Mafiascum.net

I wonder what the unique posters would look like if we limited it to mafia forums as opposed to the forum as a whole. It could be that those ~100 lost over the last year were all from mafia. and that it went from something like 450 to 350 there instead of the 550-450 drop over all.

things have just gone downhill since i stopped playing. sorry mafiascum.

I'm new and loving it here, so.



just an encouraging thought in a somewhat depressing thread

I'd say it's worrying but not depressing. The first step towards fixing anything is being aware of the problem rather than being complacent & denying it.

You know before I thought that Modbot would be something to help mods run their games, but maybe something more autonomous would suit shorter, presumably more numerous games like marathons and blitzes?

there'd still be someone behind the wheels in case of trouble, but maybe that someone would be the queue's listmod or a team of people

Unless we're talking about open games, the games in blitzes still need to be designed. Although I suppose for something short like that you could come up with some sort of formula to randomize the game and because it doesn't last as long if a few games are unbalanced or broken it's not as large a deal.

But I absolutely think that automating large portions of the moderation process, those that people typically don't really enjoy (vote counts, prods), is good.

The biggest issue I see about automation is that it's probably harder to automate replacements, particularly before we have automatic sign ups.

yeah was thinking at at least minimally we could design/use open setups like the kind we use for newbies where setups aren’t the same between games but are still constrained and thoroughly validated for balance

In post 953, Kison wrote:I'd say it's worrying but not depressing. The first step towards fixing anything is being aware of the problem rather than being complacent & denying it.

Oh I think everyone can agree there's a problem and nobody denies there is one.

It's just that if you ask a hundred different people what the problem , you'll get literally 100 different answers (okay, maybe not--some of them will share one answer: "not a damn clue"); nobody has any level of agreeance with one another on what the problem actually is, with everyone thinking it is something different if they put thought into it at all.

Probably doesn't help that other sites are actively recruiting from this site, to be fair.

In post 955, zoraster wrote:Unless we're talking about open games, the games in blitzes still need to be designed. Although I suppose for something short like that you could come up with some sort of formula to randomize the game and because it doesn't last as long if a few games are unbalanced or broken it's not as large a deal.

But I absolutely think that automating large portions of the moderation process, those that people typically don't really enjoy (vote counts, prods), is good.

The biggest issue I see about automation is that it's probably harder to automate replacements, particularly before we have automatic sign ups.

Even if each game still requires a human moderator, automating much of the busywork would make it easier by allowing a moderator with less free time to do the job,

I can't get to the site every day, so if I'm moderating, that means that the deadlines need to be long enough that a random delay in a vote count or a flip or a night end or the like won't be disruptive. (Either that or I need a co-mod.) So I couldn't plausibly moderate a Blitz by myself.

If setup design is a problem, we can work around that by a) expanding our catalogue of Opens still further and b) making more use of the NRG's predesigned setup queue (e.g. by allowing Theme mods to take a setup from there and fit a theme around it). Designing a really good setup with unusual mechanics can take months to years, but just quickly designing a reasonable normalish Closed game that doesn't stray too far from established norms can be fairly easy.

In post 958, Nexus wrote:Probably doesn't help that other sites are actively recruiting from this site, to be fair.

I haven't seen/heard of this, they are?

i think i got a pm from my good friend james3

If you receive a PM from a banned user please report.

In post 11, chamber wrote:Thoughts on just making mafblack default?

I use the sepia layout. The black doesn’t show up well on my iPad.

i think mafsilver is more welcoming

It’s okay but I find mafsepia to be the easiest on the eyes.

In post 953, Kison wrote:I'd say it's worrying but not depressing. The first step towards fixing anything is being aware of the problem rather than being complacent & denying it.

So what is the next step?

There's an absolute ton that can be done. Fewer people are coming in likely as a combination of these reasons:

1) Very little has progressed in the past 5 years. This has allowed other sites to become compelling in ways we aren't, so we must catch back up. I like moving fast, so we will catch up no problem.
2) Lack of mobile support. Wasn't as important 5 years ago as it is now.
3) SEO decline. Because of the above mostly.

It is very clear that the following needs to happen:

1) Upgrade the board. This provides a ton of benefit, primarily improved search performance & superior mobile support. This is slated to be done sometime in July.
2) Improved game signup queues. Game signups can be made more navigable & visible(right on the index page if we wanted to). But it also allows us to begin tracking important data to do #3. Once the upgrade is done, the plan is for chamber, myself & whoever else I can persuade to wrap this project up. I'm going to take over a queue to use as a testing ground.
3) Game automation. Doubtful we'll ever automate everything, with some theme games having super weird mechanics, but if we can automate the most cumbersome aspects of moderating it will (1) improve the playing experience and (2) ease the strain on moderator supply.

The above path is a no brainer. It makes the site a lot more compelling and should organically draw people in. It's just scratching the surface, though. I (and I'm sure others) have an endless supply of ideas for things that could be built or improved upon. So my #1 focus is really just myself programming & giving other programmers what they need to keep moving. Everything else is secondary for now.

A lot of non technical things can be done as well & I've suggested a few I think are incredibly important like SEO content. I think there's a lot of other stuff. I'm impressed by how excited people are for something like SUPP. It'd be great to have more events like that, but they need someone to come up with the idea & run it. If we have a flaw as a team, I think it is that we sometimes spend too much time on things that really don't matter all that much.

How much of the reduction is banned users?

Is that a serious question? We lose like 100 users month to month because of turn over (some of whom come back later). How many banned users are there? Less than 10?

Not really...
What is the time scale of the graph?

those graphs are from 2002 until the end of May, each point being a month. They show net loss and gain. When I say the 100 number, I'm talking about users that post in month x but don't post in month x+1.

Here's how I'm interpreting the data.

The first graph is how many unique id's post per month, and shows a long-term trend of -2 users per month (loss of roughly 150 active over 67 months) with a slightly steeper shorter term trend.
Second graph is how many new uniquie id's registered and posted per month. I didn't calculate the trend line but it's down a little too.
Putting the two graphs together shows there is a steady turnover of around 30/450 or a little under 7% of membership?

I'd call the turnover the issue, from just the graphs it appears the majority of new users are not staying and some old guard fade.
It's possible to go deeper by looking at join date and last post date (within the same filter) to see a scatter chart for duration of stay.

I'm going off of more data than just those graphs when I say we lose about 100 users a month. We gain back a good part of that with new registers and returning users to see the net decline that graph shows. We both agree that the registration numbers being down, and to some degree the rate they stay for, are the key issues.

In post 972, davesaz wrote:and shows a long-term trend of -2 users per month (loss of roughly 150 active over 67 months)

I don't know why you chose that period, but its more like 250 lost during it, not 150.