Normal Game Discussion Thread (Easter 2025 Update)

[biancospino]

This is the sort of things that would really benefit from using something like Triplicate Action Resolution.

But, at any rate, it isn't actually a cyclic dependence: Trackers check whether their target is performing a (non-roleblocked) targeting action, not whether that action has already started to resolve.

[rBree2]

Trackers see each other visit the other because nothing causes their actions to fail.

There’s no need for a priority in a tracker loop because action investigation actions receive their results after all other actions have been accounted for, essentially. Their results aren't sent in the middle of the night, they are sent at the end of the night.

You can think of this as Tracker A planting a tracker device on Tracker B, then going home, whereas Tracker B instead traces Tracker A's movements throughout the night.

Tracker A's tracking device sees Tracker A visit Tracker B; Tracker B's tracing of Tracker A's steps leads them to the conclusion that Tracker A visited Tracker B.

They both get the information regardless.

[biancospino]

Another interesting, vaguely related question is whether Trackers can see actions targeting a commuting Commuter. Traditionally Commuters cause actions targeting them to fail to even target (not merely be blocked (or be protected from depending on your stance)), so Trackers should not see them right?

[TemporalLich]

Mutual tracking sounds like a nonflict to me

it is technically an action conflict but said action conflict has an unambiguous resolution (i.e. both trackers get the result that their target targeted them) so said resolution should be a given (also, I am very sure this is what happens under Reasonable Action Resolution)

p-edit: I would be inclined to say a Commuter "superprotects" themself but there actually is an argument based on flavor for why an action missing due to Commuter fails to visit - the Commuter is just gone lol

on the other hand - the Tracker would just track the player to the Commuter's house which is good enough for a "Player targeted Commuter" result - so flavor fails to settle whether commuted actions visit or not

[biancospino]

While we're talking priority brackets, I would also like to suggest that Checker (which is now classified as miscellaneous) should be classified as a role-investigative. Roles-investigatives work by informing the user whether or not the target matches some property, and have a special message should they fail; Checkers also work in the exact same manner, where the property is the trivial property that always matches.

[TemporalLich]

Checker actually could be resolved as miscellaneous NAR-wise if we're only using the NAR Ladder as "these actions should resolve like this type of action because {reason}"

but mechanically, Checker is an investigative that learns nothing useful from the investigation - Checker is the blank "expected to receive results PMs and thus must be sent No Result if the action fails" role

[biancospino]

Yeah exactly. Followers and somesuch should see it as an investigative. (Not that it is an overly important point thou.)

Changing topic entirely, simultaneous is classified as a modifier, but I'm not sure it actually is. If it is indeed a modifier (as opposer to a role), that would make things like "(simultaneous Doc Cop) Vig" legal (a role that can either kill or both investigate and protect), and would make things like "simultaneous Townie" (a role that has no abilities but, if it gained some somehow — likely due to inventions — it could use multiple at the same time) illegal.
But the ability text suggests that both of those thing should be the reverse.

Which actually brings me to another point: it is sometimes the case that a modifier would be better used as a role; for instance, imagine a "Cop Doctor Roaming", who can either protect or investigate but can never target the same player twice. This is a meaningful role which cannot be readily simulated with the roaming mod (beside doing convoluted stuff like "roaming JoAT(Cop x1000, Doc x1000)", and even then it doesn't work the same for Inventions).
So the suggestion is to do the following:

• allow modifiers to apply to groups of roles rather than to single roles only. So a "roaming (Cop, Doctor) Vig" can never target the same player with a Cop-or-Doc ability but has no restrictions on the Vig ability nor on any Inventions it may receive".
  
• allow modifiers to be explicitly declared to apply to all abilities. This would promote the mod to a Role; so in theory one should invent a linked Role for that, say Hexed. So that a "Cop Doc roaming-Hexed" can never target the same target twice, neither with its own abilities nor with inventions.
  But if a whole new role is undesirable for such s niche thing, maybe just using notations like "roaming(*)" may work.

This would allow for both versions of simultaneous while keeping it a mod (by making the "simultaneous Townie" example into a "Town simultaneous-Hexed").

[TemporalLich]

Simultaneous is classified as a modifier under the logic of "if it doesn't make sense to stand alone, it is a modifier" despite acting like a role

i.e. Simultaneous Townie is not a role that makes sense except if you want a Named Townie or in the rare case of an Inventor-heavy setup

A modifier is intuitively defined as something that changes how a role works. Simultaneous is not a modifier according to the intuitive definition, and this means Simultaneous acts like a role allowing multiple roles to be used at once

Simultaneous as a modifier would indeed allow for a Vigilante Simultaneous (Cop, Doctor) to exist which can either kill a player, or heal a player and investigate a player simultaneously. I don't feel this loss of design space is too much of an issue though

MafiaWiki classifies Simultaneous as an interaction modifier which means it changes how roles are used together, Combined and Split are also interaction modifiers



as for how modifiers could be applied, I think that they could be applied to individual roles, groups of roles, or generally. Modifiers in an intuitive sense modify how roles work, and some modifiers like X-shot and Night-Specific care about how it is applied

Tracker Roaming Doctor can track players, or heal players they haven't tried to heal before.
Tracker 1-shot Doctor can track players each night or heal once in the game at night

Roaming Tracker Roaming Doctor can track players they haven't tried to track before, or heal players they haven't tried to heal before
1-shot Tracker 1-shot Doctor has one usage of Tracker and one usage of Doctor, and is equivalent to Jack-of-all-trades (Tracker, Doctor)

Roaming (Tracker Doctor) (which would likely just be written as Roaming Tracker Doctor) can track or heal players they haven't tried to track or heal before
1-shot (Tracker Doctor) has one usage of Tracker

or

one usage of Doctor, and is equivalent to 1-shot Jack-of-all-trades (Tracker, Doctor)

[Roaming] Tracker Doctor (or Roaming(*) Tracker Doctor) works the same way as Roaming (Tracker, Doctor) but inventions also count for Roaming - Roaming is intuitively understood to be its general version which would be formalized as [Roaming]
[1-shot] Tracker Doctor has one usage of Tracker or one usage of Doctor or one usage of an invention - general 1-shot (notated as [1-shot] here) means you only get to act once in the entire game even with the help of inventions

Note that modified versions of non-hybrid roles are assumed to be using the specific version of the modifier and not the general version - if the role PM says Loyal Neighborizer it's safer to assume it means Loyal Neighborizer as opposed to [Loyal] Neighborizer

As for Simultaneous and Combined, this level of modifier formalization means that Simultaneous and Combined are usually used as general modifiers (i.e. [Simultaneous] and [Combined]) and are modifiers that either need to be general or need to encapsulate at least two roles

[biancospino]

Combined is actually not usually understood as [combined]: if a "combined Reporter Inspector" receives a FV invention, then I would not expect their combined action to suddenly also give a fruit. But it would if it was [combined] instead. (Also in theory a "combined X", with only one modified role, makes syntactical sense (but not really conceptual sense): it's a role whose action combines all actions in the set {X}; ofc, being a singleton, it's identical to the unmodified X.)

(Formalizing this difference between mod and [mod

], if recognizing that [mod

] is indeed a Role, also opens design space for exotic things like a [weak]-Inventor (your target's next targeted action behaves as thou it had the modifier "weak") etc.), if Inventor is allowed to give out passives.

[TemporalLich]

Combined is actually not usually understood as [combined]: if a "combined Reporter Inspector" receives a FV invention, then I would not expect their combined action to suddenly also give a fruit. But it would if it was [combined] instead. (Also in theory a "combined X", with only one modified role, makes syntactical sense (but not really conceptual sense): it's a role whose action combines all actions in the set {X}; ofc, being a singleton, it's identical to the unmodified X.)

whoops sorry

yeah Combined is actually understood as encapsulating all roles actually

Combined Doctor and Simultaneous Doctor are technically not malformed roles (Activated Townie is a malformed role) but those modifiers are meaningless as Interaction Modifiers need to either be general or encapsulate at least two roles

(Formalizing this difference between mod and [mod

], if recognizing that [mod

] is indeed a Role, also opens design space for exotic things like a [weak]-Inventor (your target's next targeted action behaves as thou it had the modifier "weak") etc.), if Inventor is allowed to give out passives.

General modifiers would outright make Renegade Normal through the Inventor role.

I'm leaning against general modifiers being Normal (though I am willing to recategorize Simultaneous as a role so it can be a general modifier) because of these reasons:

1. General modifiers and modifiers encapsulating all roles act the same in a game with no Inventors (unless they apply to factional abilities - see reason 4)
2. An "all role" modifier usually can be written as {modifier} {role1} {role2}, X-shot is a notable exception and must be written out as X-shot ({role1} {role2}) for the all role version
3. [Simultaneous] would be the most desired general modifier, and it is already possible to argue that Simultaneous is a role
4. General modifiers could potentially modify factional actions - factional action modifiers are allowed only as specific roles afaik (currently only Juggernaut and Ninja exist)

[DragonEater70]

I think roleblocks should just cancel out, while other actions should still go.

I would venture a more sordid solution.

I suggest proceding thus:

• First try to use the priority rules. If this is enough, all good. Otherwise, if a paradox still exists, proceed. (Personally I would even go as far as abolishing priority rules entitely, but I don't think I'll have many buyers)
  
• Call a subset of all actions
  acceptable
  if there is no action in that subset which would be blocked by an action in that subset if all actions in that subset succeeded.
  
• Call the
  badness
  of an acceptable subset the number of actions which are not in it that are not blocked by an action in it. Fixed an acceptabld subset, call
  canceled
  an action that is not in that subset but is not blocked by an action in that subset.
  
• Among all acceptable subset, consider only those with minimal badness. If only one such subset exists, actions in that subset succeed (in the order required by priority and the golden rule, if relevant) and the other fail. Otherwise, proceed.
  
• If there are multiple acceptable subsets of minimal badness, use the following criteria, in order, until only one remains:
  1. Prefer larger subsets (those that make less actions fail)
     
  2. Prefer subsets with the lowest number of players who have a canceled action.
     
  3. Define the nth characteristic priority of an acceptable subset as the nth higher priority of a canceled action wrt that subset, if any such action exists (or take an infinitely low priority otherwise). Prefer subsets with a lower first characteristic priority, if equal a lower second characteristic priority, etc.
     
  4. Prefer subsets so that the first highest priority among actions not in it is the lowest possible. If equal, the second highest, etc.
     
  5. At the beginning of the game (as part as the setup), assign a different number to each possible ability that may be used by each slot during the game. Evaluate criteria iii and (if necessary) iv, but using those numbers as thou they were the priorities. This is the nuclear option if nothing else solves the tie, and is guaranteed to work since all actions have a different number. (In non-normal games, I would suggest rolling a die instead.)
  This is a little convoluted but results in the same results as the current system for all but the most egregeous of messes. In the "Town RB blocks scum RB who blocks the TRB and kills" situation, it would result in the kill going through (The two possible acceptable subsets of minimal badness — with badness 0 in this case — are {trb} and {srb, nk}, but the second is larger and so wins for criterium (i).
  The only real issue I see is that deciding the correct resolution would be (I think) O(2^n), but the number of problematic actions in any given game is unlikely to ever be greater than a handful, so I don't see this being a huge problem.

Sorry for the late reply, and maybe I'm completely misunderstanding you, but are you saying that a certain outcome has a higher badness if LESS actions are blocked? Why do you think that's a correct criterion to evaluate in adjudicating a resolution? I.e. why do you aim to have the highest number possible of blocks?

[DragonEater70]

Because I don't like the simple solution.
Imagine a multitasking RB X and a multitasking RB Y target each other (and use their X, Y abilities in some way). I strongly feel that exactly one of X and Y should go through*, but the simple solution ("cancel out blocks") would have both of them go through.

* This is because "the first RB (call them A) succeed, the other fail (call them B)" is consistent (for B is blocked by A, and A is blocked by nobody on account of B being blocked), and so is the reverse. But "A & B both being partially blocked" is not consistent at all (RB's ability is not a partial block!), and "A & B both entirely fail" is not consistent either (nobody is blocking A — since B's block has failed — so they should not be blocked!).
"Acceptable with 0 badness" is the closest approximation I can formulate of consistent; but sometimes no such consistent resolution exists, and I feel allowing a small number of cancellations is the least offensive way to cut the gordian knot.

Okay, it might be consistent in a mathematical or purely logical sense, but it's also very unintuitive. If both are the same role (Roleblockers), why should one have priority over another?

I think that in action resolution, the goal shouldn't be to find the most "logically consistent" resolution but the most intuitive resolution such that players could easily predict how things are resolved, given a certain situation.

[biancospino]

Sorry for the late reply, and maybe I'm completely misunderstanding you, but are you saying that a certain outcome has a higher badness if LESS actions are blocked? Why do you think that's a correct criterion to evaluate in adjudicating a resolution? I.e. why do you aim to have the highest number possible of blocks?

Quite the opposite actually, that criterion tried to minimise spurious blocks; the

badness

there is the number of actions that fail "arbitrarily" (i.e. due to the resolution rules as opposed to due to a succesfull block).

I think that in action resolution, the goal shouldn't be to find the most "logically consistent" resolution but the most intuitive resolution such that players could easily predict how things are resolved, given a certain situation.

Yeah, that's kind of the sentiment I arrived to more or less agree with in the end. That's kind of a balancing act thou, since clearly "logical consistent" and "easy to figure out" are both desirable but not necessarily concordant. The catastrophic rule (or the "opposite" rule that only cancels blocks instead of all actions) is probably at one extreme end of that scale, but indeed it seems hard to move from that end without degenerating into something too complex to administer.

[DragonEater70]

Anyway I think the Catastrophic Rule is fine *if and only if* it's applied when both the golden rule and priority list failed. I think the roleblocker being stronger than JK ethos is ingrained in site culture and setup balance (at least NewD3) and shouldn't be discarded.

[DragonEater70]

For some reason I seemed to believe that Reasonable Action Resolution both provided a solution *and* kept the priority list, but nope, it explicitly says that in a JK vs Mafia Roleblocker situation the JK prevents the kill.

[DragonEater70]

I don't think this situation is ambiguous unless I'm misunderstanding. The strong-willed roleblock, by nature of being strong-willed, cannot be affected by the mafia roleblocker's block. Hence, by the golden rule of NAR, the SK's roleblock resolves first (because it cannot be affected by the mafia's roleblock or kill) and prevents the mafia roleblocker's roleblock action from happening (as well as the mafia's kill). Since the mafia roleblock didn't go through, the SK's kill does go through. The catastrophic rule does not need to be invoked because there's no ambiguity; it only exists as a fallback when the golden rule fails.

actually yeah this makes more sense, the golden rule resolves a strong-willed roleblock before a non strong-willed roleblock as a strong-willed role can't be roleblocked

though I think two strong-willed roleblocks in a roleblock loop invoke the catastrophic rule while remaining successful

Wouldn't two strong-willed blocks simply succeed? So both kills would be blocked.

[DragonEater70]

Since we like debating action resolution, here's another thing that I thought of recently, I feel I may be overthinking it, but it seems ambiguous under NAR. To illustrate the issue, I will first provide an example where there is no ambiguity and then provide a modified scenario that has the ambiguity.

Scenario 1

A tracks B.
B tracks C.
C does nothing.

I think we would all agree that the result of this scenario is that A sees that B targeted C while B sees that C didn't go anywhere. If we apply the rules of NAR, it would be resolved in this fashion:

• First, we need to find an action that can't potentially be modified by other actions. In this case, A's action is potentially modified by B's action. B's action however is not potentially modified by anything. Thus, B's action resolves first.
  
• Resolve B's action: B sees that C didn't go anywhere.
  
• Now, we are only left with A's action which we can resolve and results in A seeing that B targeted C.

Scenario 2

A tracks B.
B tracks A.

This is basically the same scenario except this time, the tracking is mutual. I think most people's intuition here would say that both A and B would receive results: A would see that B targeted A, and B would see that A targeted B. In fact, I would bet most people would resolve this that way without thinking about how it would work under NAR. However, if we apply the rules of NAR:

• First, we need to find an action that can't potentially be modified by other actions, but wait! There is no such action here! A's action is potentially modified by B's action (because it affects whether A sees B's action) while B's action is potentially modified by A's action (for the same reason).
  
• Next, we attempt to apply the priority ladder, but that doesn't resolve the conflict either because both actions are action investigation actions (in fact, both actions are the exact same ability).
  
• If we were to resolve A's action first, then A would fail to see B's action (because B's action hasn't happened yet). However, B would see A's action.
  
• If we were to resolve B's action first, then we have the opposite scenario: B would fail to see A's action, but A would see B's action.

I don't really see any tracker action as interfering with another tracker action. The generally accepted way to resolve things is that "all actions resolve simultaneously". The golden rule and the priority list are only meant to be referred to when "all actions are resolved simultaneously" is ambiguous (in my opinion / extrapolated based on my own and other people's experiences). But A targeting B at the same time that B targets A is not really ambiguous in the case of tracking as there's no interference.

[TemporalLich]

I don't think this situation is ambiguous unless I'm misunderstanding. The strong-willed roleblock, by nature of being strong-willed, cannot be affected by the mafia roleblocker's block. Hence, by the golden rule of NAR, the SK's roleblock resolves first (because it cannot be affected by the mafia's roleblock or kill) and prevents the mafia roleblocker's roleblock action from happening (as well as the mafia's kill). Since the mafia roleblock didn't go through, the SK's kill does go through. The catastrophic rule does not need to be invoked because there's no ambiguity; it only exists as a fallback when the golden rule fails.

actually yeah this makes more sense, the golden rule resolves a strong-willed roleblock before a non strong-willed roleblock as a strong-willed role can't be roleblocked

though I think two strong-willed roleblocks in a roleblock loop invoke the catastrophic rule while remaining successful

Wouldn't two strong-willed blocks simply succeed? So both kills would be blocked.

either way a Mafia Strong-Willed Roleblocker and a Serial Killer Strong-Willed Roleblocker blocking each other means neither the Mafia (assuming the roleblocker attempts the kill of course) nor the SK kill succeeds - either they're blocked or they're deleted by the Catastrophic Rule - a mutual strong-willed block succeeding (due to Catastrophic Rule failing to delete them) would be the intuitive way to resolve that which makes the blocks visible with tracking or following

What would be a point of contention would be if one of those players were a Mafia Simultaneous Juggernaut Strong-Willed Roleblocker or a Town Simultaneous Strong-Willed (Vigilante Roleblocker), does the strong kill succeed or is it deleted by the Strong-Willed Catastrophic Rule (if that even is a thing)?

[TemporalLich]

I don't really see any tracker action as interfering with another tracker action. The generally accepted way to resolve things is that "all actions resolve simultaneously". The golden rule and the priority list are only meant to be referred to when "all actions are resolved simultaneously" is ambiguous (in my opinion / extrapolated based on my own and other people's experiences). But A targeting B at the same time that B targets A is not really ambiguous in the case of tracking as there's no interference.

An overly formal interpretation of NAR would say that "action investigation actions are modified by every action" (hence their position at the bottom of the Priority Ladder and why Detective and Psychologist are allowed to be in the Role Investigation priority bracket for NAR purposes (ignoring the fact that Detective/Psychologist see kill attempts that fail to even visit)) and thus would need to resolve tracking loops with the Silver Rule (though you see it as more of a Platinum Rule I think?) that says "all actions are resolved simultaneously"

[TemporalLich]

For some reason I seemed to believe that Reasonable Action Resolution both provided a solution *and* kept the priority list, but nope, it explicitly says that in a JK vs Mafia Roleblocker situation the JK prevents the kill.

yeah, both RAR and Greedy Cata would mean the kill is blocked in the "NewD3 Interaction" (the "NewD3 Interaction" is where the Mafia Roleblocker performs the factional kill and blocks the Town Jailkeeper, with the Town Jailkeeper jailing the Mafia Roleblocker)

Anyway I think the Catastrophic Rule is fine *if and only if* it's applied when both the golden rule and priority list failed. I think the roleblocker being stronger than JK ethos is ingrained in site culture and setup balance (at least NewD3) and shouldn't be discarded.

You want the "NewD3 Interaction" to resolve the same way it does in the actual NewD3 setup with the kill succeeding

this would require Lazy Cata, which is making the Catastrophic Rule resolve after the Priority Ladder and keeping the Jailkeep priority bracket in place

Both Lazy Cata and Rolestopper is Protective is needed for the "NewD3 Interaction" to resolve with a successful kill with a Town Alien instead of Town Jailkeeper, as Alien would be in the Jailkeep priority bracket with Rolestopper is Protective

I'm advocating for Greedy Cata which is making the Catastrophic Rule resolve before the Priority Ladder (and would allow for the Jailkeep priority bracket to be merged into the Blocking priority bracket)

[DragonEater70]

I don't think this situation is ambiguous unless I'm misunderstanding. The strong-willed roleblock, by nature of being strong-willed, cannot be affected by the mafia roleblocker's block. Hence, by the golden rule of NAR, the SK's roleblock resolves first (because it cannot be affected by the mafia's roleblock or kill) and prevents the mafia roleblocker's roleblock action from happening (as well as the mafia's kill). Since the mafia roleblock didn't go through, the SK's kill does go through. The catastrophic rule does not need to be invoked because there's no ambiguity; it only exists as a fallback when the golden rule fails.

actually yeah this makes more sense, the golden rule resolves a strong-willed roleblock before a non strong-willed roleblock as a strong-willed role can't be roleblocked

though I think two strong-willed roleblocks in a roleblock loop invoke the catastrophic rule while remaining successful

Wouldn't two strong-willed blocks simply succeed? So both kills would be blocked.

either way a Mafia Strong-Willed Roleblocker and a Serial Killer Strong-Willed Roleblocker blocking each other means neither the Mafia (assuming the roleblocker attempts the kill of course) nor the SK kill succeeds - either they're blocked or they're deleted by the Catastrophic Rule - a mutual strong-willed block succeeding (due to Catastrophic Rule failing to delete them) would be the intuitive way to resolve that which makes the blocks visible with tracking or following

What would be a point of contention would be if one of those players were a Mafia Simultaneous Juggernaut Strong-Willed Roleblocker or a Town Simultaneous Strong-Willed (Vigilante Roleblocker), does the strong kill succeed or is it deleted by the Strong-Willed Catastrophic Rule (if that even is a thing)?

I would rule a strong willed kill succeeds.

[DragonEater70]

I don't really see any tracker action as interfering with another tracker action. The generally accepted way to resolve things is that "all actions resolve simultaneously". The golden rule and the priority list are only meant to be referred to when "all actions are resolved simultaneously" is ambiguous (in my opinion / extrapolated based on my own and other people's experiences). But A targeting B at the same time that B targets A is not really ambiguous in the case of tracking as there's no interference.

An overly formal interpretation of NAR would say that "action investigation actions are modified by every action" (hence their position at the bottom of the Priority Ladder and why Detective and Psychologist are allowed to be in the Role Investigation priority bracket for NAR purposes (ignoring the fact that Detective/Psychologist see kill attempts that fail to even visit)) and thus would need to resolve tracking loops with the Silver Rule (though you see it as more of a Platinum Rule I think?) that says "all actions are resolved simultaneously"

Yes, I see it as an unwritten platinum rule. That's why kills don't prevent other actions for instance.

[DragonEater70]

For some reason I seemed to believe that Reasonable Action Resolution both provided a solution *and* kept the priority list, but nope, it explicitly says that in a JK vs Mafia Roleblocker situation the JK prevents the kill.

yeah, both RAR and Greedy Cata would mean the kill is blocked in the "NewD3 Interaction" (the "NewD3 Interaction" is where the Mafia Roleblocker performs the factional kill and blocks the Town Jailkeeper, with the Town Jailkeeper jailing the Mafia Roleblocker)

Anyway I think the Catastrophic Rule is fine *if and only if* it's applied when both the golden rule and priority list failed. I think the roleblocker being stronger than JK ethos is ingrained in site culture and setup balance (at least NewD3) and shouldn't be discarded.

You want the "NewD3 Interaction" to resolve the same way it does in the actual NewD3 setup with the kill succeeding

this would require Lazy Cata, which is making the Catastrophic Rule resolve after the Priority Ladder and keeping the Jailkeep priority bracket in place

Both Lazy Cata and Rolestopper is Protective is needed for the "NewD3 Interaction" to resolve with a successful kill with a Town Alien instead of Town Jailkeeper, as Alien would be in the Jailkeep priority bracket with Rolestopper is Protective

I'm advocating for Greedy Cata which is making the Catastrophic Rule resolve before the Priority Ladder (and would allow for the Jailkeep priority bracket to be merged into the Blocking priority bracket)

1. What's the greedy catastrophic rule? Isn't it just the original version of the catastrophic rule?
2. I'm confused what rolestopper has to do with the classic NewD3 interaction, can you break it down for me?
3. If "greedy cata" is the same as the rule currently on the wiki, then yes, currently JK has been made same priority as roleblocker. But why do you want this?

[biancospino]

Yes, I see it as an unwritten platinum rule. That's why kills don't prevent other actions for instance.

Kills don't prevent other actions because, while a dead player cannot

submit

an action, there is nothing saying that an action fails if it is nevertheless somehow performed by a dead player.

Anyway TL raises a good question, should the classic RB/JK interaction (if it works as it traditionally does) work differently if the JK was replaced by (1) an Alien, (2) a combined Doc RB, (3) a combined RS RB? As I understand it, under lazy cata the answer is no for (2),(3) and (1) depends on whether Alien is a jailkeeping role.

TL is postulating that an Alien, being a fusion of a RB with a role he sees as protective (namely RS), should be classified as a jailkeeping role, and hence (1) should resolve as the classical interaction.

Incidentally, moving Alien would have the effect to change the interaction between a (killing) Mafia Alien and a Town Jailkeeper. If we want both RB/Alien to work as RB/JK and Alien/JK to work as it does now, we need more granularity in the priority ladder — at that point, it may be better to just devise an ordering of roles and unlink Follower tags to priority entirely.

[biancospino]

Yes, this technically makes Macho a protective role despite it being negative utility as being protected from kills is beneficial and Macho protects you from kill protection.

Btw, if a Doc targets a Macho, does it fail? More saliently, if a Jailkeeper targets a Macho, can the Macho succesfully use active abilities that Night?
I interpret Macho's ability as nullifying the effect of any kill protection it would receive, but preserving any additional effects of the action that tried to grant that protection and hence that action itself; so that the Macho is still blocked.
If Macho is protective (which I actually agree with) then protective roles modify the effect of actions rather than deleting actions.
Ascetic is not a protective in that sense: to see that, notice how an Alien does nothing to an Ascetic, but a Vaporizer would presumably both kill and roleblock/stop it: which demonstrate that Ascetic does not actually protect againat non-kill effects.