i think it's time to try player caps.

[Flourish]

Echoing all of Zamu’s points. Also we’ve spent several years (at least on the admin side) intermittently considering player caps and deciding “actually no let’s not, it’ll make x happen and be bad because of y and z”, and have seen the results of that, so why not actually try player caps on our main servers for real, so we can actually have concrete data on how player caps affect the servers and the player experience.

[Fuppy]

I'm in support of at least giving this a try... 3 has had an extremely high pop during most days for a good while now. As Zamu said, RP basically becomes impossible at a certain point because so much is happening and you're basically limited to shallow interaction if you don't want your brain to melt. I'm hoping this would shift a good amount of the pop back to goon1, as I miss having goon1 be the server where I can just turn my brain off as things explode around me while I mop up blood.

[LeahTheTech]

I'm open to the idea of population caps but I think:
- 65 is too low
- It should only count players in game otherwise people are going to idle on the title screen to "save their spot" and we're going to have 50% idlers/observers taking up the cap
- Classic should either remain uncapped or have a much higher cap as it works a lot better at high (even 100+) populations

[Emimiyu]

Tbh, idk if it's a realistic solution, but yeah, the pop distribution sucks and it is not just RP vs Classic thing, it's just that you have to pick in between a very empty station/not satisfying pop and the servers where everyone is hanging out. People who prefer high pop goon 1 moving to 3, people who prefer lowpop on 4 thinks 4 pop is too extremely low they would rather adapt to highpop than staying in atlas with the same 6 people. It's just suck you have to pick between empty or very crowded station.

Also, regarding 4 low pop, maybe it's just rose colored glasses, but I feel like goon 4 used to be have higher pop, at least on days/hours when there are more players. Some pople moved to 3 because some issues relating metaclique in 4. Though, I heard that the issues are being handled, it still feels like population below 15 is too empty (not to mention it already includes observer) for people to come back. It's also again, sucks to pick between 80 and 10, people would rather just pick where they can interact with more people, even if it's not their favorite gameplay experience.

[Cleaverwolf]

Big support for this idea, let's give it a try!

[Silent Majority]

I am personally against it because I worry it will just cause a deduction in population.

As it stands this is sorta like the 2 thing, where there are people that will want to play with whoever they want to play with, which could cause mass depopulation of servers whenever the 'desired' one opens up.

While I can see there is a good idea behind this, and I understand it's an issue, I feel that altering rules to encourage/discourage certain behaviors to increase .

For example, if you wanted to cause a larger shift to 4 you could simply change the antag balance to have 4 roll more antags like 3 is now, and change 3 back to the old system.

You could have certain days be 'extended' days for either of the rp servers.

You could set one server to be the 'newbie friendly' one, even give a pop up explaining that.

I feel that there are some other ways to try to fix the problem, and I hate that it IS a problem, and while I'd understand a possibility of a population cap I worry this will just reduce overall players and not fix the issue.

---

ALSO this does potentially create a scenario where both rp servers could be capped and then prevent people from joining rp.

[Wrench-1]

I've always been open to the idea of a Pop Cap.
Honestly, I'd say best to trial run it, and if it doesn't work out and isn't beneficial for the health of the server, then we can roll it back.

No harm trying, honestly.

[meaow589]

Cap is reasonable but the number 65 is unsure for me the cap should relate to each job available +5-10 assistance-cyborgs. Which should around 40-50 and the number which more fair than 65-35 to 40-40 /50-50(in over crowded periods)and make server very balance no bias to choose server to play.this can deal with people who always choose high pop server since 2 server are the same

[Lefinch]

I lied, one more post, but more about how pop cap can be used to give those who are not filtered better opportunity. It is probably an idea that would need some thought and refinement, but I think at the bottom line it would allow a cap to function with more minimal interruption of player experience.

Here's how the cycle of a cap might play out if it was put in. This assumes a lot of things like Leah's point avoiding lobby counting towards round occupants (and I assume observers, ghost drones, critters etc are also not part of the count, as the issue isn't so much server load as RP capability, and those entities GENERALLY aren't part of that equation) This also doesn't focus on late-joiners, where I assume a cap is a simple case of "All slots filled"

>Round start, the lobby readies X amount of population
>I assume either a cap either prevents you readying, or "lotteries" those ready for round join.
>On Round begin, those who could not ready or did not win the lottery get some kind of message. At high population, as Zamu mentioned earlier: this could be anything from a few people to say, 20/30+ people.
>They -could- now go join an ongoing server game as latejoins. This is great as they know lots of others are going to do that, but precludes them from antag or "random job draw" options. Personally, this doesn't bother me, I'm a happy late-joiner, but I also think that what could hypothetically happen is more observers rather than players. That's a hypothetical, and I wouldn't want an idea to be shut down on guesswork, I'm only bring it up here to underline what might help my suggestion below.

Alternatively

>Round start, all is the same as above
>Simultaneously, and this is the thing that's going to make this not work immediately if it's not feasible or for many reasons people go "we can't do that" (I don't know, I just play here, but no harm in sharing an idea) a server is "readied" for the those who do not make it in to the first all.
>Those who do not make it into the first draw for the server are either direct transferred, or if not possible, offered a clickable token to join THAT starting server.
>Ultimately, those looking to start a new round now will always get that opportunity, it may just be in server 1 which is full, or server 2, which is less full, but either way their intent and experience of getting a fresh go is achieved.

I have -no- idea how remotely possible this is. I have also not considered many possible nuances that would make this a bad idea, but I want to at least offer something that I've had a little personal experience in elsewhere: Handling overflow to a populated player that still replicates as best possible the experience they were looking to have. Feel free to ignore if this is all very laughable. It worked for us in a MUD environment running on a servers jokingly ascribed to be "TI-89s daisy chained together" and it simply may make no sense here.

[swinglow20]

I had a dumb idea, but what if we put a population cap on certain days. Like, Monday through Friday the pop cap is around 65, but weekends has a higher/no cap?

[Cal]

I don't see why not testmerge it and see if it helps. The populations are very odd right now with one server having way more people.

[Zamujasa]

I'm open to the idea of population caps but I think:
- 65 is too low
- It should only count players in game otherwise people are going to idle on the title screen to "save their spot" and we're going to have 50% idlers/observers taking up the cap

65 is a number I came up with based on a rough figure that 35-45 people is ideal, and some of those will be people who died/observed.

Earlier, we had a round with about 80-90 (and at points, 100) players connected. The round generally had 70-ish people alive, with 5-10 people dead or spectating. Obviously that's a sample size = 1 moment, but it's the kind of experience I've seen before and about where I'd set it. Maybe 70.

As for people who idle and observers, we already do, and that's why the sludge factor is there, to give space for people who aren't necessarily playing but are still contributing or active in some way.

As for people idling on the title screen, I am 100% fine with giving them a "you have 60 seconds to either join the round or get booted and lose your spot" thing.

Right now (11 PM pacific) on 3, there are 39 players online, and a whole 13 of them are on the title screen -- a solid 1/3rd. Hell, I've left it on overnight by accident. We can always try a higher cap of 70 and see how it works, too.

- Classic should either remain uncapped or have a much higher cap as it works a lot better at high (even 100+) populations

The only time I'd be okay with limiting classic is if it's regularly hitting 150+ or something, and even then I'd be fine with saying "if you have over 150 players the tick lag is locked to at least 0.6 or slower"

[Chasu]

It's definitely worth a try when the time's right

[Tyrant]

I think this is a good idea honestly. The points you raise about RP breaking down are definitely real.



My only concern is how this might translate into people fighting for a spot to play on goon3. Meta shenanigans and locker naps in order to hold a place so they can play later or something.



Would it be possible to put the limit on the amount of living players in round (e.g. 55 or 50), rather than on the number of people in the server? That way people could still spectate if the 3's full. Not sure how feasible that would be to implement but I think that could solve some of the issues of observers 'taking up slots'. I know you talked about raising the limit and possibly kicking idling title screen players after a set time, which also works. Then again, those players would still appear connected on goonhub... pros and cons to both implementations.



I'm kind of curious about the logistics of a player cap as well, the actual mechanisms of it. What happens when someone tries to join anyway? Do they get shunted off into goon4 or does a pop up appear? I've really only played here so I've never encountered a player cap before.

What about if both 3 and 4 fill up at peak hours? What happens then?

These have probably been considered but I thought i'd bring them up. I think that this would overall be a good thing for RP even if some people complain. Much more controversial things have gone through, after all. As long as a PR introducing this is well thought out and covers all the points brought up in this thread, I don't see why it shouldn't be implemented.

P.S.

I am personally against it because I worry it will just cause a deduction in population.

Isn't this... the whole point? To make it less chaotic? That's the whole idea.

[Zamujasa]

My only concern is how this might translate into people fighting for a spot to play on goon3. Meta shenanigans and locker naps in order to hold a place so they can play later or something.

This is a culture thing. It will be something people have to look out for.

That said, I don't really see this as a bigger problem than what we have now. Even if 4-5 people do this a round, which seems unlikely... I dunno. If they sit in one place and don't move the whole round that'd be suspect, but that's the sort of thing that (if it becomes a problem) can be checked for. If you don't do something every x minutes, any online admins get an alert.

At that point someone would have to be getting pretty deep into the weeds for holding a spot, like macros or otherwise just being persistent.

Would it be possible to put the limit on the amount of living players in round (e.g. 55 or 50), rather than on the number of people in the server? That way people could still spectate if the 3's full. Not sure how feasible that would be to implement but I think that could solve some of the issues of observers 'taking up slots'. I know you talked about raising the limit and possibly kicking idling title screen players after a set time, which also works. Then again, those players would still appear connected on goonhub... pros and cons to both implementations.

the problem with dynamically counting living/dead is that it's very easy to go from one to the other. someone dies, then gets SRed shortly afterwards, or cloned, or uses a ghost critter spawn, or is borged, or...

setting a flat limit simplifies a lot of the logic.

I'm kind of curious about the logistics of a player cap as well, the actual mechanisms of it. What happens when someone tries to join anyway? Do they get shunted off into goon4 or does a pop up appear? I've really only played here so I've never encountered a player cap before.

that's up to us. if we keep it rp-only, we could redirect someone to the other server automatically and show a message; or allow people to connect but not actually join (i.e. only allow them to click on something that redirects them with a message, rather than automatically). i think giving people a visible "queue" type thing would lead to people just sunk-costing themselves to it given how long it takes players to disconnect, typically. (plus the grace period for rejoining.)

What about if both 3 and 4 fill up at peak hours? What happens then?

then we open up 5. or we leave 4 uncapped for the time being (until something like that happens). this is a good question that definitely deserves more attention if the tide really picks up, and might be a good reason to consider a higher cap; e.g. if both servers are full, then the cap grows to allow more, but otherwise tries to keep them balanced; sort of like how tf2 let you join an arbitrary team as long as that wouldn't unbalance them too much

These have probably been considered but I thought i'd bring them up. I think that this would overall be a good thing for RP even if some people complain. Much more controversial things have gone through, after all. As long as a PR introducing this is well thought out and covers all the points brought up in this thread, I don't see why it shouldn't be implemented.

if the point hasn't been discussed in this thread then i dont think you can consider it being discussed, if only for the sake of ensuring everyone here is on the same page. i said this in discord, but to repeat it here: shit you say in discord will not be remembered. it will be forgotten. it is impossible to find anything in a sea of random crap.

if you think something is worth discussing or mentioning then the best thing you can do is to bring it up and discuss it.

P.S.

I am personally against it because I worry it will just cause a deduction in population.

Isn't this... the whole point? To make it less chaotic? That's the whole idea.

in this case i assume they meant that it would have a net-negative effect overall and not just on 3. which is correct, imo, at least at first.