Live Action Mafia

A game of sneakiness and paranoia
It is currently Thu Jul 29, 2021 2:10 am

All times are UTC - 5 hours

Post new topic Reply to topic  [ 26 posts ] 
Author Message
PostPosted: Sun Oct 12, 2014 11:56 am 
If anybody has constructive criticisms/suggestions for future games, they should be placed here. This will also be discussed at wrap up but if you want to make sure things are written write them here, some things which will be certainly be discussed and debated at wrapup. I would appreciate if people would make suggestions in line for improving games and remember that most disagreements over roles are coming from the place of trying to make game better. I would also appreciate if people who wish to make criticisms have played game in a good faith manner and keep ad hominem attacks to a minimum. I trust that most of you will be able to do this though, as mafia does have a storied history of leading to unreasonable behavior (ask us at wrap up!), I do request that those for whom being civil is difficult please refrain participating on this thread.

game power level,
the role of godfather,

I'll start out by giving some opinions.

A few explicit points of clarification,
Remember a single game is not indicative of the balance of the set up, simulator results demonstrate that the current set up was probably mildly mafia biased, this to me, also agrees with my intuition that mafia would be winning if they currently had 2 or 3 more mafia which, were kills made more strategically they would have. While balance is certainly important, the current balancing meta seems to be doing a relatively good job, (Of the last 7 games, predicting, that this game will go to town, 4 were town victories, 3 were mafia unless I'm mistaken)

Game power level was relatively on par with previous games with the current meta (i.e. every game from last fall onward),
Power roles are, as a rule, not stronger than they were historically but far more numerous then they were until the fall game 2 years ago, in particular, there used to be a role named ironheart which had immunity from all kills except for perhaps 1 (unrechargable) kaboom.

An important to keep in mind is the following obvious equation, the proportion of town lynches which need to be accurate is the number of mafia/ the length of game. Town lynch rate can not be expected to be very successful unless town has a decent amount of power or mafia are making numerous substantial mistakes. Discourse dominated by strong townie reasoning taking advantage of moderate sized mafia mistakes, as seen this game, is very rare and game should not be balanced around it (unless we want an 80+% mafia win rate). In particular, based on what I was seeing as god, town absolutely and completely deserves to have won this game for their very strong play. (For example, catching lzahray and scgordon, being suspicious of victor from earlygame, and quickly investigating achester, alex and michiyiko).
Thus, in general, for large games such as this we have a choice between
1. very long games in terms of time,
this seems bad to me because of how draining mafia is, this game is long enough as is, balancing around 3 to 4 week games seems like an unworthwhile tradeoff to me, people get bored and don't want to put in that much time.
2. a small number of mafia with a very large amount of power
I considered this choice and worry that it puts a tremendously undue amount of stress on mafia which, given how stressful previous games have been for certain mafia might be a bad choice, it is certainly an option.
3. splitting the game into several smaller games,
This requires much more administrative power which we don't actually have in surplus, it's also rather annoying, I'm of the opinoin that this is necessary for substantially larger games then this (say, by the 90 person mark we really need to do this) but it's probably better to play a large game then make 2 small ones, even if the small ones might be slightly better.
4. Capping the number of people who can play in a given game
I strongly dislike this option, though others might disagree
5. a game like the current one
There seems to be a general consensus that town was too powerful and this has a negative effect on game.
6. just making games very mafia biased when they are large
this is an option, though one I expect people will have strong objections to.
7. another option
I haven't figured out another option I've seriously thought about, though one probably exist.

Basically there are no great options here (though we are obvious in a fortunate place where these are the potential problems rather than a lack of players)

Some other miscallenous thoughts
Genies, as written, are inelegant. As acted upon this game they are not overpowered (they were far weaker then investigators and, with the exception of people who stored numerous charges and never used them, arguably weaker than superheros). Benevolence, as written, is a trap and I'm unsure of a good way to clean it up for mafia.

The propensity for roles which give immunity from kills and investigative powers is, in large part, a function of the lack of good bonuses to give town for correct guessing, role verification, as exist for innocent children, gay knights, and to some extent prophets, is dangerous and game breaking when it makes up too large a proportion of game (see spring game 2012), and nobody has given an example of a good town incentive which doesn't fall under these three categories. Historically this was counteracted by the existence of vanilla townies, however a large majority of players seem to prefer being a powerrole to a vanilla townie.

Speaking of prophets, they are an incredibly strong role, currently the best in game. A way to make them weaker is the following variant if you smite incorrectly you immediately die. The reason this was not implemented was worry that it would be unfun for a prophet to make a smite, and then be immediately killed if they were wrong. How do current or historical prophets feel about this?

Desperados are underused by town.

Interrogate is way to strong in my opinion, a new bonus for police officers needs to be given instead.
One idea I think might be cool is that investigations can only be made during a police officer's tenure (i.e. with an elected police officer) of a person killed during the police officers tenure, and furthermore, during a mafia police officers tenure, all investigations come up negative, I haven't tried especially hard to test whether this breaks under abuse but it seems like a clean way to weaken investigations and make police officer necessary.

Late game conscripts seem bad for reasons ksedlar has conscripted, it sucks to have your alignment changed late game.

Rules are too complicated, this will partially be fixed and is partially unaleviatable with anything resembling the current rule set.

PostPosted: Sun Oct 12, 2014 1:18 pm 
Josh Frisch wrote:
The propensity for roles which give immunity from kills and investigative powers is, in large part, a function of the lack of good bonuses to give town for correct guessing, role verification, as exist for innocent children, gay knights, and to some extent prophets, is dangerous and game breaking when it makes up too large a proportion of game (see spring game 2012), and nobody has given an example of a good town incentive which doesn't fall under these three categories. Historically this was counteracted by the existence of vanilla townies, however a large majority of players seem to prefer being a powerrole to a vanilla townie.

Man, the syntax of this paragraph was hard to parse. :P

But anyway, I speak here as a new player, so my advice is heavily biased by this one game. But here is what I felt from this game.
Comments regarding each of the roles:
Basically the core of the game, but with the existence of the don and many manipulative powers, as well as fake investigators (see Ade), this should not be treated by town as infallibly as it is now (where "(s)he came up negative" is enough to dismiss NP evidence for a time). I think weakening investigations too much would make the game significantly harder for town, which may not be entirely a good thing. A moderate nerfing might be good though. One idea is to make them known to whoever they're investigating, but this would just get them killed early in the game and make late game drier.

Under the current situation this is basically a half-useful investigator (who would choose to be strongConspiracy or strongCynic??), with the one bonus being providing some protection against set-a-trap for anyone with an investigative role due to doubt about their actual role. If investigator is not nerfed (or maybe even if), superhero should be strengthened somehow.
One idea is to gain two charges for each inactive day, but that just adds to the investigation overload; another is to add more crazy roles superheroes can be, including maybe a strongProphet-Desperado (investigates, stays alive for however many charges they currently have, gets two smites within that time period, dies immediately after the second smite or when the charges run out).

I agree, underused, but with so many investigative powers going around already and the inability of claiming for fear of kaboom, this is not that useful of a role except when you know in advance that you need to get something done within the next few days. Strengthen slightly, perhaps, unless investigator is nerfed?

Underused this game, though the threat of its use played an important part in some mafia decisions, it seems. Considering this is a one-shot ability, I don't think it's that overpowered, and having immediate death would certainly make it even more underused. One possibility is to have them die that night even after a successful smite (I've fulfilled my duty in this world, now I can go meet my Maker), which weakens just slightly but encourages usage more?

IC, Gay Knights
Worked as intended for innocence purposes, I think. Inactives are a big problem for these though, since they hinder the network.
For knights (whoa it's National Coming Out week or something), inactive/evicted partners is a big issue; the remaining player should probably stay alive. One possible dynamic is to have some partnerships develop over time, perhaps voluntarily, or chosen by a cupid-like role after some in-game stuff has already happened :P

The rules for this were one of the parts I grossly misunderstood; I thought each wish only gave you one investigation, not arbitrarily many investigations on one kill. I agree that this is probably too OP, but if benevolence is modified to confirm known kills (i.e. if anyone's already gotten a positive investigation for that kill) it should be ok, allowing town a small degree of manipulation/testing if they know a killer but have not used any investigations on that person.

I actually enjoyed this role a lot, in part because the feeling of receiving invulnerability for correct predictions was nice, but also (in retrospect) because it allowed me to troll mafia (almost trapping one early on :P ) without too much worry, though I'm afraid that might not work for me anymore in future games. I think it's good to have a totally invulnerable role that's reasonably hard to activate, for the purpose of petition-carrying and whatnot; it encourages such a player to be active and doesn't break the game too much unless mafia actually fall for the trap (which they essentially did with me, though unfortunately circumstances prevented that trap from being fruitful).

Conspiracy Theorist
I agree with the general sentiment that this is the least useful role in town at the moment, but I think it's a nice thing to have (though it's played virtually no role in this game except in getting lgunder killed instead of dininno). It's only useful at all if no one knows you have it though, and there's not as much pressure to be active as there is for cynic (you just put people in your living group and publicly suspected mafia), so might be good to strengthen it with some other odd factors.

The existence of rogues allows a few townies an opportunity to try to lie to mafia. In my case it did give potentially useful (though unfortunately misused) information. Unfortunately for mafia, no real rogues were found through public sacrifice this game, which brings up the main problem for rogues: inactivity. With no one to guide/push them, it's very easy for them to just not do things, especially since they might accidentally kill another rogue or get exposed (they don't share any of the role information mafia can get, and don't have powers to cover themselves up), or even get slaughtered by rampaging mafia. Probably should strengthen.

Elected roles:
Not much use in this game (due to investigative inflation) except using their enumerative combinatorics skills and acting as a figurehead. Should be strengthened unless other investigations significantly decrease.

Police Officer
Lynch targets after their interrogation, kill targets if they're not mafia-aligned; basically, for people who don't care much about surviving long. I don't think interrogation is that terribly overpowered since it takes so long, and if they have good reasons to guess correctly, they could've guessed correctly in just leading a lynch anyway.
If interrogation gets nerfed, running for PO will probably become even more unpopular than it is now - the risk of mafia free-kabooming becomes much greater than the benefits of a cynicky triple investigation (again, see investigative inflation). If non-mafia-aligned PO is needed for ANY investigations to work, mafia will just keep killing POs, and the game would be MUCH harder for townies, but this may be OK if protective roles are strong/common enough.

A nice addition for mafia, especially when town is being silly and forgetting its existence. Current limitations are probably appropriate. Free kabooms are an idea but would probably just be too OP.

I feel like he's doing too much this game, but I'm probably biased. I would personally really enjoy being a town version of godfather (without PM ability), since my primary mafia skill seems to be enforcing consistency in reasoning among townies. (My other skills, intuitively sensing liars and trolling liars, would probably be less relevant for this.) But this would probably make town too OP, since the current limiting factor for town is basically its ability to retain information and reason correctly.


More importantly, general comments:
-If we believe ksedlar's simulations, the game should be relatively fair. The reason, I think, that it seems unfair, is that the determination of game results comes really early. It's like a stochastic process involving guerrilla troops fighting a traditional army, where early extra losses in mafia significantly reduces their chance of winning, while early lucky escapes can significantly increase their chance of winning. So for much of the game, town victory seems to be imminent, even though initially it's not imminent.
One way to reduce this feeling of fatalism is to make game determination happen later in the game, e.g. push mafia's power to later in the game. This way there can hopefully be a much higher chance of actually having comebacks (by both sides) :O

For example, mafia can have a necromantic role that, once only a third of the mafia are left, picks one mafia-aligned person from the graveyard to send their choice of a 100-digit binary code (if this doesn't break the cryptography rule) or a 22-letter alphabetic string, through the mods, to the living mafia. They can pre-agree on what this string means. This particular example might actually be really bad, but some way to make mafia's win-probability stable around 50% until mid-game regardless of early happenings would be nice.

-Another major issue is the silencing of experienced / clever players. These are targeted by mafia to throw town into confusion and allow mafia to control town, and if they remain alive are targeted by town due to suspicion. Not only does this presumably make the game less fun for very experienced players, it gives selection bias for newer players to be inactive and/or not clever, to avoid getting killed. But cleverness is something that should be the highlight of mafia, so it'd be nice if clever reasoning could be preserved, and kills were more to cut out currently contained information or power than to cut out future reasoning skills.

-Alibis have some major problems. Inactives not posting alibis scarcely get investigated (and it would generally be a waste to investigate them, especially if we want to nerf investigations), so this is another source of pressure for people to be inactive. We need to fix this somehow; perhaps warn people beforehand that being inactive is not acceptable.
But even among active players, it's hard to always have (good, confirmable) alibis when you're innocent, so some degree of investigative redundancy is still pushed for in big games. I don't know if this is fixable.

So a lot of these problems point back to investigative inflation, but some are just there on their own. I don't know if the patches I proposed would be good, but this is what I've gotten from my first game.

PostPosted: Sun Oct 12, 2014 4:31 pm 
Recruitment has a problem in that if a dying (about to be lynched) mafia tries to recruit a non-rogue, the mafia find out the result of the recruitment but nothing else happens. I.e. free recruitment. Perhaps failed recruitment should have a different punishment (maybe so mafia are willing to risk it without having a disarm happen).

I say no to the prophet part of StrongDesperadoProphet but agree that we need something else for superheroes.

PostPosted: Sun Oct 12, 2014 4:33 pm 
Derp ok I take back strongProphet-Desperado. Though I do think superheroes should have some sort of nice choice for saving up a bunch of charges and then sacrificing their life in some spectacular move. Perhaps just a strongDesperado where the number of days they survive, plus the number of investigations they make (at most one per day) equals the number of charges they have stored up or something. This majorizes regular desperado, which we could then delete as a role, perhaps.

PostPosted: Sun Oct 12, 2014 4:37 pm 
-Alibis have some major problems. Inactives not posting alibis scarcely get investigated (and it would generally be a waste to investigate them, especially if we want to nerf investigations), so this is another source of pressure for people to be inactive. We need to fix this somehow; perhaps warn people beforehand that being inactive is not acceptable.

Perhaps the solution is to be much more proactive about evicting people, or maybe to simply have smaller games.

PostPosted: Sun Oct 12, 2014 4:42 pm 
^Yes, a smaller game where everyone is active (and preferably has a reasonably sized working memory and reasonable deductive accuracy) would solve many of our problems. It would also, however, be harder for newer players to slowly get into, and poses some other problems as well (do we have to take applications? That sounds terrible.)

I guess we could divide the players into more-experienced and less-experienced games, but I think Josh has already addressed some of the problems with that above. So idk.

PostPosted: Sun Oct 12, 2014 5:05 pm 
Some thoughts.

1) I definitely think there are currently too many investigations. The only way I could think of for mafia to get away with making any kills at all in the setup from this game is to either (a) use powers or (b) be completely inactive in town square. I realize that there weren't any serious attempts this game at getting away with kills as basically all the non-don mafia admitted to making bad kills, however I think this is a symptom of them not seeing a way to make "good" ones. The only exception was ltchin->aok but this only worked because it was in the middle of a rampage which is obviously not a sustainable strategy.

If you count the number of investigations that happened on unsolved kills (basically just Ben's), they are probably all enough to cover all the unalibi'd actives. This is pretty silly. There really needs to be room for mafia to make kills without using a power. The power creep effect where more investigations keep being added only to be "patched" with more anti-investigations ad infinitum should honestly just be reversed a significant amount. One suggestion would be to have a very small number of investigations, but put them only on elected roles to ensure that they are held by actives.

2) Possibly stupid suggestion for rampages: how ridiculous would it be to have everyone in game start with the following item: "conditional tazer", works like a tazer except you must write both your target and a suspected victim on the physrep; still one time use. The effect only succeeds if your target actually killed the suspected victim, otherwise it fails (you aren't informed of the result). This should completely stop rampages but unlikely to be useful if the mafia isn't already caught. Unfortunately, it's not very elegant but maybe something can be worked out from there.

3) The current petition process really sucks. Firstly, there's the activity check of being next to impossible to get petitions completed once there's a lot of inactives. Secondly, the game is so spread out now that it's a gigantic pain to go to all the living groups and get people to sign them, which basically feels like busy work (i.e. not fun). Thirdly, the fact that petitions work like other game items and can be intercepted and destroyed by mafia seems silly to me and exacerbates the frustration/pain of going to the effort of getting them signed. Getting several piles of half+ completed petitions ripped up multiple times this game was really annoying.

I think for games this size it might just be better to move to electronic petitions and try to force face-to-face interaction more through roles (in fact elected roles would be a great place to add more interactivity, since the holders of those petitions are guaranteed active/willing).

One other thing to notice is that the existence of very powerful elected positions doubles down on the problem of experienced players dying early, since office holders are always both super high priority targets and experienced. Not sure if anything can be done about this short of removing or greatly weakening offices but it's something to consider.

4) I think desperado is a pretty lame role honestly, and I know at least Sarah (a fellow desperado) agrees. A small number of additional (but still fallible) investigations never felt useful, and in fact there was no point in the game that I would have even had the ability to catch an uncaught mafia by activating, as all the investigatable kills were caught very quickly. I also never felt that 3 days of unkillability followed by death gave me positive expected lifespan (obviously at some point I was wrong, but I don't feel it was possible for me to predict that). I'd be in favor of removing desperados for lameness.

PostPosted: Sun Oct 12, 2014 6:16 pm 
jakob wrote:
Recruitment has a problem in that if a dying (about to be lynched) mafia tries to recruit a non-rogue, the mafia find out the result of the recruitment but nothing else happens. I.e. free recruitment. Perhaps failed recruitment should have a different punishment (maybe so mafia are willing to risk it without having a disarm happen).

Yeah, it would be quite interesting to have a system where fake rogue can trick dying mafia for more benefits than I got from successfully fooling lopezv/alex :P

PostPosted: Mon Oct 13, 2014 2:03 am 
I actually think that cynic is one of the best countermeasures to the experienced players die early problem. Kill the police officer? Now 1/9 of the townies are immune to all kills tomorrow. :D (Although that assumes a certain base level of activity).

PostPosted: Tue Oct 14, 2014 10:48 pm 
The raw power of investigation is distorting the game. It lets town sit back and wait for positives to get sure lynches. Justin described it well in his post. There's two related issues.

1) Too many investigations. With Investigators, Superheroes, and Mayor, town can average ~10 investigations per day, and that's not even counting Genies, Gay Knights, Desperadoes, and Police Officer's super-investigation. (As these die, the number of the Investigations per day shrinks, but so does the pool of suspects.)

This didn't really matter this game, where town had the killers within the top few candidates. But the spare investigations they had would let them check a whole list of suspects. This means that it's not enough for Mafia to avoid being top suspects; they need to avoid being on the list of suspects at all, or it's only a matter of time until they are checked. Or, they need to have used a power to investigator-proof the kill.

2) Positive investigations are final. There's no real way for Mafia to fight a positive Investigation. They can accuse the investigator/proxy of lying, but the investigation is easy to confirm. They can claim Frame, but town should just honestly lynch anyone who has a positive regardless of Frame (though Lilly did delay lynching surprisingly long). So, caught Mafia have nothing better to do than go kamikaze and be unanimously lynched, which displaces better gameplay.

So, what to do about these?

First, I think this-can't-be-investigated powers (MtP, Don) are not a good solution. It's not fun to correctly investigate the killer only to be told "no". You fail even when you succeed. Any kill you might be totally futile to investigate and you can't be sure. It's effective for Mafia to cover as many early kills with these powers as they can, leaving a period where *every* investigation is secretly hopeless.

I think there's no way around it: investigative roles need a nerf. Not to balance these roles relative to other roles, but to balance the relative power of investigation as a game feature. It should allow town to follow up on suspects, but with hard choices about what kills and whom to investigate. Investigations should follow from existing suspicions of town, not be an engine to catch Mafia.

So, here's the arms-reduction I propose for investigation and anti-investigation powers.

- Investigators can only investigate once every two days (changing superhero appropriately).
- Remove Genie (ideally, rework it with a cookie that's not investigation)
- Remove Don
- Mafia gets only one MtP.

Ideally, one would find some other bonus to give Gay Knights, Desperados, and Mayor that isn't investigation, but as Josh said, those are hard to think of.

Now, these lower the power level, but I also propose a change to give more counterplay to positive Investigations.

- Whenever an Investigation finds that "X killed Y", any future investigation of "Did X kill Y?" comes up negative.

This means that Investigations cannot be confirmed. Anyone can lie about having gotten a positive. The idea is to make make positive investigations spark debates rather than smothering them.

(There's issues that a Conscripted Investigator would need to be unable to preempt investigations. And there might be an issue of distrustful Investigators racing to the kill site to get first dibs. So, some stuff would need to be worked out, but I advocate the direction of the proposal even if it doesn't work as written.)

PostPosted: Tue Oct 14, 2014 11:48 pm 
++++ Alex

PostPosted: Wed Oct 15, 2014 12:19 am 
wow that's neat. I like the idea of one positive investigation making others negative. though that might be stripping town of a lot of power

PostPosted: Wed Oct 15, 2014 12:34 am 
My immediate concern with second investigation comes up negative is to make it standard policy to lynch both the investigator and the mafia. This seems unfun for investigator. I'm very sympathetic to removing manipulate the press, less so to don (it seemed to add a lot of fun flavor this game), the point of don is not solely to up the arms race but to decrease the reliability of investigation. I'm concerned even with the second investigation coming up guilty it fails to do that. I'm not especially concerned its unfun for people to occasionally get misled, especially when they are unaware until post endgame they are getting misled. I'm sympathetic to the claim that investigation is overpowered but seriously question how town is supposed to catch mafia, at say a 40-50% lynch rate without a large number of investigations if mafia is playing reasonably well. Note a .5 mafia aligned player per day would result in a mafia win with fairly substantial probability. Decreasing mafia kill power lengthens game, and decreasing the number of remaining mafia both dramatically increases the swinginess of game and the pressure on the remaining mafia to not mess up, which has been a problem in previous games.

If we decide we want to nerf investigation a few other ways to do it (which met with some strong disagreement on pecker) include
Investigators can only investigate a single kill:

problems, exacerbates early kill penalties to some extent

investigators can only elect while a mayor exist, only kills made during the mayors tenure may be investigated

problems, focuses a large bit of game energy on elections

Allow each kill a fixed amount (say 2) investigations

Problems, disproportionately affects players who are late to investigate i.e. newer players.

each investigator is linked to a single (randomly chosen) mafia, only can find kills by that mafia

Problems, is especially strong nerf

There are other ideas, I think looking at ways to radically change investigations are possibly more worthwhile then looking for ways to decrease their number.

PostPosted: Wed Oct 15, 2014 12:38 am 
Jessk was talking about having two games go on in parallel when there gets to be so many people.
When the number of players gets close to 50 that idea becomes feasible.

PostPosted: Wed Oct 15, 2014 12:41 am 
This is clearly necessary, short of banning people for large enough games (say, when game would be 100, this seems clearly better to me), part of the issue is that running in parallel leads to

1. confusion,
2. not playing with some friends though it would probably be partially split by living group
3. twice the amount of Gods needed.

Points 1 and 3 being most relevant.

PostPosted: Wed Oct 15, 2014 12:49 am 
Instead of 2 gods co goding one game, it could be one god per game.

PostPosted: Wed Oct 15, 2014 12:51 am 
No i think 2 gods are still necessary per game. I would also be in favor of running the games at different times during the semester. People can specify what's more important to them, when they play or who they play with.

PostPosted: Wed Oct 15, 2014 1:50 am 
I think that if we really want investigators nerfed, Alex's once per two days (or a similar, but charge-based system) would be the sanest way to do it. I think weakening things on the player level, versus on the game level (i.e. having a nerf like investigators can only investigate with some elected position in office) prevents swinginess and keeps game from becoming super focused on elements that shouldn't be that game-critical.

Another reasonably sane option is to mess with the probability distribution of roles, AKA throw in a disproportionately small number of investigators compared to other roles. I would also be happy to remove/rework genies.

It's also important to note that having a lot of investigations is actually NOT unbalanced, in terms of raw win chances. In a game with 11/68 players being mafia aligned, brute forcing investigations is simply too slow of a way to catch mafia if kills are random. While it's fine for people to complain about lots of investigations being unfun (due to high mafia turnover rate or whatever), complaining about investigation levels making game unbalanced is just not accurate.

++ to Lucy's comment about 2 gods per game being necessary. Although there is talk of making a web app.

For dealing with larger numbers of people interested, two games might be necessary at some point, although I would prefer making a slightly scarier application process to scare off people who would end up being inactive. XD

If we do have two games, I think it might be worthwhile to have two different types of game; one high-mechanic one that is close to the current version, and one low-mechanic one that is closer to what early games were. The latter could be advertised to all the dorm lists (so new players would get the less confusing version), and the former could be advertised only to recent past player lists. At this point, I think people's game preferences are actually at least as important as other players/game times. I also don't actually think we're that limited by godding interest (IAP currently seems to have aok, Ellena, and Kemus interested in godding).

PostPosted: Wed Oct 15, 2014 9:21 am 
What if investigators got incomplete answers? Such as:

Each player is randomly assigned a multidigit ID number which is public to the town. The investigation returns a random digit from the killer's ID number. That way investigations would give some truth, but also spring debates and heated accusations. With multiple investigations we'd get more digits, but mafia could insert false digits. I think it could be a fun puzzle aspect.

Cons: it's complicated, it might be too big of a nerf. (Shorter ID numbers, ie a two digit ID number, would maintain a fair amount of power.

PostPosted: Wed Oct 15, 2014 6:53 pm 
hmm, all these are interesting ideas. a possible synthesis of some of them:
1. investigators get an inspection charge every other day, and can investigate at most one person for each kill.
2. superhero-desperado gets a charge every inactive day. investigation takes 2 charges, strongcynic takes 1, and at any point you can go desperado, unbangable, each day getting an investigation and losing a charge (or maybe 2?), dying when you're out of charges
3. gay knight stays the same - not that powerful, useful compensation for dying.
4. genie ceases to exist, or grants bang-immunity for a day?
5. prophet dies at end of 2 days if correct smiting, 1 day if incorrect.
6. conspiracy, cynic stay the same; alternatively, have conspiracy who gets attempt-killed unable to speak until the end of the day, and have cynic who gets attempted die at the end of 2 days?

another silly but possible role to discourage investigation:
schrödinger's cat: can't be banged, dies immediately if investigated for anything

also I have an exciting role idea that may break the game, so I'll post it in a few days

PostPosted: Wed Oct 15, 2014 7:05 pm 
I would really love a joker. It'd probably be my favorite role to play.
A joker wins if they get lynched.

PostPosted: Wed Oct 15, 2014 9:42 pm 
Unfortunately, with no vigilante (unless prophet counts) and no person-focused investigative role (besides interrogation), the existence of a joker would probably make town win chances plummet, since every single suspicious person would then pose a risk at lynchtime. There might be some way to mitigate this problem though, e.g. have multiple jokers, all of whom must be lynched for the jokers to win?

Another idea is the analog of a pig (an innocent role that has speech limitations) - such a player can be forced to use a certain letter at least once (preferably a moderately common one like D) in each public forum message. This is a way to have semi-confirmed innocence (i.e. fakeable, but at a price)

PostPosted: Thu Oct 16, 2014 5:08 pm 
Given how crushingly well Town played the whole game, it's scary that they had only 17 Townies remaining when they won. That's a quarter of the starting players. If the Mafia could have survived longer and gotten rid of some of the remaining vocal players, Mafia might have pushed through a win despite being very far behind. What happened?

I think Kamikazes are largely to blame. No matter how well town catches Mafia, they can only lynch one a day, and that Mafia gets free reign to kill whoever they can reach. Often, this means experienced players, elected positions, proven innocents, and known power-roles. So, the best-case for town isn't actually that good. This game, it did have the significant upside of adding excitement to what would have otherwise been a one-sided game. But, it did so in a way I find unsatisfying.

A central conceit of Mafia is that Mafia operate by being hidden. Town wins by figuring out who is Mafia and lynching them. Having known, admitted Mafia walk around in broad daylight undermines that conceit. It just feels silly. Having a Mafia kill one or two key townies in response to a positive investigation takes away from Town's well-justified feeling of success successfully caught a Mafia.

In fact, if Town were looser in revealing roles, I believe Mafia might have won largely by kamikazeing without bothering to stay hidden.

To be clear, I'm not talking about balance here. It's entirely possible to take into account Mafia's kamikaze power in determining numbers, and the GM's for this game did so. Moreover, having a large number of kills might be necessary for a large game not to last too long. It's hard though to balance such a game because it's so un-swingy: the Town-does-great case is not that far from the Town-does-OK case, so it's hard to find numbers that separate them. I also feel that balancing around Kamikazes leads to an unsatisfying game experience.

The power of Kamikazes makes soft-suspected Mafia not fight against a lynch. Why commit Mafia to your defense when you could just take out some important players now? Even if you survive, you'll be on the list of people investigated for kills, which is effectively a death sentence (though investigation nerfs would help this). Worse, a split vote that you unexpectedly gets you lynched robs you of the valuable chance to Kamikaze. Why risk that?

So, Mafia aren't motivated to fight against accusations or point fingers at someone else. What could have been an interesting debate doesn't happen as "I'm not convinced they are guilty, your accusation is kind-of sketchy" turns into "OK, half of EC just say them kill X, everyone lynch". Then, you have a boring "vote" of 35 to 1. And that incentivized additional suspected Mafia to kamikaze that day, since they then get *two days* of free killing if things go OK.

In theory, there's Prophets and Tasers to help. But, Prophets are better saved to smite key players of unknown allegiance. Tasers were, I think, a step in the right direction, but weren't used that effectively, so they don't fix the issue. Mafia can hide away and stalk around to make kills if they anticipate a taser.

I believe a comprehensive solution is needed. People have suggested lots more tasers (or Justin's personal-taser idea), which might work, but is clunky. I don't have a good a good suggestion now, but let me offer some small fixes which will only somewhat help.

- Make Tased Mafia unable to conscript. I had assumed this way the rule until Isaac pointed it out to me. The point of a Taser is to nullify a Mafia and let everyone come out of hiding in their rooms. The threat of a Conscript nullifies this.
- Make a Mafia who is lynched fail to recruit a rogue that day. It's a freebie otherwise. It also makes sense with how other day-end powers work: when Asya was lynched, she did not die as Gay Knight or else it would have been revealed.
- Make failed kills use up the Mafia's daily kills. Otherwise, it's free for revealed Mafia to run into Medkits, superheroes, conspiracy theorists, and cynics. Ironically, a revealed Mafia *wants* to him them.

PostPosted: Fri Oct 17, 2014 11:35 am 
++ to all Alex's suggestions.

In particular, I think making failed kills use up daily kills would do a great job of reducing kamikaze danger, since a lot of kamikazes happen after someone fails to kill someone.

I also agree that it's quite surprising that game ended with only about a quarter of town alive, given that town killed mafia aligned players on 11/14 days of game (not counting day 1, since nothing happened).

PostPosted: Fri Oct 17, 2014 6:57 pm 
I agree with suggestions 1 and 2, and strongly disagree with suggestion 3 town does not need an additional reward for catching mafia, it also removes from the interest, if town notices a day without a kill, it becomes clear that the person probably lied (it's not worth mafia putting themselves in this much pressure almost ever) if it is genuinely unclear, you lynch both of them almost certainly, it's not actually that subtle or interesting. Having a large number of kills is in fact, fundamentally a feature of having larger games, town this game, while it played very well, was incredibly poor with it's use of tasers, town does not have a right to stop mafia from sacrificing themselves, while I agree the multi-day publicly revealed mafia situation is somewhat silly (though I think how negative it is is dramatically overstated, it's a possibly minor annoyance, it's not game breaking in any way) I feel that relatively simple fixes (allowing occasional double lynches instituted by the mayor say) are more than sufficient to solve it. If town is especially unhappy about the possibiity of kamikazes they are free to use prophets to stop them. I also feel like this proposal makes game feel much more trappy for mafia. If we're intending to reduce the trapiness of game (which is largely how I view the anti-investigation impetus) I don't think that making kill fails stronger is a good way to do it. I also think that allowing mafia to go out with a bang serves the valuable purpose of making mafia who are caught or likely to be caught still have a valuable purpose to the mafia allowing mafia who have made relatively minor mistakes (i.e. victor running for police officer) able to make the choice to self sacrifice, an interesting strategical decision, killing key players is an admitted problem though I feel like the solution to this should be broader (cynics also partially serve this purpose).

PostPosted: Sat Oct 18, 2014 11:40 am 
An intermediate idea would be that a failed kill does not use up team mafia's daily kill(s), but that the mafia who made the failed kill is not able to make another kill that day.

Even with that modification, I don't like this particular idea. I think my distaste stems from my intense dislike of trap roles (by which I mean mic/receivers + anything of the form 'I do not die'.) I don't want to make running into a trap much more sucky than it already is.

More importantly, while I am NOT a fan of unstoppable kamikaze rampages, I think tasers do give the town the necessary tool to deal with abusive cases. I'd be open to modestly increasing the taser-count. I also think that modifying the lynch mechanic is an intriguing idea, in particular to allow for the possibility of multiple lynches in one day.

My preferred way to deal with trap roles would be to make traps few and far between. I would eliminate desperados, mic/receivers, and medkits. Then, only conspiracy theorists and cynics would have potential immunity, and they have a relatively low chance of catching a mafia unless they play cleverly. We could even get rid of kabooms at that point.

Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 26 posts ] 

All times are UTC - 5 hours

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group