Improving the Circus Maximus Queue

Queue improvement


  • Total voters
    26

a fairy

is a Tournament Directoris a Site Content Manageris a Community Leaderis a Community Contributoris a Smogon Discord Contributoris a Contributor to Smogonis a Top Smogon Media Contributoris a Top Dedicated Tournament Hostis a Social Media Contributor Alumnus
Community Leader
Hiya

Currently, mafia-styled games held on the forum (OC, NOC, Popcorn/Viva/multifac etc etc) are all lumped in one mostly-FCFS ordered queue, where no two games run simultaneously. We didn't even need a queue for a while, then Revivial picked up steam rather than slowing down (can you believe JOAT10 was hosted because we had nothing in the queue at the time) so we made a queue. Then OC games started, the mausoleum emptied and all the boomers revived themselves to play again, and judging by the queue, y'all don't seem to be just passing through!

The queue is not working, not well at least. It can be improved, and there's no reason to shrug and keep our NOC-structured queue when we can accommodate the distinctions. Here's a poll and discussion thread for how to improve our queue, I've thrown in some logical ideas I had at the start of the poll, I'm not against adding more possibilities.

Votes are fully free - you can vote for any option you want, you can see what others have voted for, and you can change your votes if you want. The poll currently doesn't close. Not all of these options are mutually exclusive from each other, though some many contradict - it'll just give mods idea of direction forward.
 
I voted the option of separating noc and oc queues because they feel like complete different games and some people might prefer one type of game to another one. I understand that people that love both might not like the idea but at least alternating between then is a good idea because the queue currently goes from 3 oc in a row to 6 noc in a row to 4 oc in a row again. This option is probably the quickest way of making the queue thinner.

My other vote was the requiring games to list their design and desired player count in queue. Games higher in the queue shouldn't be a surprise to the player at least in basic things like "role madness 15p". Games lower in the queue are fine with little to no info to let the hosts decide their game.

I don't like the making the hosts give game length because I feel like it's too hard and miscalculations add up the longer the queue goes.
 
My two cents: first of all, I agree that the queue seems to be having some problems at the moment. Including MM3, the current queue, as of 4/1, is as follows:
OC (MM3)
OC
OC
NOC
NOC/OC
NOC
NOC
NOC
NOC
OC
OC
Open NOC slot
OC
OC

As we can see, it is very lopsided. Several OC games in a row, followed by a huge span of NOC games, and then another span of OC games afterwards. I think this creates two related problems:

1) Repetition. While I'm sure all of the hosts have put in hard work to make each of their respective games as unique as possible, it's no secret that OC and NOC games inherently play differently. Repeating the same type of game entirely during certain spans (assuming the games are run consecutively) can leave those who only like one of the two game styles without a game to play for several months.

2) Staleness. In a similar vein, playing the same kind of game over and over can quite simply get old. This is more salient when the same OC/NOC format is repeated (ex.: 3 OC 1v1 games in a row) but can still be a factor even if the structures differ (ex. a multifaction OC followed by an 1v1 OC)

I think all of fairy's proposed solutions would address this issue and like most of them. The only one that I have some reservations about, however, is ironically the first option: to run games simultaneously. There is a lot of overlap between the NOC and OC playerbases, and running games simultaneously may prevent players that desire to do so from participating in both.

Personally, I think simultaneous games can be implemented with the caveat that the only games that run simultaneously be "small" games. In other words, when a "Big" game is scheduled, it should be the only game running. This will allow a) everyone who wants to play in the Big to be able to, and b) allow those who don't make one of the "small" games to join the other.

Finally, my discussion above mostly pertains to what I believe should be done on a go-forward basis. To fix the currently lopsided queue (at least pertaining to the next 2-3 games), I would ask the mods to reach out to the respective hosts to see if they would be agreeable to a reorder of the queue. Generally, I feel like no more than two of the same type of game should run consecutively.
 
Last edited:

UncleSam

Leading this village
is a Forum Moderator Alumnus
Hard agree on separating the queues. The player bases are not that similar and I just don’t see a huge need to split them up outside of when the OC is a big game. Those can be special cases and handled separately / on an individual basis.

Also agree on listing the formats so that players know what to expect.

If one queue or the other starts to suffer in terms of signups then maybe we re-think things. But it definitely seems like the no-brainer option is to allow for two games at once for the time being.
 

Celever

the bear becomes the ringleader
is a Forum Moderatoris a Community Contributor
Moderator
I've spoken to both a fairy and Tommy about this in DMs already lol, but I for sure think we should be alternating NOC and OC. I actually think the game queue is otherwise fine.

Splitting off NOC and OC completely is not a sensible idea. OC and NOC are both games that eventually result in burnout, so alternating between them either offers breaks entirely for people who don't want to play one or the other, or it offers variety for people who want to play every game. The playerbases do, ultimately, have a lot of overlap, and that's actually a good thing. OC mafia has seen a great resurgence, but I don't think it would maintain so well if we started chaining OC games for months upon months upon years, which is what splitting the host queues demands.

Running a NOC and an OC simultaneously will always make one game suffer in some way. Either the playerbase gets split, making a big OC game like what we just had totally impossible, or players try to double up but inevitably end up focusing on one game more than the other, leaving a low activity or dead slot in one game.
 
I feel like 2-3 NOC:1 OC would probably be a good ratio to lean towards, given the differences in game length between the two.
The main reason I don't fully advocate for a completely separate queue is that, as some others have noted, the larger OC games akin to MM3 and MRM would suffer from it.

For instance, I didn't wind up playing in MM3 in part because I knew there was a mash coming up on MU that I knew I'd want to suffer play through, and, since said sorts of games tend to require large amounts of commitment, I knew I wouldn't be able to focus on both at the same time. A similar effect, albeit a bit weaker, would likely happen on a larger scale should the queues be separated.

The main con of larger games, to me, is that if your slot (or the game's activity) dies, you don't really have much to do for a while. But that issue could easily be lessened by opening the next game's sign-ups a bit earlier than normal in the case of those larger games.
 

Yeti

dark saturday
is a Community Contributor Alumnus
imo the people who will play larger OCs are very unlikely to play NOCs, so i think depending on the game size NOCs can run concurrently to OCs without cannibalizing the playerbases much. if it's a Big-proper then it would probably need to run solo.
 
In the past we were able to support separate queue for OC and NOC, and hosts and moderators were able to react to burnout and adjust as games finished by moving things around in the queue, taking breaks, or having everyone focus on a big instead of trying to run a NOC concurrently.

I think that having separate queues and making anticipated player count and format for each game public are the most important things, and almost everything else can be flexible and adjusted based on perceived demand and timing considerations
 

Duskfall98

Votecount Specialist
is a Forum Moderator
Moderator
I think in reality there’s two player groups currently from my perspective:

NOC + OC players
Just OC players

With some limited exceptions. So far the two big games have had probably close to an even split of those two groups making the game, zorbees current signups is roughly 2/3 the people who play NOC. While there’s definitely a large group of OC only players, these are also probably the group that aren’t planning to play every single game constantly back to back.

So I definitely do think to have two full queues is going to effect the numbers of both game types, for NOC most of their players are willing to play both but also for OC half their players want to only play OC, but not regularly so it does seem to me that regular OCs would be relying on the same player group as NOCs.
 

Ditto

/me huggles
is a Tournament Director Alumnusis a Forum Moderator Alumnusis a Community Contributor Alumnusis a Contributor Alumnusis a Battle Simulator Moderator Alumnus
Note: This post and these opinions are that of a barely functioning walking OC rotting corpse as a fairy so delicately put it and someone who does not see themselves playing much NOC.

I voted that we should separate the queues for the same reasons as others have said, so I won't really repeat those, but I did want to bring up 2 things.

1. This resurgence of old players could easily lose steam and momentum. This could either be compounded by lots of games that leads to quick burnout, but at the same time having too much time between games might also lead many to drift away from the site again. I don't know which is more likely, but I don't think there should be some grand restructuring of systems for something that might not last.

2. People are talking about NOC players and OC players and crossovers, but I do think there is a subsect that was around in the past and is likely to arise again. Big OC players and OC players. There were many people who would only play in the Big OC games and wouldn't join smaller ones, and a good amount of the people that returned I think fall into that group. Smaller games require less players, but I also think they will draw less people to sign up. I would feel really bad for all the people who are putting effort into games that are in the queue and then have all that work be for naught because of player drop off.

Smogon Mafia is in a weird state right now and it's great that there's been this resurgence of older players, but I think expectations should be tempered.
 
I think in the past there was even a separate big game queue to stagger those specifically and ensure the other games could fold into that schedule.
 

UncleSam

Leading this village
is a Forum Moderator Alumnus
Yeah actually thinking about this more I think we should just have a big game queue separate from OC / NOCs that are smaller.

The small game queue I will leave to the discretion of the moderators. If games are getting way too many signups than imo the queues should be split, if not then maybe one queue is fine for all smaller games (25 or less).

The big game queue should be separate and imo should have about a month between games, if not a bit more (and should be planned around the smaller ones). Probably this means 1-2 smaller games between big ones, give or take (with either a smaller NOC and a smaller OC running simultaneously or one then the other).

I think it’s fine to run 2-3 smaller NOCs (as opposed to smaller OCs) back to back. I don’t think it’s fine to run five smaller games (regardless of NOC / OC) back to back, because those are where the real player base differences come in.
 

StupidFlandrs48

Looking for myself. Have you seen him anywhere?
is a Pre-Contributor
Despite the fairly narrow scope of my experience, Mafia Mafia 3 made it very clear to me that we can't treat that kind of thing as the same game compared to NOCs with 1/3 of the scale or less. MM3 felt like an event where the host's imagination went wild with crazy ideas that tried to shake up the game of mafia in a lot of ways, and while that experience is definitely appealing, it shouldn't be the baseline for a new player due to the complexity and time commitment. The current system of treating all games as the same and only holding one at a time will not work mainly for this reason: it deprives new or not-fully-active players of a good entry point for weeks or even months at a time, which is just unacceptable.

Big, ambitious OCs like MM3 should be given respect by being removed from the main queue and scheduled on their own, with only light considerations for overlap with NOC and small-scale OC timetables. Having one small game happen at a time, but saving big stuff and complex small stuff to be run alongside the main queue sometimes, not all the time, is probably the best way forward to keep things moving and improve the variety of experiences this forum can offer. One game at a time for two months, and one game running alongside another special game every third month or something like that.
 
Ideally, split the queues and build in a buffer in the OC game queue to prevent burnout.

It appears to me that that there are two player bases, as Tommy says. An NOC base of people that tend to be more regular and willing to invest more time into games that are also willing to join OC games, and an OC base of older players that prefer less frequent, larger games.

My (ideal) preference would be a world where the Smogon mafia community can support multiple concurrent (2-3) NOC games as well as have an OC game running about half the time. I say this as somebody that wants more games to join tbh. I don't really get burned out by mafia that easily. I understand I'm in the minority on that.

I think a more realistic outcome is having at least one NOC game running at all times and an OC game running half the time. I think this would be a good solution
 
yes, let's spread out NOCs and OCs so the same gametype isn't running 5x in succession. also move the queue to the main subforum. the 3 seconds it will save will improve everyone's qol thx

i think we just need to have the mods facilitate the queue and get signups going for smaller games in quicker intervals. for example a noc game with 3-7 players left is likely few enough where the next game's signups could get going. mods can follow games to get a sense for when they are wrapping up and get the next game signups posted so there's not too much deadtime. if people start getting burned out and less interested in playing then they can make the decision to slow games down too.

i was thinking you could maybe get away with a noc and oc running at the same time but personally i probably wouldn't have time to do both so perhaps not a great idea. but i do think we can overlap the games a bit more than currently.
 
As someone who can only play NOC, please seperate cues.

I have no input about how OCs are run, but NOCs are in a really good state rn. I think we can run them cocurrently with OCs.
 

Nuxl

new message from your psychologist
is a Community Contributor
mostly noc player reporting in, i second/third/whatever
- alternating the NOC and OC queues
- starting games near the tail end of a hosted game or at the end of one

Don’t really think we have the playerbase to run more than one noc + oc atm and would heavily recommend not multitabling games (or having more than 1 game once) as a whole. think it will

1. increase the rate of burnout (eventually people get sick of playing mafia over and over and need a break) which affects community retention between games.

[I also think spamming the same game type over and over will hurt playerbases of one gametype as a whole (OC mainly, either it’s too much OC or too little when NOC queue starts)

but 2. a (possibly selfish) game quality reason where people multitabling noc + oc (or say more than one noc like m2h said) could worsen the quality of either of their performances in the games depending on their commitment to finding their win condition in both games

thnx guys for continuing to sign up and host games, it’s been awesome to see circus get spark back
 

a fairy

is a Tournament Directoris a Site Content Manageris a Community Leaderis a Community Contributoris a Smogon Discord Contributoris a Contributor to Smogonis a Top Smogon Media Contributoris a Top Dedicated Tournament Hostis a Social Media Contributor Alumnus
Community Leader
Here's what we're doing at the start. This may change in the upcoming months as we further evaluate how the changes affect interest and scheduling.

Firstly, we wont be running games simultaneously. It's our perspective that we have two groups of players currently - people who play whatever game they see get posted pending their available, and people who want to come around for the larger OC games with breaks in between. We think that running games simultaneously is actively detrimental for the reasons stated above by other folks contributing to the thread. However, it also means our larger OC games will come up faster, which we suspect will cause some of those folks who come around for those to burn out.

We do think, however, that NOC and OC games are distinct enough, and should have some moderator guidance to ensure that we don't have surpluses and droughts. While we can't guide people who submit games, we can build a schedule using subjective judgement and a general guide on what a healthy schedule might look like.

Our new schedule is based around the larger OC games being proposed. We have larger games, roughly in the 25-35 playercount games, which are more difficult to fit into, but can probably work alongside another one of its kind well enough. Then, obviously we have the capital "B" Big games, the 50ish playercount games (see Viva and MM3) that should stand on their own to avoid major burnout for things of those sizes.

We are going to create cycles of games, starting now moving forward, based on historical trends since Revival begun and what we have in queue currently. Our "normal cycle" will include 2 NOCs, 1 smaller OC game (think zorbees' current game), and then a capital "B" Big game. Our larger games will fit into the OC slots, but we're going to avoid a larger game in the same cycle as a capital "B" Big game, when possible.

With this structure, games will need more information when going into the queue, which we'll ask for some patience as this is a new requirement and a large number of games already in the queue are light on information from a previously much more informal approval system. However, the information will be flexible, as only NOC/OC and the expected playercount is taken into account when scheduling, rather than percieved game lengths.

Additionally: this is not set in stone, and is based off of the current queue, what we've seen brought up by people, and historical Revivial behavior. We're going to be flexible based on what gets submitted, what we think will be the best for the community's interest, and what the hosts are able to do schedule-wise. The cycles are an ideal balance in our eyes, replacing the broken FCFS system, but this is just our baseline that we will adjust and make exceptions to as needed.
 

Users Who Are Viewing This Thread (Users: 1, Guests: 0)

Top