[REVIEW] Open Setup Reviews

[Cogito Ergo Sum]

Alright. Say we go ahead and run this, are we allowing double ups on roles? Can there be three N1 Docs for example?

Yes. (The set-up doesn't actually make sense if you didn't allow it.)

I still get the feeling that town will spend their time on D1 trying to break the game instead of playing mafia, regardless of whether they find a beneficial claiming strategy.

Hmm. Let me think about it. Isn't that a problem a lot of Open set-ups are going to face regardless?

[Hoopla]

Isn't that a problem a lot of Open set-ups are going to face regardless?

To a degree, but good open design shouldn't feature an early massclaim/claiming strategy or a synchronisation of actions. Open setups can be pretty limited in some senses, as the nature of confirmed innocents and optimising night actions can be so powerful, there needs to be incentive for people to not declare their roles, whether that is through uncertainty/randomness of the setup, scarcity of PR's or a mechanic that makes claiming suboptimal (like the mechanic in Masons and Mafia).

I don't really see anything that is preventing town from thinking that they can get an edge by trying to optimise night actions than play it blind. There is some randomness, but I think there is enough probable information to make networking of night actions profitable, imo. I'm not sure of an exact strategy, but if all the information is out there, I think you could organise actions in a certain way to test combinations of players and force scum to shoot certain people.

[Cogito Ergo Sum]

You know I know all that stuff, right? The "problem" I mean is that it's going to be a distracting topic of discussion early on.

I do not believe there's sufficient networking possible to overcome the inherent downsides of mass claims. Seems to me like the main thing mass claim will accomplish (excepting a 5-8 spread of claims) is telling the scum who to shoot. There's too many players to protect and too little protective power for that not to be case, as I see it.

[quadz08]

From what I can tell of this discussion (and I may be wrong), it seems that you both think massclaiming isn't necessarily optimal here. The difference is that Hoops wants to prevent it from Game 1, while CES seems to think that letting it organically grow into the meta for this particular setup is the way to go.

I think Hoops' method is the better one in this case, for the record.

[Hoopla]

You know I know all that stuff, right?

Nope.

~~

You can't protect everyone - but you can protect the most valuable players (such as claimed Cops for that night or confirmed innocents), which forces scum to shoot from a less desirable pool of players, and when scum need to use their nightkill with great efficiency in this setup, you can see why even one or two players being safe for the first couple of nights is valuable. If there are more innocents/Cops for that night alive than protections available (even if scum is the claimed Doc for that night, they can't just go ahead and kill the towniest player), you can have those claimed Doctors flip a coin between who to save which still partially nullifies scum's NK if they don't want to take a chance.

I think a sensible plan is for N1 Cop(s) to investigate N5 Docs/Cops, which creates innocents out of the least desirable NK choices for scum, enabling greater use out of the N2/N3 brigade. N2 Cop(s) then investigate the N1 Cops to see if their information is reliable.

Like I said before, it seems obvious to me that town will employ a claiming strategy of some kind regardless of whether it is actually beneficial.

[Hoopla]

From what I can tell of this discussion (and I may be wrong), it seems that you both think massclaiming isn't necessarily optimal here.

I'm not sure what I think, but I probably wouldn't oppose a massclaim as town if I were playing.

[Hoopla]

While I wait for more input on the Diffusion of Power setup, I'll bring up a different setup.

Upon seeing Xalxe's Marathon game of Switch, it reminded me I wanted to reintroduce this setup into the mix, because it isn't broken or overly unbalanced - I think the Switch mechanic needs tweaking though.

It doesn't make sense for the SK to have the opportunity to veto every switch, particularly when it's vulnerable to the Cop and the Vig at night. It means the decisions at night are fairly automatic, as the SK is more at risk than the mafia by gambling on letting the Cop or Vig use an action (presuming the mafia don't Switch them off) - the SK should leave both those switches off each night.

I think a much better mechanic is the Mafia having control of the switches while the relevant mafiosi is alive, but the SK instead of vetoing the switches it wants off, simply reverses the state of the switches that the scumteam leaves them in. This makes a more interesting dynamic at night that isn't completely predictable. It's also deliciously wifomy.

I like the night start for the novelty of it, but I think the SK should be innocent to the Cop for balance reasons - it is already up against potentially two kills a night and no kill immunity, so it should at least be safe from the Cop. I also think with a nightstart, 13p (an extra VT) would make the game a little less swingy overall and slightly reduces the chances of a D2 massclaim being beneficial to town if they catch the right set of night actions on N0 and N1 (one more VT slightly thins the ratio of confirmable town roles to non-confirmable roles - a problem that sometimes can affect open setups).

[Tierce]

I think a much better mechanic is the Mafia having control of the switches while the relevant mafiosi is alive, but the SK instead of vetoing the switches it wants off, simply reverses the state of the switches that the scumteam leaves them in. This makes a more interesting dynamic at night that isn't completely predictable. It's also deliciously wifomy.

I poked at the setup when Xalxe ran the marathon and this was an idea that came to me. Seems feasible, because this way both the scum and the SK are left wondering what the other side would do and have to plan accordingly, both on night actions and claims. It's a more interesting result to analyze and leads to a level of gameplay I'd like to try.

[izakthegoomba]

So the mafia and the SK form a kind of XOR gate for the PRs?

That's practically bastard... I love it!

[IceGuy]

I don't like the SK being investigation immune. He should be affected by the PRs just as the scum team. I would make him NK-immune (but not vig-immune).

Another potential problem I see is the predictability of the scum's actions. For instance, the game will almost always start with the cop and the vig being switched off, and the doctor switched on (the doc switch mafioso making the kill).

Also, I'd really, really like to bring this to a 13-player, daystart game.

[Staeg]

No, the night actions are not predictable because the SK can reverse them

[IceGuy]

I mean predictable for the SK.

[quadz08]

I think that the SK should be 1-shot bulletproof. That way, he still cares about the vig-switch, while balancing him out a bit.

Additionally, I think limiting him to a maximum of 3 actions per night (all 3 switches and no kill, 2 switches and a kill, or any combination of fewer switches+kill) could be a good idea. Not sure how that works out, balance wise.

[Hoopla]

Another potential problem I see is the predictability of the scum's actions. For instance, the game will almost always start with the cop and the vig being switched off, and the doctor switched on (the doc switch mafioso making the kill).

How is the SK to know what state the Cop/Vig switch is left in? The SK reversing the state of the switch means there are no guarantees.

[Hoopla]

Additionally, I think limiting him to a maximum of 3 actions per night (all 3 switches and no kill, 2 switches and a kill, or any combination of fewer switches+kill) could be a good idea. Not sure how that works out, balance wise.

This only makes a difference when all three switches are alive.

[IceGuy]

How is the SK to know what state the Cop/Vig switch is left in? The SK reversing the state of the switch means there are no guarantees.

He won't know, but it's most beneficial for both parties in the beginning.

[Hoopla]

If the SK has immunity to the Cop, it's beneficial for the SK to leave the Cop active, while the mafia want it inactive - that creates a guessing game. Mafia should like having the vig active more than the SK, because the mafia has the security of a three player team over a lone SK - this could create a guessing game. It's in scum's interests to kill the SK as soon as possible to have control of all the switches, so leaving the Vig active is in general a +EV move if they can manage it.

[Staeg]

I disagree with mafia wanting to leave the vig on because of his chance of hitting the SK - the mafia are also more than three times as likely to get shot (due to associative tells and all).

[Hoopla]

Perhaps, but it should be in the mafia's interest to make the game as short as possible - as a general rule, the more days the game goes for, the easier it is to find scum during the day. Games that have a good chance of not going beyond D3 typically favour scum.

[Hoopla]

I don't like the SK being investigation immune. He should be affected by the PRs just as the scum team. I would make him NK-immune (but not vig-immune).Another potential problem I see is the predictability of the scum's actions. For instance, the game will almost always start with the cop and the vig being switched off, and the doctor switched on (the doc switch mafioso making the kill).

Alright - how about this solution to minimise predictability:

SK pregame gets to choose one of these perks:

1) Investigation Immunity AND Vig Immunity
2) Mafia NK Immunity

~~

I was originally going to suggest a choice between Investigation Immunity and 1-Shot BP, but kill protection is the obvious choice. This seems to be balanced perks, as the Cop/Vig could both be dead by D2 anyway, making having double-perks irrelevant. The second choice only gives one immunity, but you'll probably be facing at least three or four mafia NK's to win the game.

This also has the benefit of mafia not knowing whether the SK wants the Vig/Cop turned on or off, which now ensures the switch dynamic isn't stale.

[Cogito Ergo Sum]

Okay. I've come up with a unintrusive if inelegant change to Diffusion of Power that should remove mass claim's most meaningful benefit - tell the scum how many Cops and Docs there are. That way the 5-8 split should always be avoided; the town can gain no information from timing of claims (e.g. if a claim makes for a 7-5 split, that claim would look townie otherwise); scum can even all claim the same role if there are only 4 of those which means the only way for anyone to be cleared by set-up logic is if the town lynches 2 scum from a group of 6.

Just flipping Doc/Cop instead of Nx Doc/Nx Cop prevents hypocopping by the bye.

[Hoopla]

I think both of those solutions could work.

[callforjudgement]

So, I think I found a breaking strategy for scum in Scumhunters 8p. Day 1, one of the scum claims in-thread.

I know this seems ridiculous; but Lovers' Mafia (4:2 nightless white flag) is considered balanced; and if the claimed scum gets lynched (which they probably will be, because town has no motivation to claim scum), the game becomes 4:2 nightless white flag, but with the strongest townie missing (they get NKed), and weakest scum missing (because that's who the scumteam will sacrifice). So scum can drive the game to a scumsided setup without much of a problem. Also, scum claiming D1 means that there's no way you can use D1 play to tell who the scum are because everyone has the same information.

Town

might

be able to defeat this by lynching someone else instead, in which case the setup is effectively 5:2 nightless in which the town win outright on a correct lynch D1. I think this is also scumsided, but am less sure. Also note that the lynch threshold is higher by 1 day 1, due to the claimed scumbag (who is presumably trolling and refusing to vote, or self-voting, in order to not give away information). This probably doesn't really matter so much, though, unless the game goes to plurality lynch (in which case it definitely does).

[Alchemist]

Hoops, can you make me some interesting small-player opens for skype/ventrilo mafia? :3

[Hoopla]

So, I think I found a breaking strategy for scum in Scumhunters 8p. Day 1, one of the scum claims in-thread.

I know this seems ridiculous; but Lovers' Mafia (4:2 nightless white flag) is considered balanced; and if the claimed scum gets lynched (which they probably will be, because town has no motivation to claim scum), the game becomes 4:2 nightless white flag, but with the strongest townie missing (they get NKed), and weakest scum missing (because that's who the scumteam will sacrifice). So scum can drive the game to a scumsided setup without much of a problem. Also, scum claiming D1 means that there's no way you can use D1 play to tell who the scum are because everyone has the same information.

Town

might

be able to defeat this by lynching someone else instead, in which case the setup is effectively 5:2 nightless in which the town win outright on a correct lynch D1. I think this is also scumsided, but am less sure. Also note that the lynch threshold is higher by 1 day 1, due to the claimed scumbag (who is presumably trolling and refusing to vote, or self-voting, in order to not give away information). This probably doesn't really matter so much, though, unless the game goes to plurality lynch (in which case it definitely does).

The game always requires town to lynch two scum throughout the game - if they do it on D1, they only need one more from two possible scum. If they miss D1, they need two from two. Both outcomes from the D1 lynch involves the scumteam strengthening their team, as they either get rid of the weakest teammate or the strongest townie. I don't think giving town a 4:2 whiteflag scenario is optimal, as normal 4:2 is clearly better. You brought up the town's counter-strategy, in the sense, they get an autowin if they lynch a different scum than claimed scum D1. I don't think claiming scum is a good idea, really.

Hoops, can you make me some interesting small-player opens for skype/ventrilo mafia? :3

I don't really know what's balanced for face-to-face and voice-chat mafia. We have a wiki of interesting setups you're welcome to experiment with.