/tg/Station 13

Servers being at near max population due to the influx of new players has made the exponential damage caused by every extra traitor in a round more obvious.

A single antag can decimate the station using 0 TC using tools on the station, and having 14-18 of them during a 90 player round is not really something that can be managed.

The current server config has the traitor scaling coefficient at 5, and changeling scaling at 7. The coefficent works using the formula

Antag # = Server pop / Antag coefficient

To put it in perspective, the server's current security scaling coefficient to determine round-start sec officers is set to 6. Meaning assuming every security post being filled (which is rare) there will always be more traitors than sec on a traitor round.

Consider raising the antag coefficient for longer rounds. or adding some sort of cap to the formula so rounds dont immediately go to shit at the 20 minute mark.

You could break it into two coefficients if you wanted to. One for normal pop and one for high pop, that line to be arbitrarily set. Then you could have current antag coef on normal pop and a smaller one on high pop.

To balance that you could have more mid round antag spawns on high pop. More late join traitors, higher chance of a blob spawn, pirates, xenos, etc.

You could break it into two coefficients if you wanted to. One for normal pop and one for high pop

No! this is cancer. there are at least 15 configs we have that are such bullshit population based configs.

We need to cut the cancer out at the root: ConfigMappings.txt. An idea I had for a config that instructs the server to change configs when certain values get above or below a threshold. use it on population, server load, admin count, etc.

Just throw more math at it.

Instead of Y=x/5, we could have something closer to y=x/(5+0.05x) In this way, we get a smooth curve that tapers off a bit as we reach higher player counts. The coefficient scales up with player count.

Assuming the antag code rounds up (which I suspect it might, since you still get one antag when there's only one player, and that is at a y=0.2); The exact formula can, of course, be changed to get the exact curve we want.

Personally I think traitors should just have their more destructive items removed from their arsenal. The problem isn’t how many traitors there are but more so how many syndicate bombs can now be bought since there’s now a fuckton of traitors

Yep. Traitorlings is a terrible mode for the reasons above.

>Traitor bombs everywhere and plasmafloods literally every traitorling round

>Ling kills the head of staff so the shuttle isnt being called

this isnt even fun for the antagonists the majority of the time. Such a hellfuck of a mode

Edit: To more tie this into the thread theme. I feel like this change wouldn't work. The reason the shuttle gets called is not because there's no sec to deal with all of these antags. It's because these antags are simply too destructive that sec no longer has the reigns to deal with them. Thus the shuttle gets called. It'd be better to just remove it as the best alternative. The immortal antagonist, and the Boom Boom your station is doomed antagonist are just too powerful of a combination.

Shadowflame909 wrote:Yep. Traitorlings is a terrible mode for the reasons above.

>Traitor bombs everywhere and plasmafloods literally every traitorling round

>Ling kills the head of staff so the shuttle isnt being called

this isnt even fun for the antagonists the majority of the time. Such a hellfuck of a mode

Edit: To more tie this into the thread theme. I feel like this change wouldn't work. The reason the shuttle gets called is not because there's no sec to deal with all of these antags. It's because these antags are simply too destructive that sec no longer has the reigns to deal with them. Thus the shuttle gets called. It'd be better to just remove it as the best alternative. The immortal antagonist, and the Boom Boom your station is doomed antagonist are just too powerful of a combination.

The point would be to reduce the chances for stupidly high damage by reducing the number of traitors in a round that can cause the damage. Balancing TC is a code solution, not a policy one.

One traitor can still fuck over an entire station using 0 TC. The point is to make it so there are less people in a round trying to fuck over a station so non-antags can manage to deal with it.

lings are the only solo antag i enjoy playing against as sec it's nice to have antags around that will almost guaranteed need more effort than stun>done as it is with 90% of traitor encounters. i cant speak for everyone of course but it's nice to have variety.

MrStonedOne wrote:

You could break it into two coefficients if you wanted to. One for normal pop and one for high pop

No! this is cancer. there are at least 15 configs we have that are such bullshit population based configs.

We need to cut the cancer out at the root: ConfigMappings.txt. An idea I had for a config that instructs the server to change configs when certain values get above or below a threshold. use it on population, server load, admin count, etc.

It doesn't already do that? I kind of assumed you'd switch config files for different states.

WarbossLincoln wrote:

MrStonedOne wrote:

You could break it into two coefficients if you wanted to. One for normal pop and one for high pop

No! this is cancer. there are at least 15 configs we have that are such bullshit population based configs.

We need to cut the cancer out at the root: ConfigMappings.txt. An idea I had for a config that instructs the server to change configs when certain values get above or below a threshold. use it on population, server load, admin count, etc.

It doesn't already do that? I kind of assumed you'd switch config files for different states.

Each indvidual config checks for population itself if my shitty ability to read code actually proves correct. So you end up with 15 checks when it could be done by one population check switching to different pre-sets.

Again, take my explanation with a grain of salt, I dont know how to code worth a dick.

as soon as you code it mso