Traitorchan/DA Roundtype Probability

[Arete]

Bringing the discussion from singulo over here, using this data.

In the last 30 days, we've had 670 rounds from among the types traitor, changeling, DA, and traitorchan. We've had 364 rounds from among all the other types combined. When you consider that traitor-type rounds tend to last much longer than ops, rev, wizard, blob, or malf, I think it's fair to estimate that something like 80% of the server time is being spent on traitor-type rounds. I would like to see other roundtypes happen more often, and I think most players feel the same way.

Lowpop rounds currently force one of the traitor-type game modes because the others aren't considered viable for a lowpop crew. That's probably fine, although I think rev might still be fun even at lowpop. The adjustment I'd suggest is to change the probabilities for highpop rounds to be weighted more heavily toward the other game modes. Traitor-type rounds at high population just don't involve most of the crew in anything interesting.

[Bluespace]

Some admins take it upon themselves to force gamemodes to mix it up a bit, especially if they last 3 rounds has been TRAITOR TRAITOR TRAITORLING etc.

[bandit]

I'd personally like to see shadowlings, gang and/or abductors added to the rotation. Is this seriously just a code freeze thing?

[Remie Richards]

No, those 3 modes are out because they ARE or WERE in heavy development and not meant for standard every round play.
Some of them are *probably* fine to put in now.

[PKPenguin321]

Some admins take it upon themselves to force gamemodes to mix it up a bit, especially if they last 3 rounds has been TRAITOR TRAITOR TRAITORLING etc.

Which is a good attitude to have, I recall having days with four malf rounds in a row (byond RNG) and it just gets stale
No mode should be repeating itself that often, and traitor/ling/DA shouldn't be exempt.
Also, put shadowling into the rotation already :V

[Incomptinence]

No, those 3 modes are out because they ARE or WERE in heavy development and not meant for standard every round play.
Some of them are *probably* fine to put in now.

Consider it "testing" and have your black little coder-heart beat faster.

On a serious note while abductors is not a Michael Bay movie on the surface it seems fully functional and I can't see a way it could upset anyone even if it is still in development.

With the potential to work with a single team of two it could even go on high end lowpop for a little spice or additionally work as a random event also.

Looking at the stats it is good vanilla changeling has been made rare or even removed from rotation. 6% success rate on just 10 bloody rounds how you suck so bad ling?
After that we are instead looking at 316 trait-chang rounds to go with the 322 traitor rounds, 22 double agents, 10 changelings and all of this squeezed into 1069 rounds about 62% of our raw rounds is spent on this stuff ignoring duration.

[Amelius]

Traitor is just probably the most interesting roundtype that allows for the most variation and creativity on the part of antags and crew, and so it makes sense that it's the most played. Shadowling isn't in yet for some reason (WHY ISN'T IT IN?!).

What I CAN tell you is that I don't think I've ever had a fun wizard round since summon guns/magic was essentially removed (I looooooved the old survivor 5x summon guns / magic thing because it got the entire crew involved), and it's gotten to the point that I just suicide as a non-priority role as soon as I hear 'WIIIIIIIIIIZARDDDDDDDD!', since almost every single wizard has at least one mindswap apprentice (that strings the round on foreeeever), tries to play 'peaceful' (extended without admin events), goes robeless (mindswapping), or plays passively and evasively until someone suicide bombs them (EI NATH, zzzz). The rounds are long, drawn-out, involve a single (or handful of) antagonists that usually don't actually manage to cause much mayhem, with plenty of 'stealth' options that result in even longer rounds. Then, it mulligans into extended a good third of the time, drawing it out even more.

Then blob itself is structurally repetitive, and hence boring, on both the part of the blob and crew.

Definitely not looking forward to seeing more wizard or blob. Single antags a shit.

[PKPenguin321]

Some admins take it upon themselves to force gamemodes to mix it up a bit, especially if they last 3 rounds has been TRAITOR TRAITOR TRAITORLING etc.

Which is a good attitude to have, I recall having days with four malf rounds in a row (byond RNG) and it just gets stale
No mode should be repeating itself that often, and traitor/ling/DA shouldn't be exempt.
Also, put shadowling into the rotation already :V

Is pseudo RNG possibly not too much work to ask for picking the round type without picking the same or a similar round type as the last one or more? Surely it'd save admins an occassional headache?

This idea's been suggested in the past but been turned down since it would make it easier to metagame the roundtype
"Last round was cult, so we probably wont have one this round." etc.

[Incomptinence]

Although as we get more round types that ability to count one out matters less and less.

[Steelpoint]

What's the point with new game modes if they're going to be left in 'testing' for ever.

Put all the new game modes on the full rotation (which in off itself is a great trial by fire).

[callanrockslol]

> make them less common

Why not make the other gamemodes more common?

[Cheimon]

Yeah, please add the new game modes to the rotation. Abductors in particular doesn't seem to have any real problems, but it's being excluded apparently just for being new.

[Tokiko2]

Abductors not only work well, they also works really well in lowpop which direly needs more variation.

[DemonFiren]

Which is to say, Basil direly needs more variation, and ho boy it does.

[Amelius]

Put in shadowlings. Pleaaaaaaaaaaaaase.

[mosquitoman]

Put in shadowlings. Pleaaaaaaaaaaaaase.

Shadowlings are useless. Nobody knows what to do when enthralled. Some people even repair broken lights.

[DemonFiren]

I guess that's because people haven't had the practice.

What you do when enthralled? Follow your bossling's orders. Bust lights everywhere. Bring more meat to the bossling.
Basically, same as allium, but you have to remain stealthificate.

[Durkel]

Just remove DA from the rotation and we're golden. There's no reason to have a second traitor type with gear restrictions.

Also, shadowling blows never add to the rotation pls.

[Lumbermancer]

When you consider that traitor-type rounds tend to last much longer

If it wasn't for traitor I wouldn't even queue for civilian jobs for that very reason.

[Incoming]

Mulligans were created as an effort to give traitor/chan level longevity to quicker modes, but the community didn't really take to that aspect (at least not yet). They just say "WHOOPS THE WIZARD DIED LETS GO HOME" before the system can really do anything. Then they bitch to me that mulligans don't do anything and it just makes them wait an extra 15 minutes to shuttle out after they end the antag.

But yeah one of the big hopes for the system is that eventually it will make most modes into something that could have a decent length even if the threat is something that could normally be quashed in a matter of minutes. If this happened the "relative time" of playing traitor/chan/traitorchan would drop significantly without proverbially shitting in the punchbowl of people who like the more moderate pacing of those modes. Still a long way from that though.

[Not-Dorsidarf]

On the topic of mulligan: Is your ultimate objective to make every round be a 4-hour wild ride, then disable the emergency shuttle until the fourth hour?

[Incoming]

Maybe 2 hours would be good, 4 is a long time and a lot of people don't really have that kind of time commitment just lying around. I'd never suggest "forcing" a round to be longer than the playerbase wants it to be, I'd just like to give some tools to blunt the desire of rush calling that dominates so many rounds these days.

The whole system is time constrained to 40 minutes by config settings right now, which is fair considering given that right now it really only works for extending laugher rounds where a solo antagonist bites it immediately.

[Lumbermancer]

Mulligan would be better if there was a 5 or 10 (or random) minute cooldown between killing of "main antag" and server distributing then new tators among the population. I think it would help the flow of the round.

People wouldn't instantly scramble to exploit the chaos and destruction caused by previous threat, instead engineers would engineer, doctors would heal and clone etc. It would be like a soft-restart of sorts.

[Incoming]

Mulligan would be better if there was a 5 or 10 (or random) minute cooldown between killing of "main antag" and server distributing then new tators among the population. I think it would help the flow of the round.

People wouldn't instantly scramble to exploit the chaos and destruction caused by previous threat, instead engineers would engineer, doctors would heal and clone etc.

This is already a thing, there used to be a "tranquility period" of 3 to 7 minutes before the new antags were distributed. I have scaled it back to 2 to 5 minutes though because when faced with a peace after a storm, people with access to shuttle consoles tended to say "whoops now it's boring lets go home". By the time it was time to actually hand out new antagonist roles, the shuttle would already be inevitable and the new antags would be screwed over.

Once a wider variety of threats could come from mulligan I'd probably push this period back up a bit (at least for the more disruptive antags) since seeing a wizard 2 minutes after defeating a nuke squad would be a bit much for anyone.

The problem I've found is that there's an attitude that "running from the antags" with the shuttle is as viable a strategy as actually trying to deal with them. If you rush the threats so they can't work on their objectives they'll still "fail" even though the station did almost nothing to impede them. Add on that quick calls mean more chances for antag rolls for the callers and you'll see that, depressingly, a lot of the time people would rather mash the reset button than try to deal with emergent threats.

I feel like these emergent gameplay situations are one of the few forces that can keep the game interesting for players once they've seen all the standard things space station 13 has to offer. People complain the game is boring but some of them are actively trying to keep gameplay as formulaic and samey as possible, though they probably don't realize this.

[Stickymayhem]

Let's make the station lose if they run without dealing with the threat then. In the case of traitor chan, it should count the number of antags on the shuttle. With greentext they count as 2, without they count as 1. If the total is greater than 75% of the number of antags the station loses.

Some system like that could lend more importance to objectives if people care, but even just surviving is good. Die gloriously could count as 1 I suppose, since it's mostly unavoidable except by keeping them from hurting themselves.

[PKPenguin321]

Not too bad of an idea, but it kind of promotes a valid-hunting mentality by essentially making the game say "The crew got all the valids! The crew wins!".

[Saegrimr]

Make that a security specific thing. People will validhunt anyway, no need to actively promote it for the entire crew.

[onleavedontatme]

Let's make the station lose if they run without dealing with the threat then. In the case of traitor chan, it should count the number of antags on the shuttle. With greentext they count as 2, without they count as 1. If the total is greater than 75% of the number of antags the station loses.

Some system like that could lend more importance to objectives if people care, but even just surviving is good. Die gloriously could count as 1 I suppose, since it's mostly unavoidable except by keeping them from hurting themselves.

I don't think people will care much if the station "loses"

I think people call the shuttle early because

A) They want to roll for antag

B) If they don't and the badguy kills everyone they'll get yelled at in deadchat while everyone sits around bored watching the murderboner

C) It's just plain tedious and takes too long to revive everyone/fix things

D) 2 hours or whatever in the evening when people have work and school is a significant amount of time. People can just get bored after 2 hours and want to quit as well. Who wants to be stuck as the chef for two hours?

E) (This might just be me) I like a happy ending. If we killed the malf AI I want the round to end with all the survivors living happily ever after, not suddenly becoming changelings and eating each other.

[Oldman Robustin]

Since I've been back I think I've seen every post that I made in my last couple weeks a year ago.

Traitor still sucks, and can't do shit.

Traitor/ling/da/traitorchan make up a retarded percentage of the game considering how boring they are.

AI makes the game less fun.

There's no transparency or player input on gametype weighting.

https://tgstation13.org/phpBB/viewtopic ... 953#p26953

Time is a flat circle, blah blah, can we actually get some results this time?

[onleavedontatme]

Since I've been back I think I've seen every post that I made in my last couple weeks a year ago.

Traitor still sucks, and can't do shit.

Traitor/ling/da/traitorchan make up a retarded percentage of the game considering how boring they are.

AI makes the game less fun.

There's no transparency or player input on gametype weighting.

https://tgstation13.org/phpBB/viewtopic ... 953#p26953

Time is a flat circle, blah blah, can we actually get some results this time?

Feature freeze is coming to an end, hopefully some of the stuff we discussed in those traitor/AI threads will happen

[Not-Dorsidarf]

I'll also be making a varying shuttle types system, where different alerts give you different shuttles which travel at different speeds. So the Red Alert shuttle, which gets here faster (for, say, when nuke ops are seen dragging the captain off) is smaller and cramped, for just getting the maximum number of people off-station ASAP, but punishes people who just call the red shuttle because they want to get home faster (Because an entire high-pop round's population won't fit without serious rioting). Meanwhile, calling a shuttle on Green gets you a more "luxury" shuttle ride back home, helping sec not get overwhelmed by shittery due to more space/seats/etc.

[Bombadil]

I'll also be making a varying shuttle types system, where different alerts give you different shuttles which travel at different speeds. So the Red Alert shuttle, which gets here faster (for, say, when nuke ops are seen dragging the captain off) is smaller and cramped, for just getting the maximum number of people off-station ASAP, but punishes people who just call the red shuttle because they want to get home faster (Because an entire high-pop round's population won't fit without serious rioting). Meanwhile, calling a shuttle on Green gets you a more "luxury" shuttle ride back home, helping sec not get overwhelmed by shittery due to more space/seats/etc.

Inb4 everyone just lays down on the floor

[Erbbu]

The crew has usually been very apt at stuffing a dozen people into the two or three tiles of an escape pod.

[MisterPerson]

Complaining about the weighing is pointless here because this is about the codebase, not the server's settings. If you're going to make suggestions, keep that in mind.

[PKPenguin321]

where else should we put these kinds of threads? i don't think they would fit too well in policy discussion
subforum for config settings when?

[Incomptinence]

New round types being added as server config options is a little coding related I believe.

[onleavedontatme]

New round types being added as server config options is a little coding related I believe.

They're in as config options already, just set to an extremely low chance or disabled

[Arete]

So, the poll has been up for a week. Even given that the forum population isn't perfectly representative of the playerbase, it seems like we have a pretty clear majority here. Can we get a headmin or two to weigh in on the chance of changing the probabilities?

[Incomptinence]

Cross posting from the traitor balance thread cause it is more on topic here

Using the 69% win rate of revolution is kind of unfair, the number of rev rounds has become SO low that it's kind of too small a sample size to get a real number for.

The "actual" rev win rate is traditionally just about 50%, which is great. If there actually IS a tendency to revolutionaries winning more, it's probably because the roundtype has become so rare that the meta paranoia of heads has worn down a bit.

We really need more revolution rounds

Most round types apart from traitorchang and traitor are also pretty low probability.
1007 completed rounds, 29.89% traitorchang, 29.49% traitor, 7.94% wizard, 7.24% extended (events maybe), 6.85% nuke, 3.27%cult, 3.07% malf, 3.07% double agent, 2.48% blob, 1.29% rev, 1.09 gang (test), 0.69% shadowling (test), 0.59% ragin mages (badmin), 0.49% abduction (test), 0.49% monkey (what). You can do this shit with a calculator it is simple.

Keep in mind this doesn't reveal the actual server weightings because they aren't percentages and low pop heavily skews towards traitor types which is fine because the round type is near totally unbalanced in high pop by the diminishing returns equation. Sadly success rates by population isn't covered in solo antag stats I would be pretty interested in seeing how a scarcity of traitors deals with a full crew.

[Bombadil]

Holy fucking shit. Thisis why i never see cult anymore

[Erbbu]

So basically only ~40% of rounds are not traitor or traitorchan. And I bet a sizable part of those extended rounds are not event rounds. Holy crap.

[Oldman Robustin]

So is anyone able/willing to actually do something about this?

This is probably the simplest bloody request that you can make on these forums. No new code, no changes to mechanics or game balance, just change the weights ffs.

3% cult? 1.3% Rev?

Jesus christ these are probably the only gametypes besides Nukeops that have memorable outcomes.

Meanwhile the highest non-traitorling gametype is freaking wizard and extended.

What do we have to do to actually get even the simplest most agreeable changes implemented around here?

[Amelius]

Cross posting from the traitor balance thread cause it is more on topic here

Using the 69% win rate of revolution is kind of unfair, the number of rev rounds has become SO low that it's kind of too small a sample size to get a real number for.

The "actual" rev win rate is traditionally just about 50%, which is great. If there actually IS a tendency to revolutionaries winning more, it's probably because the roundtype has become so rare that the meta paranoia of heads has worn down a bit.

We really need more revolution rounds

Most round types apart from traitorchang and traitor are also pretty low probability.
1007 completed rounds, 29.89% traitorchang, 29.49% traitor, 7.94% wizard, 7.24% extended (events maybe), 6.85% nuke, 3.27%cult, 3.07% malf, 3.07% double agent, 2.48% blob, 1.29% rev, 1.09 gang (test), 0.69% shadowling (test), 0.59% ragin mages (badmin), 0.49% abduction (test), 0.49% monkey (what). You can do this shit with a calculator it is simple.

Keep in mind this doesn't reveal the actual server weightings because they aren't percentages and low pop heavily skews towards traitor types which is fine because the round type is near totally unbalanced in high pop by the diminishing returns equation. Sadly success rates by population isn't covered in solo antag stats I would be pretty interested in seeing how a scarcity of traitors deals with a full crew.

> 8% wizard.
> Extended above nuke, double cult, malf, DA, and triple blob, rev, gang, and shadowling chances.

Fucking why. No wonder I feel like every round is wizard if it's not traitor/ling. Lower wiz chance, lower traitor/traitorchan chance to half that (15% apiece), boost nuke, cult, rev, add shadowling, and gang to the rotation. Bleb, malf, and wiz should have the lowest chances, since they're single antags (but still realistic).

[MrStonedOne]

The show round type odds in show server revision config setting is on, but it doesn't work

If somebody fixes that, you all can see the odds.

[Incoming]

The show round type odds in show server revision config setting is on, but it doesn't work

If somebody fixes that, you all can see the odds.

If I had known this before, I would have fixed it there. Will fix.

Report yer bugs dang it.

edit: https://github.com/tgstation/-tg-station/pull/9824

[MrStonedOne]

Thanks incoming.

Starting next round, you all will be able to see the game type odds via the show-server-revision verb.

[invisty]

As someone who's been playing heaps of lowpop games at dead'o'clock, I have to say gang is much better suited to low pop than Rev is, since rev depends heavily on the crew-to-heads ratio. Gang works well with or without heads/sec.

60% traitor/chan overall is just... I don't know man. Come dead'o'clock, that's 100% traitor/chan. It's boring as fuck, people go braindead, the server population only dies more rapidly unless the admin (if there's an admin) starts pressing buttons.

Cranking up some alternate game modes for low population would be a good start.

Adjusting existing game modes based on server population would be another good step. Wizards with round_up(1 + (crewpop/10)) spells, perhaps?

Gangs with conversion cooldowns based on server pop?