Why not massclaim?

[Farkran]

Disclaimer: sorry if this topic has already been explored, i skimmed the titles and ran a search for "massclaim" but couldn't find anything. I am surprised this has never come up though - i feel this is like the oldest topic in the world and i am fairly sure this must have been a hot discussion some time in the past, so if there is already a thread about it please close this one and redirect me there, thanks.

On the offchance that this is new, or if we want to discuss the subject in the recent meta:

WHY SHOULDN'T PLAYERS ALWAYS MASSCLAIM IMMEDIATELY?

Specifically in games where a massclaim is possible (i.e. not mountainous-like) and not directly contrasted by game mechanics (i.e. avalon mafia). What is there to lose by massclaiming immediately, and what is there to gain? A rough analysis of pros and cons follows:

PROs

• Force scum into lockclaiming into something, with the least possible amount of information and the least amount of time to think what of could work best. They will have to come up with something that could work, without any pre-existent softclaim, hardclaim or flip. The chances of a counterclaim, or dumb/unbelievable individual claims rise greatly.
  
• Lynch accuracy. Depending on the PR/VT ratio, you can determine if you want to lynch in the PR pool or the VT pool and on which days. If you have a provable claim (FN, vig...) you may want to give them a chance to confirm themselves or die before they are lynched. If there are too many VTs, you may want to avoid lynching PRs entirely. In either cases, you narrow the lynchpool significantly.
  
• Be done with setup spec as soon as possible, so that you can focus on dayplay. It will be tiresome for a few pages, but then you're done forever.

CONs

• You teach scum where to kill. If there is a doctor claim, you can wave him goodbye.
  
• You teach scum what to expect. If there are roleblockers, soft investigatives, vigs, etc - scum can work around those.
  
• Massclaiming is boring.

Is information about who and what the PRs are, so important to keep hidden, compared to the benefits provided by outing them? In which circumstances? If we come to a conclusion that massclaiming is almost always beneficial, as i now tend to think it would be, how do we counter massclaim-based solve strategies?

My recent experience with Avalon Mafia taught me the importance of keeping your identity hidden at all costs and i had fun, so i wanted to ask for some other people opinions on the matter.

[Farkran]

This is the way I see it:

The purpose of PR's, by and large, is to reduce the size of the lynchpool and/or increase the number of lynches town gets. Cop? Anyone you get an inno on is removed from the lynchpool, and if you get a guilty the lynchpool temporarily shrinks to 1. Doctor? Two saves and town gains an extra lynch, plus you can save other PR's. Vig? Two shots give an extra lynch, and you are out of the lynchpool just by virtue of existing. Pretty much all PR's come down to influencing one of those two factors in town's favor. Scum PR's, as well as the scum NK, serve to influence those factors in scum's favor-- be it by roleblocking a cop, or shooting a cop's inno, or protecting a partner from a vig. The difference is that town PR's generally benefit town just by surviving (cop and vig are pro-town pretty much no matter what, barring really bad vig shots, and doc is useful as long as there's ANYONE that you obviously want to keep alive), whereas scum actions need to be aimed precisely. A scum RB is no good unless you can hit a town PR with it. Nightkilling VT's might help your dayplay but it won't stop the cop from clearing half the town and autowinning.

By massclaiming, you allow both town and scum to use their roles in the optimal way. The difference is, town were probably able to use their roles in a MOSTLY optimal way, whereas scum were very likely not using their roles optimally at all. So a massclaim provides a small marginal utility to town, and a much larger one to scum.

Obviously, this changes over the course of the game, which is why you see everyone massclaiming in or near LYLO. This is because as the game goes on the scum get more information and are probably a lot closer to using their roles optimally on, say, D4, compared to D1. So the massclaim helps scum comparatively less, and it still helps town because setup spec can provide a small boost to dayplay (or occasionally a large one, but balance is usually nebulous enough that this benefit isn't big).

These are good points in a vacuum, but wouldn't a massclaim narrow the scum pool in a similar way? The reasoning also assumes that the PRs are useful - a lot of the time, at least in Normal games, they aren't. Stuff like Backup X, Gated-night X, soft-investigatives, doctors, roleblockers... is their night action really better than what you gain by outing them? Moreover, i did say that some roles would be able to prove themselves even after having claimed unless there are roleblocks in action (which is not always the case).

Now there are roles that can be very useful if kept hidden (masons, mostly), but i'd say most other roles aren't better hidden than they are outed. Also, if all roles are outed immediately, most of them will still be able to act, depending on the setup. I mean, if there are 2 investigatives, one of them will likely act and will be able to produce conftown (or softinno). If there is a doctor and a investigative, the investigative will likely act.

Regarding lying town, that's just bad play in 9 out of 10 cases, but i know it happens so i don't have any counter argument to that. It's just bad, and you should never do that in a massclaim. When there is no massclaim running, that's a different matter.

Regarding setups that punish massclaims, yeah, the premise is that i would like to see more of those because i think they're fun!

[Farkran]

No, i don't know what it is, sorry

[Farkran]

I don't mean to say those roles are useless. They do have utility, i was just wondering if said utility would be better - on average - than having a d1 massclaim. Of course there will be circumstances where not massclaiming would have been better, but if you measure that against all the times it wouldn't, maybe massclaiming would be the better choice in general. I am also confident that setup designers would start to work against that, which was one of the points of my original post.

For instance, i am a strong supporter of the Janitor mechanics, at least partially (i.e. flip alignment but not role, maybe?), and i had huge amounts of fun in avalon mafia where scum could win by guessing who a specific role would be even if they all died (like it happened). On all the normal games i played (2), i feel like massclaiming would have been better than not doing it, and i have seen it working in town's favor in TM2020 too.

Like... in my experience, scum is usually not ready to massclaim early, but townies are so jealous of their secrets that they think anyone fishing for roles in the main thread is automatically scum, even when their role is utterly crap. I have literally seen that happening to me at least 3 times. Did anyone really see a scum player fishing for roles in the main thread instead of discussing opinions in their PT?

[Farkran]

4 to 5 PRs in mini normals also matches my experience.

I can agree that masons have a free pass to hold their claim or fakeclaim VT, i disagree about a vig. I have seen too many times vigs being unable to confirm themselves due to double killing the same target with scum, being roleblocked, target being healed, etc.

[Farkran]

If you mass claim day 1 in a 13 player game. Given that there are 4-5 town PRs like you guys said, it's much safer for mafia to claim VT. Generally 3 mafia in a 13 player game so you have 3 mislynches before lylo.
Given that all 3 mafia claim VT, and you believe all PR claims are truthful. That's still finding 3 mafia in 8-9 players.
You'll start getting paranoid about one of the PRs being mafia at some point in the late game surely. So does mass claiming really improve the odds of town winning? I don't think so.
Mafia always have tools to deal with town PRs, ie, if town have a watcher, mafia will at least have 1 shot ninja and so on. So while mass claiming might(huge if) make the lynchpool smaller initially, it completely shuts down all the town PRs at night.

I wouldn't say always, but it's true that often there are watcher-ninja couples and the likes. However, i mean, that's pretty much a point that reinforces the theory rather than contrasting it? 1. if this is true, mafia having a ninja will guarantee that they know there is a tracker-like role; 2. if the PR results aren't reliable, why wouldn't you do better off by claiming? Let's keep in mind that this whole argument is based on the CURRENT meta, not what setups could or should be.

Also i haven't seen counters to:

1. Counterclaims are far more likely to happen, especially if scum is forced to claim first (this is luck-based if you follow playerlist order) - this may force them into claiming VT, which means lynching in the counterclaims or VT pool will be ideal.
2. If all PRs claim at once, at least most of them will be able to act and get results anyways - investigatives targeting in the VT pool is also ideal based on point 1.

Then again, you aren't necessarily required to forgo all the dayplay aspect. You just reduce the lynchpool, then you still have to pick correct, and that generates analyzeable interactions.

[Farkran]

I have no experience with Large Normals, is there an archive like there is one for Mini Normals?

It would be interesting to know the variance in roles for claimability purposes.

Like, in 13p games it's pretty much a given that there are more than 3 and less than 6 (non-inclusive) town PRs = 30% to 38% of the players

VS

3/13 scum = 23% and 4 lynches available for the town before losing (lylo after 3).

What's the proportionality in larges, assuming only one scumteam? If the variance is proportionally higher, massclaiming would be much worse than it is in Minis. The possibility of third parties or multiple scumteams also make massclaims less reliable.

In opens/semi-opens and micro, i think the argument is still valid where applicable.

In the current newbie matrix, it's probably not efficient because the PR variance is too high at 11% to 22%

VS

22% scum, 4 lynches available