Antag Rolling Cooldown

[bryanayalalugo]

Usually, for a long time now, a certain amount of people usually end up rolling antag a lot in a row. This doesn't usually bring much diversity from the same players who keep rolling antag, especially if they are the kind that are notorious in killing all of security/command.

My suggestion is to add a "cooldown" or less chances for a player to roll antag after being one. This would give a chance for other players to try also to be an antag.

[Kyet]

At any given moment, roughly half of the server population have their antag preference turned off.

With 50-60 people connected, you can expect around 40 actual crew members. The rest are ghosts, players in lobby, etc.

With 40 crew members, of which 20 are eligible to be antags, and maybe 8 antags are selected per round... that gives each player with the preference enabled a roughly 40% chance to be antag.

The numbers don't always work out this favorably, but this should give you some idea.

Overall, the chance for people who have the preference enabled to be antags is surprisingly high.

 

We can't put a cooldown on being an antag, because if we did, there would not be enough antags.

The better solution is for more people to turn their 'be antag' preferences on, so we're choosing from a broader pool.

[bryanayalalugo]

Then how about people who haven't rolled antag to have a higher chance to roll it? If said people are not numerous/present, then it could then default to those who already rolled antag recently.

[Kyet]

That system would be exploitable.

People could simply split their play between multiple accounts. Since each account is played less often, its also gone longer without being antag, and thus more likely to get special treatment than the player would be if they only played on one account.

 

That system would also have bad incentive for players.

Effectively, you'd be programming the game to prefer giving antag status to the people who play less often, at the expense of the people who play more often. That would create a disincentive against playing regularly. Effectively, you'd be punishing the people who play here most.

 

There is a way to solve both of these problems. You could make it check how many play hours a player has since they were last antag. That at least eliminates those two problems. It makes it so that the system doesn't change players' odds over the long haul. It simply spaces out a player's antag rounds more evenly amongst their normal rounds, so they're less likely to have several rounds in a row where they're always antag, or always not antag.

The issue with doing this, though, is that it introduces an entirely new problem: metagaming.

If its very unlikely that the people who were antags last round will be antags this round, everyone can use that information to metagame, and decide who to trust.

Worse, if you know which players have the antag preference enabled, and you know which players of those were antags last round, your odds of guessing this round's antags increase dramatically. To use the 60/40/20 example again (60 players on server, 40 in-round, 20 with antag role enabled), if you can eliminate 8 of those 20 (because they were antags last round) you are left with 12/20 candidates for being antag this round. If we're again picking 8 candidates, that means if you just guessed that everyone on that list is an antag, you'd be right in 8/12, or 2/3, of the cases. That's a dangerous level of predictability. Yeah, it relies on knowing who has preferences set which way, assumes no churn/replacement of players between the rounds, etc. Still, if you play a lot you'll start to recognize who plays as antags (who has the preference on) and from there its a very simple cross-check against the list of antags shown at the end of the last round. You won't always be right but its potentially a powerful new metagaming technique, and we don't want to introduce new opportunities for metagaming.

The current dice-roll method for determining who gets antag is fair. It may sometimes produce streaks of rounds where you're always antag, or always not antag, but randomness sometimes does that. We can't mess with it without introducing more problems than we solve.

Edited January 24, 2018 by tzo

[Ralta]

I'd much rather try and entice existing players to roll antag. A lot of people are scared to try it, or get frustrated at the 'must-win' attitude that seems to permeate a lot of the out of game channels.

Mind you, I personally don't roll antag very often any more because I enjoy playing a regular crewmember far more, but that's me.

[ZN23X]

I think it's a matter of, there are certain players who have less RL responsibilities, who can play 8-13 rounds a day, who always have antag on, who APPEAR to roll antag more than others because of that.

I always have antags on but only play 1-3 rounds a day so I only get antag every few days.

I'll admit it'd be nice if they used thier frequent antagging to be risky and experiment with something new rather than do the same safe routine over and over again...but if thats what entertains them...different strokes for different folks I suppose.

Particularly the hiding in the engineering pod routine. It's so anti climactic. The round is about to end and a new one begin. If you are wanted for being a dangerous criminal, go have one final battle with security for better or for worse. Give the people  (many of which you killed) what they want.

Edited January 24, 2018 by ZN23X

[Jountax]

Off topic, but for what it's worth I usually try to get somewhat creative. Or at the very least don't do what most antags usually do.

On topic, I think Tzo pretty much hit the nail on the head and kept going until he'd built a house, a front porch, a tool-shed and even an above-ground pool in the backyard.

Edited January 24, 2018 by Jountax

[Calecute]

I agree with tzo, there a lot of problems with this suggestion, and the antag streak is caused mainly by few people with antag on.

Being kinda of new to the game, playing for like 3-4 months I still find antaging very hard. I got jailed or killed every time I was antag. The only time I greentexted I was arrested with an emag and the blueprints in the beginning of the round, politely insisted that I had found them on the floor, Tetra Vega was HoS and cut me a break, letting me go with a Tracking implant. Ironically she was my target. RD teleported gunned her into space. I broke into evidence but blueprint was not there. Someone had left it in the bridge. So I quite cheaply broke in when people got out for the shuttle, got the blueprints and ran for engie pod. Quite undeserved greentext, and my other antag rounds were worse than that.

One time I was a vamp and got got by a sec officer while I sucked my first victim. Bad luck I guess, he was patrolling nearby and heard the wirecuffs come on. Other time I was a vamp I was walking maint met an officer and he harbattoned immediately, didn't even know I was wanted. I wasn't even gonna attack him. I tried to fight back but backup came in form of a borg, and for reasons unknown that borg blew up right next to be and I died. Other time I contortionist suited my way into caps quarter, picked the gun up, and realized I couldn't put it in my pocked, so couldn't get out the way I came in. Of course by the time that happened, I was bolted in. So I dropped the gun and tried to escape without it, but I had taken too much time assessing the situation. Sec borg came in and got me.

I think someone should try to do an antaging guide to help people with this, I mean I don't even know how every antag item works, and I took the time to test a few on a private server. I think one of the main reasons few people antag is that it’s difficult to do so, people try it, get fucked, and stay away from it.

Also ZN23X, I hate the eng pod antags, always fucking breaking into my beloved atmos. 

[EldritchSigma]

The antag selector is purely based on chance. It is a statistical illusion that certain people roll antag more often than others.

Example:

You have 50 crew members in total actually being on the station. For simplicities sake the roundtype is traitor. For further simplicity we say that 1/10 of the active server popultation (50 *1/10 = 5) are rolling traitor. That means, provided that everyone has antagonist turned on gets a 1/10 chance to roll Traitor. Now lets go to the more extreme cases: Only a fraction of that population as antagonist turned on. As in 20 people out of the 50: Now since the active server population is still 50 people there are still 5 antags. But now the chance that you get to roll antag out is 1/4 because onnly 20 people are even considered for antagonist roles. 

Now to that what is called "gamblers illusion" or more commonly gamblers fallacy: Say you are 1 of those 20 people who have antag turned on. And you dont get antag that round. Round passes and the next round begins. Same situation, 50 active players, 20 people with antags turned on. Now you dont get it again. Same thing repeats for another time. Previous chance rolls do not affect any current chance rolls whatsoever. No matter how you put it. It is possible to not roll antag a few thousand times in a row. Same goes in the other direction. 

Basicly:

The antag selector does not attempt to even out antag selections based on players. It does not have any preferences nor does it have any kind of factors that could lower your chance of rolling antag. The Antag selector doesnt care how often you got antag in the past couple days. But it also doesnt care if you rolled it like 10 times in a row today alone. It is just rolling a dice and nothing can be done about like it is.

 

Now should it be changed: In my opinon it should not... if you implement factors that increase or decrease the chances of rolling antag you can be absolutly sure as hell that people will figure them out and abuse them. Behind every character is a human being after all. And not everyone has the same decency. Some are decent, going so far to say they are nice (and that comes from a misanthrop), but lets face it, most are just stacking odds in their favor one way or another alongside other things what people do not like to see yet possess the same intrinsic attributes they dislike. But I digress.

Antag Selector is alright like it is, if people would turn on antag there would be broader selection of people to possibly become one and thus more diversity. Its not the system to blame, its the people to blame.

Edited January 24, 2018 by EldritchSigma

[Shadeykins]

The fundamental problem here is the number of people who actually have their antag prefs enabled.

I'll tell you right now, it's less than 50% of the online population at any given time.

On a 100 person round, I would say maybe 15-30 people (at best) have their prefs for that antag type enabled.

In a traitor round at 100+ pop you may have 10-12 traitors, this means, depending on how many people have their antag enabled, you're at times more likely to be selected for antag than not.

I'm sure someone could actually pull up figures on this, I'm just basing my numbers off what I've seen firsthand when trying to manually replace traitors (I usually have to cycle through about 15-20 people who have their prefs disabled, before coming across someone who doesn't).

[TrainTN]

6 hours ago, Ralta said:

I'd much rather try and entice existing players to roll antag. A lot of people are scared to try it, or get frustrated at the 'must-win' attitude that seems to permeate a lot of the out of game channels.

Mind you, I personally don't roll antag very often any more because I enjoy playing a regular crewmember far more, but that's me.

This, more or less. Not a lot of people enjoy playing antag, myself included, because antag is very difficult. It should be difficult, but it is why few people play it. You're either good at playing antag and play it regularly, or you're not good at antag so you never play antag so you never get better at playing antag so you rarely retain new antags in the player pool. It's a role where even the slightest mistake or just simple bad luck can screw you over and get you either immediately executed by Security, or thrown into the permabrig for the rest of the round and left to rot and/or suicide. An early-round error can leave you dead or imprisoned for upwards of an hour or more, and Security can and will go every extra yard to sniff out any possible wrongdoing on your part because it's not their job to give you a break, it's their job to beat you in the head with a stun baton. So you end up in this situation where you see the same people as antag over and over, not because the system is improperly weighted, but because they're the only people willing to do it, and are probably the only people capable of doing it too. Turning the system against those people isn't going to open more opportunities for other players, because other players just don't want it. My antag rounds have never gone good because I'm not good, and getting good at antag takes a lot of time and experience and failures to learn from. So I prefer to just play the game normally, I get more out of it by playing it safe and I spend fewer rounds dead and imprisoned and humiliated.

The idea of playing the bad guy is fun, but it turns out playing the bad guy is extremely hard. Especially when killing people is messy and unpredictable and usually not even worth it because reviving the dead is VERY EASY to do, and stealing high-value items usually involves killing people for them anyway.

Edited January 24, 2018 by TrainTN

[Tayswift]

I'll quote @FlattestGuitar who has done some analysis on antag fishing:

Quote

It's basically impossible to catch someone antag fishing. There are a handful of people who spend almost half of their time playing as an antagonist, while the average player should only get around 10-12%, based on some calculations I ran back in the day. The numbers could be outdated, but there was a clear and obvious difference between certain people.
The amount of bans handed out for antag fishing by the Paradise staff team as far as I'm aware isn't even higher than like five, because even with the numbers supporting your claim that someone is antag fishing, there isn't a definitive way to prove anything unless the person outright states it as a fact.

So randomness is the most fair way of doing this, but there are people who will tend to leave the round earlier if they're not antag, resulting in a lopsided amount of time being antag. In other words, antag selection is entirely fair (law of large numbers), but what you're noticing is some players having noticeably more antag playtime, which changing the antag selection won't solve.

The solution is to encourage people to antag more. Part of this is probably allowing you to have an "antag character" selected, so that if you roll antag, you play with that character instead. That'll solve a lot of issues such as IPCs not being able to roll changeling/vampire.

Edited January 24, 2018 by Tayswift

[TrainTN]

1 minute ago, Tayswift said:

I'll quote @FlattestGuitar who has done some analysis on antag fishing:

So randomness is the most fair way of doing this, but there are people who will tend to leave the round earlier if they're not antag, resulting in a lopsided amount of time being antag. The solution is to encourage people to antag more. Part of this is probably allowing you to have an "antag character" selected, so that if you roll antag, you play with that character instead. That'll solve a lot of issues such as IPCs not being able to roll changeling/vampire.

That doesn't sound fair. That could be metagamed incredibly easily if you can tell who is never antag and who is always an antag. Isn't this supposed to be a roleplaying game about paranoia in space? It's supposed to be a game where almost anyone could be a traitor, it's why they exist in the first place. If I can just meta who is always innocent and who is always guilty, what's the point?

[Tayswift]

2 minutes ago, TrainTN said:

That doesn't sound fair. That could be metagamed incredibly easily if you can tell who is never antag and who is always an antag. Isn't this supposed to be a roleplaying game about paranoia in space? It's supposed to be a game where almost anyone could be a traitor, it's why they exist in the first place. If I can just meta who is always innocent and who is always guilty, what's the point?

Our slot system allows you to generate a random char/name. So you could set "changeling" to always be a random human, while "traitor" to be your IPC/whatever slot you have declared ready for. It's really not a meta-able thing unless you want it to be.

Edited January 24, 2018 by Tayswift

[Calecute]

I'm not sure how would you do this with the slot system without fucking up matchmaking. 

[Tayswift]

1 hour ago, Calecute said:

I'm not sure how would you do this with the slot system without fucking up matchmaking. 

What is matchmaking?

[Calecute]

Matching peoples preferences to their jobs/roles.

You can have as many characters as you want, but you just choose one to ready with

[Calecute]

I mean, as far as i know, i might be wrong

[EldritchSigma]

Might be a bit off topic:

What would be really nice that the antag selector could be decoupled from the character you select, as an entirely optional opt-in/opt-out thing. That means you create your characters as per usual, set their preferences, but in addition set the characters antag preferences. Now if you select for instance with Goodguy McGoodwill, a person who could never do evil and is truly incorruptable, but the antag selector chooses you he selects another character of yours to play on, for instance Joerge Mwerlins the evil dude.

It is codeable... yes. Yes I have tried that myself, but it is a pain in the ass to implement something like that (the most I have got out of this was per character antag prefs and that was only applied to the slot one character...). And I am more versed with tgcode than with paracode... but atleast its not impossible, someone who is really good at paracode and has alot of dedication could do this. 

[Calecute]

Yeah, if matchmaking works the way i think it does this would make it better, you could also do it in a way that assings jobs first and them character, witch would also make it better.

[TrainTN]

I'm still against making antag options specific to characters. It removes a lot of tension knowing that this specific person can always be 100% trusted. Not that it's not possible already if you can quickly alter your options before readying for the round.

[LordSplinter]

I think The only way of people not metagaming it if people don't get to know about this system, a alternative there could be it's that admins would manually antag you when you play alot and even playing alot you don't get it(when I was with my PC,it happened alot of times that I didn't get antag for 40 or 30 rounds while playing about all day everyday)but the actual system works good,you can probably get a bunny feet collar with you while playing

[Shadeykins]

On 1/24/2018 at 2:25 PM, TrainTN said:

I'm still against making antag options specific to characters. It removes a lot of tension knowing that this specific person can always be 100% trusted. Not that it's not possible already if you can quickly alter your options before readying for the round.

To play devil's advocate, anyone who has their prefs off can always be 100% trusted.

And there are a LOT of people who have their prefs disabled.

[TrainTN]

3 hours ago, Shadeykins said:

To play devil's advocate, anyone who has their prefs off can always be 100% trusted.

And there are a LOT of people who have their prefs disabled.

You're right. I just think making it a character-specific thing would exacerbate the issue. If you know Character A is always a goody-two-shoes but Character B can be any antagonist in half the rounds he's in, one of them is going to get a lot more suspicion than the other all the time. I think it takes a bit of fun out of it, it's already a bit of a problem where you can almost predict who is an antagonist depending on who is online, but it's not like you can force people to be bad guys if they don't want to.

Unless it's a conversion game mode. Let's bring back Revolution.