Mini Normal 1954: Iambic Tetramafia: Endgame

[mhsmith0]

mhsmith0, mastina

And... The elegance and symmetry not to mention the fact the setup was so easy for me to guess suddenly makes sense.

No offense jj, just that this is exactly the kind of weird setup that makes perfect sense in hindsight I expect from them.

thanks

[mhsmith0]

Neighborizer and PT enabler are pretty fundamentally lynchable roles.

Neighborizer in particular isn't even slightly clearing, and a paranoid town can/should be VERY concerned about the possibility of a scum neighborizer rolling them (which basically happens every time you get a scum neighborizer lolol)

[mhsmith0]

But yeah, scum need to not shoot their traitor. That's inherent to a game design with a traitor. Mastina can talk more about that

[mhsmith0]

Town basically had:

1) 2 masons (ok that's helpful, though lack of day chat nerfs their power substantially since they can't really coordinate)
2) A largely insane cop (but who knows he's insane)
3) A generically non-clearing role (neighborizer) who in practice will super nerf the insane cop since he'll be EVEN MORE likely to get a guilty that isn't really one, AND you get the realistic chance that scum will be guiltied while in a hood and therefore have a great excuse (which is why, for instance, you should never give scum a vanilla cop or any other role that has a gun when town gets a gunsmith unless you're intentionally trying to nerf the gunsmith)
4) An explicitly NU role that, when dead, fucks with the power set
5) 5 VTs

But yeah a lot of what was going on here was the potential for the PT cop to falsely guilty masons or neighborhood people and/or to falsely clear the traitor. That potential results dysfunction was a pretty non-trivial portion of scum's power in this game.

[mhsmith0]

History of traitor games in 10/3. Note that I'm not sure when traitor became only normal if non-recruitable; 1543 a solo traitor won, so clearly it was after that...

Spoiler:

1401 town
joat (gunsmith/vig/rb), odd night vig, 2x masons
traitor, goon, 1-shot bp

1402 scum
bg/tracker/joat
vs traitor role cop/goon/encryptor
(here traitor could be reruited, which is non-normal, idk if that was true then tho)

1483 town
tracker, 1-shot vig, gunsmith
vs traitor, goon, 1-shot ninja

1486 town
1-shot cop, ic, even jailkeeper, cultwach admin (idk what that was)
vs traitor, goon, rolecop

1543 scum
even night compulsive vig, doc, 3x neighbors
vs traitor, joat (rolecop, roleblock), nurse

1585 scum
tracker, jialkeeper, bodyguard
vs traitor, goon, strongman

1642 scum
gunsmith, bodyguard, ic, 2-shot vig
vs goon, encryptor, 2-shot roleblocker traitor

1740 town
?????
vs traitor, goon, jailkeeper

1762 town
3-shot jailkeeper, gunsmith, 1-shot vig
vs goon, goon, traitor

1791 town (was an open game, not really comparable)

1814 scum
non-consecutive jailkeeper, joat (voyeur, follower, neapolitan, bodyguard, motion detector, neighborizer), weak doctor, 2-shot commuter
vs traitor encryptor, goon, goon

1895 town
cop, fruit vendor, doctor (and 3-man hood including cop as a member)
vs traitor, odd night rb, even night rb

1938 scum
JOAT (doc/jk/bg), 2-shot vig, gunsmith
vs bp traitor, joat (cop/rolecop), neighborizer

I'd say that town power doesn't particularly stand out as greater here than it was there. Scum won 1814 with less power (and I'd say town had more including three distinct ways to prevent kills, though jk/weak doc interactions could be potentially nasty for town) as an easy example.

PS 1543 was an absolutely horrible design and if anything I'm impressed town made it to 3/1 mylo given how little they had to deal with.

[mhsmith0]

Here you have a number of PRs that vaguely fit together, but no real proof that any of them are town-aligned other than masons and (barring a fairly lol town) the PT cop. Neighborizer and PT enabler are explicitly non guaranteed town roles. They can't be "vouched for" by anyone else, and even if they CAN prove their existence there's no proof at all that they're necessarily town roles.

Or to put it a different way: in mini normals, it's pretty common for towns to lynch PRs WAY more clearing or important than neighborizer or PT enabler. The realistic lynch pool was 7/10 town roles, the potential for the PT cop to out himself without ever getting anything useful was substantial (and the potential for him to also out the masons was pretty substantial too), and scum had just a little bit extra power to potentially screw with town night actiosn if needed (for instance, if they wanted to shoot a mason n2 and save the pt cop for n3 or something)... for a mini normal that's pretty balanced.

[mhsmith0]

The neighbourizer while it is lynchable, I generally see it as town in most games so far. Yet to see a scum one.

viewtopic.php?f=53&t=29549

Skimming it you see something like a 3:1 ratio of town to scum neighborizers, and the ratio is smaller if you exclude stuff like JOAT where one of the shots is neighborizer (or a boosted role like weak neighborizer), ditto if you count only from, say, 1700 to now (1400-1700 had almost exclusively town neighborizers, more recent games have had much closer splits of the role)

And scum neighbors and neighborizers just crush these games, so town SHOULD be very wary of a neighborizer most of the time. If they setup spec'd their way to that role bieng necessarily town, they were both incorrect and wandering into very dangerous territory for whenever they'd next have a game vs a scum neighborizer.

[mhsmith0]

And scum neighbors and neighborizers just crush these games, so town SHOULD be very wary of a neighborizer most of the time.

I don't think I've played a single game with neighbors where at least one of them wasn't scum.

It happens with reasonable frequency. Especially when you have more like a 2-man hood.

And of course if you have a town neighborizer, then it depends who they target

[mhsmith0]

FWIW, I'd say that if the game WAS unbalanced, it's perfectly fine to say that the game is unbalanced. People fuck up on setup design (sometimes pretty egregiously too). It happens.

The issue I have is that THIS game design wasn't particularly unbalanced.

Traitor (and traitors in normal games are EXCLUSIVELY non-recruitable) is generally worth something like half a goon (mainly because it can't be recruited). The even night roleblocker is a little bit extra for the scum team, ditto encryptor, but you're still looking at a generally low amount of scum power. So essentially, you look at town power on 3 goon setups...

1921 - cop and jailkeeper - highly concentrated power, but hard to argue that it's less power than town got here
1908 - miller double voter, odd night cop, 3-shot gunsmith - double investigatives, the miller double voter is somewehat self-clearing (and at any rate will divert cop/gs investigations elsewhere by nature), more town power than this game
1905 - weak jailkeeper, ascetic JOAT (cop, doc, vig), loud role cop - more town power than here
1890 - 3x masons
1875 - cop, jailkeeper, 1-shot bp - more town power than 1921 due to the bp
1843 - 3x masons
1833 (2 goons, 1 encryptor) - hated, 2x masons, watcher - more town power than here, watcher and masons in particular have strong coordination potential

(for kicks, you can also potentially look at 1929, an 8/3 game where scum power was very comparable to this, but town got a watcher, 2-shot vig, and 3-shot doctor, though again that was 8/3 not 10/3 so the lack of an extra day doubtless played too)

That compares to a 2-man masonry (WITHOUT daytalk, which nerfs utility quite a bit), a cop who knows he's questionable sanity but is actually REALLY questionable on sanity, a net neutral utility neighborizer, and an explicitly NU role that shuts down a good part of town's communication powers when it dies (though in fairness, it also enables the PT cop to be stronger, though frankly it's pretty unusual for the PT cop to not be outed by the time the NU role dies barring like a d1 lynch or something)

I think ultimately the interactions between the town roles was really poor, the potential for the PT cop to completely blow things up on town was substantial, and town had zero active defense against scum's nightkill (no watcher, doctor, bp, jailkeeper, etc etc etc)

I'd also note that there's a reasonable expectation that somewhere along the line the traitor is going to blatantly strongarm a town mislynch, possibly even via some kind of fakeclaim (like a cop guilty or the like), so barring a really early shot on the traitor, scum team really shouldn't be too disadvantaged by the possibility of a late game traitor shot. When the traitor is the role that makes it to endgame, something went drastically wrong for scum.

[mhsmith0]

This would be pretty brutal as an open. Five claimable and confirmable town PRs is a gimme win for town. Well, I suppose some towns might lose it anyway, but it'd take a lot of town dysfunction to blow 10/3 with 5 effective ICs.

[mhsmith0]

Town is massively under buffed in comparison to the games you cited

And the neighborizor in this game acted like some sort of Miller cult and trying to argue it had any town utility at all is extremely questionable

Yep. Neighborizer in general is not much help to town, other than rare setups with a good stack of investigative power where somehow you might coordinate (and said setups are kind of rare)

"Games which are otherwise Normal may still be more suited to a different Queue. For example, a game with a large number of Doctors may be better run as an Open Game (after review in the Open Discussion Thread) or as a "Hospital" Themed Game."

Tbh im surprised this setup was approved normal by the above clause (at no detriment to the fact that it was a pretty good setup)

Nothing about this setup was particularly odd in terms of normality. No multi-doctors (and 2 docs probably ok if you can balance it, but like 4 is blatantly non-normal), nothing else that was particularly bizarre. You had role interactions that were a bit challenging, but that also is pretty normal.

Like, a 10/3 with 3-5 town PRs and something not crazy weird in terms of scum PRs, and with just one graylisted role, is just on the surface normal, so you'd generally need quite a bit (like, REALLY weird roles / role interactions and the like) to argue it wasn't.

PS
5 town PRs of (on average) poor strength against a goon, a very limited roleblocker, and a traitor that is effectively a godfather to the investigative. That's generically around balanced.

I'd also note that in 10/3 games, you have town records of:
0-6 with 3-4 VTs (i.e. 6+ PRs, though a decent % of those the "PRs" are neighbors which really aren't PRs)
1-5 with 8 VTs (i.e. just 2 PRs, i.e. just 2 PRs is unbalanced AND pretty non-normal)

in a more standard range (5-7 VTs, 3-5 PRs)
7-15 with 5 VTs (and 5 PRs) 32%
22-36 with 6 VTs (and 4 PRs) 38%
31-36 wiht 7 VTs (and 3 PRs) 46%

in general, one of the more notable tendencies in normal games is for game designers and/or reviewers to give town a big stack of garbage power roles (and/or really stack the scum teams with power) and just presume that a higher # of PRs is itself substantially protown, when in fact the data pretty strongly shows the most balanced setups we run in 10/3 mini normals are 7 VTs, 3 town PRs.

[mhsmith0]

Yea I don't know why traitor should auto lose. In stack the deck traitor traditionally picks up night kill power if all goons are eliminated and that's also a 3:10 setup, so what's up with that? One could argue that the PRs given to town this game rival the max spec for stack the deck.

This setup was town sided in that sense, but it also shouldn't be used as an excuse to hide the fact that scum also played rather poorly to get lynched back to back the first 2 day phases.

stack the deck is essentially unique in how traitor is treated. In general traitor is considered half a goon with the reason being that traitors can die to scum team and get endgamed if the other scum die first (some mids let traitor be recruited if scum team shoot but that’s nit a “normal” role)
Hilariously in stack the deck most scum teams recruit the traitor, which is objectively anti win con but there are a lot of dumb scum out there

[mhsmith0]

It is both: The setup was townsided, and scum wasn't really good(and I have to apologize here for my play, it was abysmal)
Yes, the setup was townsided! The interactions between these roles were far better than the first look seems to show, excluding the Neighbourizer.
The breaking role here was the Town PT enabler. Yes that one. Remember? Jingle claimed very early that there must be masons. That was the problem with the PT enabler. It was, with PT cop alive, anti town. The claim broke the game. The early claim meant two things:
Mafia knew that there were masons(probably).
The PT cop knew that there were Masons, making them reluctant to claim.
The town got a basically confirmed town that the Mafia would NOT want to kill(at least not until the PT cop died).
A PT innocent was good for town(yes there was the traitor, but the traitor did not need to die for mafia to lose!) as they could have been no mafia killing team member.
Traitor surviving is a very very weak false innocent.
The only, yes ONLY roles the weakened roleblocker could block were the Neighbourizer and the PT cop. The latter was the only good target, the Neighbourizer would have been a bad target.
The Encryptor is a fairly weak power role for Mafia in this case, especially because it

Yes the Neighbourizer was lynchable.
The enabler, if they were claiming D1(which did happen!)?
No way.
The masons? Sorry... doable, but uhhh...extremely hard to pull off. If you are caught lying its game over.
The PT cop?? That one was so outlandish yet did fit so good... very hard to lynch. Then you get these hard clears from PT cop(yeah one is fake but the fake one could not be used for anything else but one BS claim then ppl would think"traitor".
This game was not obviously townsided.
Now, I did not say that game was badly designed. I, however, will say that there was a problem that was hard to spot; and it is more in line of problems that probably just come up during play. It happens, and I am not angry about the reviewers missing it. But this game should be an example further reviews should take into account when scanning for less obvious problems.

In this game you had three largely unlynchable roles: pt cop and masons. The utility of the PT cop was overall low and always had substantial odds of outing himself and/or the masons very early (and it’s far from unrealistic that pt cop or masons the the other mislynches along the way due to the “guilty”. Such an outcome is super proscum both because it outs town PRs well before endgame AND because it protects the traitor from getting shot from like n2 through n4 barring a REALLY stupid scum team.

As for balance... a reasonably balanced game against a three goon team (sans day talk) would be something like cop, bodyguard, one shot bp. There the cop gets an n1 shot off, has a bodyguard on him in case he gets outed early, and the bp is claimable and somewhat self clearing as the third PR.

There you get two largely unlynchable slots (bp can be lynched but it’s still a reasonably credible claim most of the time), a third (bg) is self resolving and protects the cop, and of course cop gets explicit checks (2+ most of the time).

Such a setup is objectively more town power than this one, but it’s still balanced, mostly because most MS towns are garbage. Without a LOT of help, ms mini normal towns lose. A LOT. You can certainly decry that aspect of site meta, but it’s a thing that absolutely exists, so “hey town got some potentially useful stuff” is not an argument for it being townsided, it’s an argument for it being balanced.

Fwiw I think you guys are also underestimating the potential nastiness of a town clear on the traitor as well; a traitor unleashed by effectively being a cleared godfather can strongarm town lynches reasonably effectively, and the scum team can realize the traitor was NOT actually truly cleared, and can potentially recognize traitor play via what he’s doing with townclear status.

[mhsmith0]

The danger to town of a cleared traitor is in large part the fact that traitor, while a normal role, exists in a very distinct minority of mini normal games. As of the last time I updated my spreadsheet, there were about 160 completed 10/3 mini normal games, and only NINE of them had a traitor. It's a rare role in the mini normal queue, so "oh maybe that seemingly cleared townie is a traitor" is a thought process that MIGHT exist for town, but it's a situation where the information edge to scum team (i.e. they KNOW there's a traitor and the PT clear is meaningless wrt whether someone is a traitor) is potentially important. Of course the scum team could simply blow their advantage by shooting the traitor anyway, but that possibility is why a traitor is worth less than a goon, and why the game was balanced to give town somewhat less power than they'd get against a 3-goon team.