https://x.com/i/status/2082731167334158755
What in the everloving fuck is deadlock about
1 Like
I would never consider disabling proxy submissions because it is both detrimental to logistics and wolf experience. The actions would still be micromanaged but now someone is getting pinged 18 times in wolf chat and probably gets low-key flamed if they donât show up in time and have an important action. Just sort of sucks.
Also the advantage the informed minority has is information and time zones exist. It is not practical for all players to wait for all the information to come in from the day and then submit their actions. However, players will frequently make decisions that maximize their chance of winning at the expense of fun and their own health, so some players will stay up way later than they should to submit and sometimes the micromanaging leader is inconsiderate keeping them up because they havenât found the best target yet, etc. etc.
But yeah as Lissa mentioned this is a thing you design around. This is part of the reason why Iâm a big believer in placing constraints on wolf actions. Actually pressing the buttons is much cheaper so you make it harder for them to use it on game theory optimal targets either because itâs something like a slank vigil that has some restriction out of their control or it requires them to do something in the game thread that requires audacity or cleverness to get the target they want. The latter is something you canât do too much, but wolves holster those sort of actions on accident a lot. Iâm also a huge advocate for constraints on strategists and to some extent leaders as well because yeah theyâre really powerful!
(This is less germane to the topic at hand, but you also have to consider the cognitive load youâre putting on wolves in design. The average wolf is usually more mentally drained than the average villager, so you donât really want to make the night action puzzle overly complex in certain areas. Like you could give a villager whose role is a literal puzzle, but wolf roles are like pieces of the much larger puzzle that is all of their roles combined and you want to limit yourself to making it less obvious where to place that piece.)
Also I keep looking for ways to smooth out holstering variance for villagers like a village strategist was a design target for Anni 2025 for a long time. Didnât end up making it into the game itself for a number of reasons, but I was trying to make one.
(There are a number of reasons why something like a town strategist is really, really hard to implement. Itâs tricky to avoid it leading to âenfranchised veterans get to have all the funâ situation where the sick roles go to big names who then all die in 2 cycles and itâs potentially worse for the village because all of their good roles are dead. Itâs also hard to figure out how to give this power to an uninformed player. I sometimes see a thing where players can sign up for âhigh intensityâ and affects the sorts of roles they can be randed into. This can sometimes act like a strategist in that youâre more likely to have a complex role on a player who can handle it, but some of the most important roles are low intensity like vanilla cops and vigis. You could perhaps change this to like âhigh responsibilityâ, but ehhhhhhh.)
4 Likes
Anyways Iâm feeding my radio ice pop things and iberian ham to turn a profit rn, ask me anything
2 Likes
Maybe in a mash
But in smaller count games my experience is that people treat âstrongman but only against certain types of protectionâ as ânormal kill but slightly better if you can predict the typesâ (i.e. they donât care most of the time) and âguranteed killâ as an actual strongman (when included ofc)
Citation: Heroine Virtuous
2 Likes
Not MLA-formatted, smh.
1 Like
Zugbot has basically this
Nobodyâs ever used its system in a game though
3 Likes
Yeah I donât think that level of granularity is actually that useful. It would work better if game pieces were face up, but the game wouldnât make sense if they were. Like the -1 roleblocker canât know if theyâre targeting someone with +1 or +2 priority/willpower/focus whatever. Thereâs usually no feedback so itâd just be a blackbox where you use the action and your target is either blocked or theyâre not blocked. This is basically how it works now where you have like Strong-willed that prevents you from being blocked, and itâs a fine tool in the toolbox that usually isnât used in high volume even in mashes, but is a useful one to have access to. But then you think about the next level like Level 2 Roleblocker and Level 2 Strong-will, and how often youâd really want to use those, and you could ask the same question about Level 3 and Level 4 etc. The system is more flexible, but is less conventional and adds complexity for what seems to be little useful design space. Iâm not saying that you canât make something interesting with itâIâm sure you could come up with some good setups in factâbut adding the machinery to support that into the system has a non-trivial comprehension cost. Most of the time players can ignore it, but if itâs like something thatâs possible that may or may not be in play, it just sort of complicates mechanical inferences.
4 Likes
I think if you pretend toki pona style the only numbers are 0, 1, and 255, and maybe 2, itâs a decent way to allow implementation the typical suite of roleblocker strongwiller commuter unroleblockable in a way thatâs intuitive to both humans and computers. The goal of Zugbot is to be able to support arbitrary abilities, so modbot style doesnât work as well.
I dunno how working in the greater design space it affords with weird numbers like 2 and 3 and 5 would be, Iâd have to try it.
7 Likes
I just brought it up because it was basically exactly the thing tutuu was describing
3 Likes
Modbot provides a very specific framework for vanilla-y games with particular rules; the idea of Zugbot is to provide a framework for hosts to do basically any kind of game whatsoever, barring really weird bastard stuff.
This system lets you implement differences like âdoctors protect against only one kill vs. all killsâ, âstrongman works against doctor and BPV but not commuterâ, etc. without hardcoding them in to the system and making a ton of work with player <-> host <-> developer communication.
I thought it was weird at first but after thinking about it I like it in that context. No it would not be strongly valuable in a manual game, but as specifically a framework for a bot intended to support arbitrary roles I think itâs flexible and allows for the stuff hosts want.
5 Likes
That being said I still havenât put in the mental effort to understand how focus and redirection would work lmao
4 Likes
Iâm being perceived (guy at the checkout counter is commenting about how many cherries Iâm buying).
4 Likes
Itâs not even that much; itâs, like, three bags.
3 Likes
My girlfriend campaigned for cherries even though I donât like them much and them ate 0
4 Likes
In her defence I only ate 2/3 of the apples I bought; in my defence I went on vacation and she likes apples too
3 Likes
welcome back town of salem
4 Likes
I was mostly responding to the concept tutuu pitched. Zugbot working that way is fine since, as you say, it can easily simulate conventional dynamics, and disguise the abstractions that run it under the surface. I think it would be better if the ability to use numbers like 2, 3, 5, etc. were bounded to games that explicitly and publicly advertised that possibility because players need to make mechanical inferences from time to time, and the boundaries have a lot value for that purpose. I also wouldnât bake that sort of flexibility into that subsystem myself for design philosophy reasons that Iâve decided I donât want to spend the time working out, but this is certainly a âreasonable minds can disagreeâ thing.
5 Likes
In order to properly replicate town of salem you must then utterly refuse to let wolves actually have a strongman
3 Likes
that is perhaps the worst yamoto player ive ever seen
4 Likes
You havenât seen anything yet
3 Likes