I've noticed that TappingStones is not currently on the Blacklist or Probation list. I was under the impression that he was to be added?
IIRC we were discussing whether or not to formally blacklist him when he announced his intention to leave the forum and never return. Since he's been true to his word, we never needed to actually add him to either list.
I wonder if the problem lies in the distribution of responsibility - at the moment, game hosts are the people most responsible for calling out and addressing negative behaviour (flaming/trolling/playing against win conditions/lurking/etc) and they are also responsible to handing out punishments of stern talking tos, replacements, or modkills. The problem lies that the game host also bears the latgest burden of dealing with those punishments in having to find replacements or deal with thei game becoming unbalanced by a modkill. This leads those game hosts to err on the side of caution and be conservative with punishments, which leads to inconsistency and line-toeing.
A way this could be remedied would be to try and reduce the responsibility of game hosts - for example, if we could foster a culture of reporting posts that cointain flaming then we could instate a rule that says "if you get an infraction for flaming, you must be replaced". While that approach would be more rigid, it would see much more consistent reactions from hosts - as can be seen in other forum wide rules like posting role PMs, which hosts know (usually) means mandatory modkill. This would, of course, put the burden of deciding "sufficient flaming" onto the shoulders of the forum mods, but they are a) more trained in making that call and b) don't suffer the conlfict of interest around finding replacements/destabilising the game that a game host has.
As an aside, I understand mods not wanting to red text offending posts, but some form of engagement would be required to make reporting feel more impactful. Perhaps something like a summary every month of mod activity (X posts were reported, Y warnings and Z infractions were handed out), just to provide something tangible to indicate that mods are reading reports.
Alternatively (or even additionally!) the council could be more of a first port of call for contentious behaviour. If standard procedure is for a host to raise grey-area behaviour with the council and they make a decision (involving but not driven by the host) that results in a judgement of "force replace or not", then the host again bears less responsibility for resolving the punishment and the council take any flak. This could be facilitiated with a formal inbox, discord channel, or designated "host manager" council member. I think a lot of the time hosts don't contact the council until things have already escalated, and even when they do, the council resolves things in a very ad-hoc way. A more formal process (even if it didn't have rigid guidelines) would hopefully make things more streamlined and consistent.
Punishing lurking is more difficult because that is ultimately up to the game host. But I think a strong start is to examine what level of activity we really do expect - as Silver stresses, 1 post every 48 or 72 hours is really not actually enough. In parallel, any kind of system to encourage replacements would help bolster any kind of enforcement overhaul. Auto-ins are no longer sufficient incentive, so how can we encourage more people to replace into games?
Overall, I think part of the issue is that game hosts are responsible for identifying unwanted behaviour, handing out punishment, and dealing with negative consequences of that punishment. This understandably leads to reluctance to commit in the hope that saying "cool it guys" will stop it. Splitting that responsibility burden with other parties who have experience and have volunteered to shoulder occasionally being the "bad guys" in controversial decisions should at least make punishments come earlier in the process and more consistently.
Hi everyone. I am regrettably announcing that after just over seven years, I am retiring from the Mafia council. It's been a satisfying honour to help guide the subforum through its ups and downs over that period, but after so long I can no longer muster the motivation and engagement required to properly fulfil the responsibilities of the Council. However, with the current combination of experienced hands and enthusiastic new blood, I have every confidence in the Council and the subforum going forward.
As part of my retirement I will be transferring control of the discord server to current Council member SilverCrys, who will probably look at instating some moderators on it.
I will still be around on the discord, and I will doubtless continue to play occasionally. I am still keen to review games, and will always be open to discussing mafia theory and mentoring.
My sincere thanks to everyone I've worked with: councillors, secretaries, and forum mods alike. Special thank you to all the players and game hosts past and present: without you we wouldn't have our wonderful little corner of the internet, and I am so proud that I have helped look after it.
We could do a sort-of FTQ where we just stick all the games in the queue on a first come, first serve basis but do a review of game status and put the best/most finished game at the front every time a Specialty fires, so we know which one will be next. I guess that leaves the chance that a game never fires, but it gives us the most quality control? It'd effectively just turn the specialty queue into a rolling FTQ you can submit to at any time, but you never know when your game will fire until someone contacts you. Obviously has it's own downsides to just having it be regular queue, but.
I fear that it would be too much book-keeping and time consuming to be honest, as well as difficult to measure. It's already enough hassle to keep tabs on what games are ready to run and getting people to finish them, to have to hands-on review games and calculate their relative readiness and to deal with people complaining about being moved around every time doesn't seem like it would be worth it. If we push the best to the top, then it's just a rolling FTQ but worse because people have to have a game perpetually able to run but it might never fire - at least with the FTQ you have more awareness of time scales.
I'd vote in favor of reinstituting the speciality queue, with the requirement that a first draft of the setup be reviewed prior to entry.
I'm in support. I think we ought to discuss how we go about populating the list. It seems straightforward to fill the top three slots are filled with the winners of the most recent FTQ (in order TBD based on who is most ready to run). But after that, do we just open up a first-come-first served list or give precedence to hosts who entered the FTQ but did not win?
For future FTQs, I suggest we maybe try and incorporate entries from all queues to pick a best game to jump all the queues as part of an event - say we run it once a year and combine it with an invitational player list/team event/cross-site advertising. This would be instead of just doing it ad hoc (how it used to be) since I don't think we really have the player the player demand to support occasionally running a large game on top of the queue structure. Obviously if we do get a surge there's no reason we couldn't run an ad hoc FTQ to fill it in.
I could see bringing back the Specialty queue, since it does seem a bit counter-productive to turn away games that people actually want to run when the other queues are so bare. I do think that we ought to require games to be mostly complete before being added to the queue though, because it sucks to think you have lots of games to run but in reality none of them are ready. This should be ok because the FTQ has similar (but under-enforced) criteria, and at least effort put into games that join a queue that effort has no risk of being effectively wasted as it can in the FTQ.
Whether we would still need an FTQ at all is another question. It was intended originally to allow really exciting games to skip long queues (which we don't have any more) and to stimulate game design through competition (which we might still need?) - so maybe it can be retired as a regular for now and be run occasionally if we need to spur some creativity.
For those who would like to apply to the Council, please send answers to the following questions to the MafiaScribe account. What we're primarily looking is people with the experience, standing, and demeanor to calmly and reasonably settle disputes, the ability to work within a team, and with a vision for the subfoum and - importantly - the drive to see it accomplished.
What do you consider to be the Council's purpose?
What makes you a good fit for the council?
What is your philosophy of leadership?
What are the subforum's biggest problems/risks and how would you go about addressing or preventing them?
What are the subforum's biggest strengths/selling points and how would you go about preserving and leveraging them?
I think we should encourage upcoming hosts to manage player posting fairly strictly so that we can see what systems do and don't work.
How do you propose we motivate this without replacements or modkills becoming any easier? I don't believe hosts don't strictly monitor activity because it is too hard or they don't care - they don't do it because the enforcement system we have is difficult to implement.
Queues will fill rapidly enough, if it's a game people are interested in.
I think it may be a good experiment to create a schedule for at least certain game types. That gives people a date to put on their calendar rather than check in daily, so we'll be able to capture those players, and gives us the capacity to run things outside the queues. As is, it's incredibly frustrating seeing a game you don't want to join fill up glacially slowly, with no other options in the wings to kickstart that process. We cannot continue with that system. We cannot let the interest level in one game, dictate the interest level in mafia for the entire subforum, for months on end. We won't survive.
By "Queues" we mean the queues for the games themselves, which are basically empty. How can we schedule games when the only queue with more than ONE game ready to fire is the Micro queue (with two)?
I think Bur makes a very salient point: any changes we might make to the queue system are somewhat moot unless we can also refill them.
With regards to activity, I think there are multiple factors compounding to produce our current situation.
Meta - Taking cues from other sites, people are now used to deadline scrambles to produce lynches (which is made worse by our majority lynch requirement as opposed to plurality), and the hyper-activity used to on other sites can make it difficult for MTGS regulars to stay caught up, and for everyone to cope with the inflated post count.
Incentives - We are lacking in incentives to smooth out activity or end the day before the deadline. There are no real arguments for ending Day early that are effective, and the deadline is the only driver to compromise and settle on a lynch. These both lead to the most active parts of play clustering around the deadline with the rest of the Day spent on vanity wagons or just waiting for CFDs.
Enforcement Mechanisms - Probably the most important point, I think our enforcement mechanisms are insufficient.
One post every 72/48 hours isn't really a reasonable level of activity (7-4.6 posts in a two week game Day is really not enough). Prods as a mechanism worked fine when it was to remind people that a game existed or to limit the length of break people could take, but players sticking to the letter of the law are not really participating enough.
Modkills/replacements are challenging for mods to use: we don't have many people available for replacement (long gone are the days of overfull signups), and modkills can have a huge and unfair-feeling impact on a game. This leads to people not being punished as frequently or consistently as we might like, which means they're not dissuaded.
In a similar vein, the smaller playerbase means that the probation list goes pretty much unused because mods need warm bodies to fire their games and don't (rightly or wrongly) feel like they can be too picky about players. Similarly restrictions like "one game at a time for chronic lurkers" isn't much of a restriction when we only run a couple of games at once most of the time.
So I think the problems with activity are a result of the smaller playerbase both allowing lurkers into games more frequently as well as making them more challenging to punish, while we also lack incentives to drive even participation throughout the Day and as modern metas drive activity to concentrate on the end of Day.
Solutions
Well, obvious solutions would be things like more strictly enforcing replacements/the probation list, or even baking modkills into games by default despite how challenging these might be (very). If done as a concerted effort (and people really made an effort to be available to replace) and the forum was ok with maybe firing games less often in exchange for better playerbases then it could have a real impact. However, enforcement isn't the only (or best?) solution - I think we need to try and restructure our incentives to participate evenly throughout the Day and avoid deadline scrambles. We could do this by having slightly shorter deadlines with unused time carried over to the next Day (to incentivise ending the Day early), or by mandating more posts during the prod period (5, say) - or even a cap on posts for each prod period to reduce spam.
I even consider a more complex approach where each few days the votecount is taken and people at the bottom are removed from the pool - this forces the possible lynches to reduce over time so the deadline scramble is at least between a couple of choices rather than the whole playerlist, and prevents people spending all Day on one-man vanity wagons. There's obviously issues (especially how to allow people to be re-added to the pool after being removed, and how much it changes the game from traditional Mafia), but maybe something more radical is needed to keep activity consistent throughout our long phases.
We could even offer actual rewards for things like hosting, or replacing in, or not getting prodded, or for being voted sporting/fun to play with. Every time you perform the desired action, you get entered into a draw and at the end of the year we give someone Curse Premium, or $20, or we interview them, or we let them play a game while being mentored by someone highly respected, or they get to design a game with someone highly respected, or they get auto-in to an invitational, or the nearest council member buys them a beer. There are lots of possibilities.
But ultimately that's just taking our current system and leaving sign ups open for even longer, albeit with a hard cut off date. We already have the PCL for signalling interest far in advance, and it sees very little usage. I also definitely don't think it gives us more flexibility - we have to commit to a game X months before it runs, as opposed to the current system which allows us to essentially pick whatever game we think would work best (at the cost of less notice).
We also don't want it to make it more difficult to host games (e.g. requiring backup mods and setups finished fat in advance), as we have few enough games to run as it is.
Stickies
Definitely could consolidate these, I think even right now we could:
Unsticky the Theory & Discussion thread.
Combine the Replacement and Player Contact list thread (Discord is a more active place for replacements, and announcements of games tbh)
Create a new (locked) thread under the banner of "Resources" which includes the general rules, information for new players, our articles, and the Vote Counter
This would take us to the Council thread, a PCL/Replacements thread, Hosting/queues thread, Resources thread, and the open sign ups.
Staff
I think saying Iso is our only hope is a bit disingenuous consider admins Cythare and Wildfire are both pretty familiar with Mafia, and we have our local mod Bur, plus other faces on staff like ExpiredRascals. But obviously I don't know what goes on behind closed doors, and I can only hope that if this discussion was actually serious then the MTGS staff would involve the Mafia community at large.
Game Scheduling
I really don't think this is a good idea. Fundamentally, I don't think the solution to "our sign ups take too long" is "ok, let's make the sign up period longer". While it'd be nice to know when a game will fire when you sign up, I think it is impractical to ask people to commit X months in advance, and you will have many problems with people forgetting/no longer being available but not outing/mods not being ready or falling through leaving people effectively signed up for a game they don't want to play, or even people waiting months and months only to get randomed out. The significantly simply solution to knowing when a game would start would just be a hard cap on sign up length.
I also think that scheduling when games will fire is a very difficult process: game length is variable, which means it will be very difficult to gauge how many people will actually be available to play when a game fires, and I think it will be very difficult to avoid periods where either a game could fire but isn't (because some people are free but the fire date hasn't hit), or the inverse and overlaps of games that strain the playerbase. Basically, I don't think you can reliably predict when a game should fire, and I think the resulting scheduling issues would cause more harm than good - it reduces flexibility in a system that leans quite heavily on the ability to look at what games are actually ready to fire and the availability of the player base to put a finger in the air and pick a game to open.
I think scheduled games would be an excellent idea for a very small forum that struggled to get enough players for a single game, so as to coordinate the few members it has to get something off the ground, but we are some distance from being in that position.
I think ultimately what we want to always have some sort of sign ups open and for games to fire quickly. Perhaps the problem is we do not have the numbers to support this policy: when we hit current capacity we either have signups which take a long time to fill (usually requiring an ongoing game to end), or we just have no sign ups open at all (and then people just walk away because there's nothing going on). This is sort of the problem the PCL was supposed to fix, but it clearly hasn't. Scheduling games would partially fix this by ensuring that games fired at regular known intervals, so that people when to come around, but at the cost of flexibility and being worse at responding to demand or playerbase utilisation. Deadlines on signups would at least force the next game to refresh until it hit one that was exiting enough to bump demand up to actually fire, but risks skipping over games simply because the playerbase is at capacity (which I think is what has happened to Osie's Micro - people are occupied with the team games and Shadow's Normal, so demand for any new game is low).
Based on the numbers I put together in 2016 and conversations with admins and some people from MU:
-MTGS user numbers as a whole more or less recovered from the Curse takeover in about a year, whereas the Mafia sub-forums have continued to languish. Not sure on the MTGS user numbers at the moment, but forums in general are suffering stiff competition from reddit.
-Mafia forums seem to be doing just fine: MU is very active, although I think MafiaScum is having some issues (but that's heavily cultural in that they have MTGS deadlines with MU-level activity which is causing absolutely massive games filled with low-content posts). MTGS is very much struggling to recruit from MU because despite our imaginative and usually well balanced games, the deadline length is something people really don't want to endure - MTGS is very unusual in the aspect of our deadlines.
So I don't think our decline is necessarily a consequence of a decline in general mafia interest, forum use, or MTGS itself. MTGS is doubtless no longer at the height of popularity it once was, but it is still proportionately far better off than our subforum.
On the rolling sign ups, I think it definitely has some of the drawbacks you mentioned: games can take a while to start, and people might just stop checking back if the same game stays open for too long. But if we are going to shelve games that don't fire within a time period, we have to be very sure that we are ok with abandoning our historical "everyone gets a go" policy, and also avoiding disincentivising people from making games because they think "what's the point of all the effort if it won't even get played?". We would also need a lot more games in the queues or the faster turnaround will stall out quickly. But at least the shorter queues make it easier for someone to have a game fail to fire and then make changes and get another shot.
An alternative might be having a rolling queue for each game type - at least then a stagnating game only blocks one queue as opposed to all of them, and it would allow us to more easily gauge what the popular game types are. I know the Player Contact List sort of fulfills this function, but it doesn't really in practice. Downside would be a lot of sticky posts (maybe a subforum for them with a BIG OBVIOUS THREAD pointing at them?).
There isn't a formal process as such, but I think all three of you have enough of a history building and running games that I'd be happy to see you all on the approved reviewers list. Just make an effort to team up with an experienced second reviewer.
IIRC we were discussing whether or not to formally blacklist him when he announced his intention to leave the forum and never return. Since he's been true to his word, we never needed to actually add him to either list.
A way this could be remedied would be to try and reduce the responsibility of game hosts - for example, if we could foster a culture of reporting posts that cointain flaming then we could instate a rule that says "if you get an infraction for flaming, you must be replaced". While that approach would be more rigid, it would see much more consistent reactions from hosts - as can be seen in other forum wide rules like posting role PMs, which hosts know (usually) means mandatory modkill. This would, of course, put the burden of deciding "sufficient flaming" onto the shoulders of the forum mods, but they are a) more trained in making that call and b) don't suffer the conlfict of interest around finding replacements/destabilising the game that a game host has.
As an aside, I understand mods not wanting to red text offending posts, but some form of engagement would be required to make reporting feel more impactful. Perhaps something like a summary every month of mod activity (X posts were reported, Y warnings and Z infractions were handed out), just to provide something tangible to indicate that mods are reading reports.
Alternatively (or even additionally!) the council could be more of a first port of call for contentious behaviour. If standard procedure is for a host to raise grey-area behaviour with the council and they make a decision (involving but not driven by the host) that results in a judgement of "force replace or not", then the host again bears less responsibility for resolving the punishment and the council take any flak. This could be facilitiated with a formal inbox, discord channel, or designated "host manager" council member. I think a lot of the time hosts don't contact the council until things have already escalated, and even when they do, the council resolves things in a very ad-hoc way. A more formal process (even if it didn't have rigid guidelines) would hopefully make things more streamlined and consistent.
Punishing lurking is more difficult because that is ultimately up to the game host. But I think a strong start is to examine what level of activity we really do expect - as Silver stresses, 1 post every 48 or 72 hours is really not actually enough. In parallel, any kind of system to encourage replacements would help bolster any kind of enforcement overhaul. Auto-ins are no longer sufficient incentive, so how can we encourage more people to replace into games?
Overall, I think part of the issue is that game hosts are responsible for identifying unwanted behaviour, handing out punishment, and dealing with negative consequences of that punishment. This understandably leads to reluctance to commit in the hope that saying "cool it guys" will stop it. Splitting that responsibility burden with other parties who have experience and have volunteered to shoulder occasionally being the "bad guys" in controversial decisions should at least make punishments come earlier in the process and more consistently.
You too, you are invaluable to the smooth running of the game queues and your assistance has been incredibly appreciated.
As part of my retirement I will be transferring control of the discord server to current Council member SilverCrys, who will probably look at instating some moderators on it.
I will still be around on the discord, and I will doubtless continue to play occasionally. I am still keen to review games, and will always be open to discussing mafia theory and mentoring.
My sincere thanks to everyone I've worked with: councillors, secretaries, and forum mods alike. Special thank you to all the players and game hosts past and present: without you we wouldn't have our wonderful little corner of the internet, and I am so proud that I have helped look after it.
Here's to the future.
Eco.
I fear that it would be too much book-keeping and time consuming to be honest, as well as difficult to measure. It's already enough hassle to keep tabs on what games are ready to run and getting people to finish them, to have to hands-on review games and calculate their relative readiness and to deal with people complaining about being moved around every time doesn't seem like it would be worth it. If we push the best to the top, then it's just a rolling FTQ but worse because people have to have a game perpetually able to run but it might never fire - at least with the FTQ you have more awareness of time scales.
I'm in support. I think we ought to discuss how we go about populating the list. It seems straightforward to fill the top three slots are filled with the winners of the most recent FTQ (in order TBD based on who is most ready to run). But after that, do we just open up a first-come-first served list or give precedence to hosts who entered the FTQ but did not win?
For future FTQs, I suggest we maybe try and incorporate entries from all queues to pick a best game to jump all the queues as part of an event - say we run it once a year and combine it with an invitational player list/team event/cross-site advertising. This would be instead of just doing it ad hoc (how it used to be) since I don't think we really have the player the player demand to support occasionally running a large game on top of the queue structure. Obviously if we do get a surge there's no reason we couldn't run an ad hoc FTQ to fill it in.
Whether we would still need an FTQ at all is another question. It was intended originally to allow really exciting games to skip long queues (which we don't have any more) and to stimulate game design through competition (which we might still need?) - so maybe it can be retired as a regular for now and be run occasionally if we need to spur some creativity.
How do you propose we motivate this without replacements or modkills becoming any easier? I don't believe hosts don't strictly monitor activity because it is too hard or they don't care - they don't do it because the enforcement system we have is difficult to implement.
By "Queues" we mean the queues for the games themselves, which are basically empty. How can we schedule games when the only queue with more than ONE game ready to fire is the Micro queue (with two)?
With regards to activity, I think there are multiple factors compounding to produce our current situation.
So I think the problems with activity are a result of the smaller playerbase both allowing lurkers into games more frequently as well as making them more challenging to punish, while we also lack incentives to drive even participation throughout the Day and as modern metas drive activity to concentrate on the end of Day.
Solutions
Well, obvious solutions would be things like more strictly enforcing replacements/the probation list, or even baking modkills into games by default despite how challenging these might be (very). If done as a concerted effort (and people really made an effort to be available to replace) and the forum was ok with maybe firing games less often in exchange for better playerbases then it could have a real impact. However, enforcement isn't the only (or best?) solution - I think we need to try and restructure our incentives to participate evenly throughout the Day and avoid deadline scrambles. We could do this by having slightly shorter deadlines with unused time carried over to the next Day (to incentivise ending the Day early), or by mandating more posts during the prod period (5, say) - or even a cap on posts for each prod period to reduce spam.
I even consider a more complex approach where each few days the votecount is taken and people at the bottom are removed from the pool - this forces the possible lynches to reduce over time so the deadline scramble is at least between a couple of choices rather than the whole playerlist, and prevents people spending all Day on one-man vanity wagons. There's obviously issues (especially how to allow people to be re-added to the pool after being removed, and how much it changes the game from traditional Mafia), but maybe something more radical is needed to keep activity consistent throughout our long phases.
We could even offer actual rewards for things like hosting, or replacing in, or not getting prodded, or for being voted sporting/fun to play with. Every time you perform the desired action, you get entered into a draw and at the end of the year we give someone Curse Premium, or $20, or we interview them, or we let them play a game while being mentored by someone highly respected, or they get to design a game with someone highly respected, or they get auto-in to an invitational, or the nearest council member buys them a beer. There are lots of possibilities.
We also don't want it to make it more difficult to host games (e.g. requiring backup mods and setups finished fat in advance), as we have few enough games to run as it is.
Definitely could consolidate these, I think even right now we could:
This would take us to the Council thread, a PCL/Replacements thread, Hosting/queues thread, Resources thread, and the open sign ups.
Staff
I think saying Iso is our only hope is a bit disingenuous consider admins Cythare and Wildfire are both pretty familiar with Mafia, and we have our local mod Bur, plus other faces on staff like ExpiredRascals. But obviously I don't know what goes on behind closed doors, and I can only hope that if this discussion was actually serious then the MTGS staff would involve the Mafia community at large.
Game Scheduling
I really don't think this is a good idea. Fundamentally, I don't think the solution to "our sign ups take too long" is "ok, let's make the sign up period longer". While it'd be nice to know when a game will fire when you sign up, I think it is impractical to ask people to commit X months in advance, and you will have many problems with people forgetting/no longer being available but not outing/mods not being ready or falling through leaving people effectively signed up for a game they don't want to play, or even people waiting months and months only to get randomed out. The significantly simply solution to knowing when a game would start would just be a hard cap on sign up length.
I also think that scheduling when games will fire is a very difficult process: game length is variable, which means it will be very difficult to gauge how many people will actually be available to play when a game fires, and I think it will be very difficult to avoid periods where either a game could fire but isn't (because some people are free but the fire date hasn't hit), or the inverse and overlaps of games that strain the playerbase. Basically, I don't think you can reliably predict when a game should fire, and I think the resulting scheduling issues would cause more harm than good - it reduces flexibility in a system that leans quite heavily on the ability to look at what games are actually ready to fire and the availability of the player base to put a finger in the air and pick a game to open.
I think scheduled games would be an excellent idea for a very small forum that struggled to get enough players for a single game, so as to coordinate the few members it has to get something off the ground, but we are some distance from being in that position.
I think ultimately what we want to always have some sort of sign ups open and for games to fire quickly. Perhaps the problem is we do not have the numbers to support this policy: when we hit current capacity we either have signups which take a long time to fill (usually requiring an ongoing game to end), or we just have no sign ups open at all (and then people just walk away because there's nothing going on). This is sort of the problem the PCL was supposed to fix, but it clearly hasn't. Scheduling games would partially fix this by ensuring that games fired at regular known intervals, so that people when to come around, but at the cost of flexibility and being worse at responding to demand or playerbase utilisation. Deadlines on signups would at least force the next game to refresh until it hit one that was exiting enough to bump demand up to actually fire, but risks skipping over games simply because the playerbase is at capacity (which I think is what has happened to Osie's Micro - people are occupied with the team games and Shadow's Normal, so demand for any new game is low).
-MTGS user numbers as a whole more or less recovered from the Curse takeover in about a year, whereas the Mafia sub-forums have continued to languish. Not sure on the MTGS user numbers at the moment, but forums in general are suffering stiff competition from reddit.
-Mafia forums seem to be doing just fine: MU is very active, although I think MafiaScum is having some issues (but that's heavily cultural in that they have MTGS deadlines with MU-level activity which is causing absolutely massive games filled with low-content posts). MTGS is very much struggling to recruit from MU because despite our imaginative and usually well balanced games, the deadline length is something people really don't want to endure - MTGS is very unusual in the aspect of our deadlines.
So I don't think our decline is necessarily a consequence of a decline in general mafia interest, forum use, or MTGS itself. MTGS is doubtless no longer at the height of popularity it once was, but it is still proportionately far better off than our subforum.
On the rolling sign ups, I think it definitely has some of the drawbacks you mentioned: games can take a while to start, and people might just stop checking back if the same game stays open for too long. But if we are going to shelve games that don't fire within a time period, we have to be very sure that we are ok with abandoning our historical "everyone gets a go" policy, and also avoiding disincentivising people from making games because they think "what's the point of all the effort if it won't even get played?". We would also need a lot more games in the queues or the faster turnaround will stall out quickly. But at least the shorter queues make it easier for someone to have a game fail to fire and then make changes and get another shot.
An alternative might be having a rolling queue for each game type - at least then a stagnating game only blocks one queue as opposed to all of them, and it would allow us to more easily gauge what the popular game types are. I know the Player Contact List sort of fulfills this function, but it doesn't really in practice. Downside would be a lot of sticky posts (maybe a subforum for them with a BIG OBVIOUS THREAD pointing at them?).