system where one of them just wins and resolves while the other does not, or have both of them fail. Could assign priority by choosing one of them to resolve first always (which is probably what you should do given night actions work in automated games even if I think this is a poor system overall), give priority based on alignment if possible (e.g. wolf JK/RBs resolve first), timestamp priority (the action submitted first resolved first), or random priority (you randomly pick one to resolve).
Most of those have issues, but as I mentioned I think realistically you're going to go with the "OoO" approach where one is going to arbitrarily be given priority over the other. I think it's important that whichever one loses is blocked completely so a blocked JK doesn't somehow not block but does protect. I think the way the strict OoO system works is RBer becomes effectively unblockable to JKer so a JKer can protect but not block an RBer which feels pretty $%#!ty, but that is the cost of doing it this way.
In an ideal world there'd be dependency resolution where an RBer is blocked if targeted by a JKer and vice versa (assuming no cross-targeting), but this makes cross-targeting have less of an obvious solution. The simplest solution is to just kind of pick of one to win, like RBer, and have the JKer fully blocked in the event of a cross-target. My preferred option is have all dependency cycles like that just mutually fail because it's a solution that works in all such cases, particularly those where the same role cross-targets like RBers cross-targeting or something like two bus drivers creating a cycle.
Glad we're finally coming around to talking about the timing of sub outs lol
Also all these posts subtweeting some suspected RWSTFO mfers should go grab some receipts and file a report. Same with people you suspect of frivolously subbing out or doing it for $%#!ty reasons. That's going to be pretty effective thing you can personally do to help fix the problem, and even if mods don't think it's actionable right away, you've brought that player to the attention of the mods and they are more likely to scrutinize them in the future.
Huh. I never thought about just reporting the modbot post like I thought that'd be weird, but I guess the thread it creates can edited to make it clear it's for a specific player later on.You can already do that. Report the Modbot Sub Out post and add details there, or use the Report User form for random reports.
Think about the sample size required here to have a statistically significant finding. Any player with that many sub outs is already banned for frequency reasons.
This is gonna sound stupid but could a report button be added to substitutions specifically so someone cleanly report a sub they believe to be based on alignment or something "frivolous"? I imagine suspicions of someone RWSTFOing aren't communicated a lot to the mods directly and rarely with literally links to the sub in question.
The only instance I think of is Sanari who I remember saying in a sign up thread they'd sub out if they weren't a villager and got yeeted on the spot which was pretty funny and well deservedwhat do you mean? if we know unequivocally that someone is subbing out because they randed wolf?
the only two cases of that i can think of are hydre and monstr, both of whom we banned for 3 months
but i think it’s generally case by case and we often can’t know if a sub is actually a RWSTFO
Edit: yeeted as in banned
Edit 2: found it lol
Banned right after thisthen i assume you're ok with not playingJust be forewarned. I have randed wolf my last several games and I generally hate playing scum so I suffered through it, but I will not do it again. Not one more time.
And you $%#!ed about people misrepresenting your positions earlier lmaoI reallly think you're overthinking this, cause a lot of these seem very simple to me. and i would wager they would go a very long way to fixing the problems on MU.Versions of them might be practical fwiw but I don't have the energy to articulate where I think they'd need to be changed. On its face a lot of those are probably not feasible though.
They are certainly better than doing nothing and just pretending that sub outs aren't a problem and that the disparity doesn't exist, which seems to be your plan as of right now unless I missed something
Can a mod clarify the current RWSTFO handling? Is it warning -> ban like I think? I didn't see a mod confirm when I mentioned that earlier, but I could've missed it.
Versions of them might be practical fwiw but I don't have the energy to articulate where I think they'd need to be changed. On its face a lot of those are probably not feasible though.
Ship of Theseus? Nah, we have the Game of Theseus. If you sub out every slot in the game at least once, is it the same game?
Amrock I'm gonna level with you a lot of those suggestions are not gonna happen because they'll impose too much of a burden on mods or hosts, and other ones shouldn't happen because they're too rigid and narrow to be practical.
Also the sub rate by alignment has not always been skewed towards wolves, and has been balanced and skewed towards villagers in the past with the same policy which implies the cause of the current discrepancy is independent of the policy. It doesn't mean the policy couldn't be better, but I don't think it's fair to say the current situation is inevitable under the current policy. And besides reducing the total number of subs is the end goal here and that on its own would reduce the significance of a discrepancy whether or not it existed.
Okay in card games or whatever you don't know what's in the opponent's hand. You only know the cards that have been turned face up, and have either perfect or approximate knowledge of the cards in the starting deck. If you're trying to figure the odds of the cards in your opponent's hand, how they play tells you way, way more than the underlying raw probabilities. When I used to play Magic regularly I often knew a few cards with high confidence just because nothing else made sense, but pretty frequently I had a good idea on the composition of the entire hand. Yes, with no other information the best answer is going to be whatever the mean values are, but you have actual play from a slot it out weighs sub out statistics by a country mile in terms of significance.This is not how statistics workYou should assume it's rand because the rates are skewed across a ton of games in the sample. Not all games are going to have the same rate. Some games all subs are gonna be town, some games are all gonna be wolves. You shouldn't rely on the mean probability of a population for a slot that's literally made post min as the content of those posts are going to be more significant regarding the slot's alignment than rate in the abstract. It's just kind of lazy thinking?
You should assume it's rand because the rates are skewed across a ton of games in the sample. Not all games are going to have the same rate. Some games all subs are gonna be town, some games are all gonna be wolves. You shouldn't rely on the mean probability of a population for a slot that's literally made post min as the content of those posts are going to be more significant regarding the slot's alignment than rate in the abstract. It's just kind of lazy thinking?Nutella why would I assume it’s rand when it’s literally not rand lmaokey word being ASSUME, point is to assume the slot is =rand
What do you mean Visor-kun, b-baka.Why.
So.
Tsundere?
🥺👉👈
Y-yeah? Do you think I like sub outs or something?
If you can't make a half decent wolf case against something entirely bullshitting it it's a skill issue. Just wolf em onto the stand and out of them if you're really convinced that they should eat rope for subbing out. It's not hard, and should be made even easier if you're actually right.
Enjoy your cursed knowledge and please don't spread it in games 🙏
Even with empirical data it's still a frustrating thing to argue about in a game. Just read their posts lol. The policy is in place because you could compile your own empirical data showing a particular player has subbed out as mafia 3 times and is likely gonna be banned for RWSTFO but that's also disallowed because it sucks and is the kind of thing that can ruin a game. You can make up reasons and still believe it, but not everyone is gonna know, and ignorance is bliss in this case.
RWSTFO is an instant ban if mods think that's what you're doing on literally a second offense (with the first being a slap on the wrist and a firm "stop it") IIRC
No I get what you're saying. I'm saying that the framing is misleading because people are going to read that and take away that a sub is 67% to be a wolf, Wisdom already did. I'm saying the phrasing is important to ensure those who have a looser grasp on statistics aren't lead to the wrong conclusion.You are making the base rate fallacy by assuming that from my post"Twice as likely" is incredibly misleading. A player that subs out is 63% to be a villager down from 75% (rand). It's still about twice as likely to be a villager than a wolf. This framing is important. You shouldn't have to talk about it lol.$%#!ing this
times a million
it deserves a different thread and I don't want to talk about it here but like
thisthisthisthisthisthisthisthisthisthisthisthisth isthisthisthisthisthisthisthisthisthisthisthisthis thisthisthisthisthisthisthis
Wolves are 2x as likely to sub out than villagers does NOT equal that of all subs, wolves are 2x as frequent than villagers
@Apogee help me explain lol
"Twice as likely" is incredibly misleading. A player that subs out is 63% to be a villager down from 75% (rand). It's still about twice as likely to be a villager than a wolf. This framing is important. You shouldn't have to talk about it lol.$%#!ing this
times a million
it deserves a different thread and I don't want to talk about it here but like
thisthisthisthisthisthisthisthisthisthisthisthisth isthisthisthisthisthisthisthisthisthisthisthisthis thisthisthisthisthisthisthis
This thread is EIMM and there are no villagers.
As someone who has supported the position of "players can sub out for mental health reasons sometimes" as well as other ones that on the surface agree with some of voices coming in here suggesting some very lenient versions of those positions, I wanted to make my own thoughts on it more clear.
If you're someone who regularly plays games, and if you find yourself in game where you have just $%#!ing had it where there's some semi-toxic exhausting bull$%#!, the thread has lost all fun, and you need a break from stress for IRL reasons, I'm not gonna hold it against you if you do that once every like 2 years. That kind of leniency is earned because you generally show up and reliably put in the work through thick & thin, uphill both ways, and through rain, sleet, hail if required etc. This is something that also just happens to be compatible under the current policy system, and definitely do think a "free pass" system should ever be formalized ever ever it's a terrible idea (no offense to FoL and you did change it).
My position is that you signed up for a game, and that means you play the damn thing to the end. Doesn't matter if it's not fun anymore, doesn't matter if you're getting pushed and you don't like it, doesn't matter if it's the 10th wolf rand in a row, doesn't matter if you're stressed, etc. etc. Unless there are some extreme outside stressors or bad $%#! going on you signed up to play the $%#!ing game so play it. The thread burning down in a toxic hell hole mess obvious changes things, but I'm talking about more (unfortunately) run-of-the-mill threadhell stuff is part of the game. You learn to deal with it 99% of the time or you shouldn't sign up.
I think it has to be this way. Anything less and I think sub rates will be higher than is unacceptable. I don't think automatic sign up bans are good way to reduce sub out rates as discussed, but I generally align with that side of the aisle's positions on the seriousness of the commitment being made when you type "/in". Mods can correct me if I'm wrong, but I also believe this is generally in line with their position as well, and policies, whatever they may be, will be implemented under this same guiding philosophy. Any arguments that fundamentally do not align with this philosophy are sort of outside of the scope of what this thread is trying to accomplish. Changing the site philosophy on the seriousness of the commitment made to signing up for a game is just a way different topic than discussing a policy that reduces sub out rates that's in line with the existing philosophy.
yeah this 100%. Hally and I continue to mindmeld even outside of mafia games who knew
The Great Sub Policy Debate of 2022
v1.8
Change Log:
- 1.8 - Mod on Duty moves from REJECTED to SOFT REJECTED.
- 1.7 - Added more mod thoughts on Mod on Duty.
Chloe claims joat
Lute claims brainbomb
Issue
"The sub outs are too damn high!"
Request for comments regarding "the sub out issue", specifically regarding potential site policy changes as well as ideas on how to reduce game integrity. Discussion has shifted to be almost exclusively towards discussing policy.
Current Policy
Mod: Mods discuss current policy, enforcement difficulties on their end, and what they're doing in attempts to improve things already under existing policy in post #253.
To summarize
- All replacements in all games for any reason are logged, and have been logged forever for all time.
- Mods seek a reason for all sub outs first from the host, and then from the player themselves if the host doesn't know.
- Action: If the sub out reason is frivolous, the player who subbed out is warned.
- If the sub out reason is valid and there are no other recent sub outs, mods do not contact the player e.g. the player's house was hit by a hurricane and they had no internet.
- Action: If there was a recent sub out, mods will contact the player regardless of the given reason(s), and give an (informal?) warning to the player.
- Action: If there are multiple recent sub outs, a more serious warning is given or ban if deemed appropriate given the context.
More context can be found here:
Sub Out Statistics AKA Nerd $%#!
Spreadsheet of raw game data courtesy of Makaze: https://docs.google.com/spreadsheets...sVw/edit#gid=0
(still going to wait a bit to add more numbers here)
Proposals
The Amrock14erProposal - SOFT REJECTED
An automatic sign-up ban occurs with the discussed length being between 1 week and 1 month. No exceptions are given based on reason.
Host forced or unrequested sub outs must be reviewed in most circumstances.
- A host forced sub for 0 posting generally would not require review. Sign up ban applies.
- Any subjective decision made by the host involved in subbing a player out requires review. Sign up ban may or may not apply.
- Players may be subbed out due to host error or for game integrity reasons that subbed out player is not at fault for does not result in a ban.
Other notes:
- The sign up ban will not force a player to be replaced in any other games they are playing in.
The belief is a sign up ban will:
- Create a deterrent against "frivolous" sub outs.
- Reinforce the seriousness of subbing out.
- Provide a break for players with non-frivolous sub out reasons.
Those against a sign up ban are concerned that:
- The incentive can be a negative on the well-being of players who sub out for mental health or other real life reasons that don't prevent a player from participating, but participating has much greater (negative) personal consequences.
- Increases the stigmatization of sub outs for the same above mental health or other real life reasons.
- Creates potential game integrity issues regarding the host and subjective decisions to sub a player. The host's decision may be or may appear to be unduly influenced by the knowledge the player they're subbing out will be banned (they don't want to get their friends banned, they do want to get people they don't like banned).
- Players that would otherwise sub out will instead participate minimally i.e. making the post minimum, and likely make posts of questionable quality.
Mod: Moderators have weighed in on this proposal.
- Automatic sign up bans (and the consequential unbans) would have to be implemented manually under the current system (needs to be confirmed).
- Moderators do not believe it would be effective at significantly reducing sub outs.
- Moderators have indicated they believe it's otherwise impractical to implement, and have indicated that they will not be implementing this policy.
Mod On Duty - SOFT REJECTED
A moderator is assigned to each game, and is available for players to talk to, and can act as a counselor of sorts (I believe this is the gist, but it's been awhile).
Mod: While a moderator may not be on duty, mods are investigating having someone else on duty to perform a similar function. Additionally the mods are investigating having a moderator available at all times to be contacted in cases where something needs to be escalated for an moderator to handle.
Latest follow up:
Sub Out Delays & Reminders (?) - APPROVED
The game host when requesting a sub (via the button on the forums) will be automatically given a message that instructs them to ask that the player can wait 24 hours before subbing out before finalizing their decision, and remind the player that they should only sub out if absolutely necessary, it can negatively affect game integrity yadda yadda.
Alternate: A "Request to be Substituted" button that gives a canned spiel to the player who wants to be subbed out, and requests that they wait 24 hours before contacting the game host.
Mod: Moderators have indicated the first version of this will be implemented in some form.
Casual & Competitive Games
There is not a cohesive or consistent proposal here, but the gist is have some games with the current sub out policy, and typical rule set, and another class that some combination of more stricter either sign up or posting requirements, and possibly different sub out policy enforcement level e.g. subbing out of a "competitive" game results in an automatic sign up ban (or perhaps from only competitive games).
This concept is being mixed into a shared host blacklist, but there is no suggested shared guidelines or policies between the same hosts other than loosely having similar philosophies regarding replacements.
Mod: Moderator comments have assumed these would be all host enforced, but haven't said either way if some kind of official separate game class is a possibility or not.
Updates from Moderators
Mod: Moderators have indicated they are working on changing internal procedures specifically geared towards attempting prevent sub outs before they happen i.e. shifting from a strictly reactive approach to one that's both proactive and reactive. See post #939 for more details.
Oh lmao I didn't read it that closely whoopsOkay you do realize Mod on Duty is not happening right? At no point have the mods said it's happening. This exercise has been fully academic lol. Why would mods reconsider a rule for a policy they're not implementing?Well, that transcript Alison gave me is just one example of what I would be fearful of with this mod-on-stand
If I were ever to play a game with this type of policy I would need assurance that they are mostly acting as venters and wouldn't say anything more than "yeah that sucks" or "sorry you feel that way", and definitely not to the level of informing individual players that others were talked to.
(Again, we literally have a rule on MU that says you can't say if a mod talked to you for this very reason so if you really think this is concern trolling @ hally maybe you should talk to your mod team and remove this rule)
Seems like they are considering it in a different form which I would still have the same concerns about, though I do trust roro to get it right since we seem to be on the same page
I'll go update the summary post rq
Originally Posted by Arete
