Something_Smart's Mini Normal Review

[callforjudgement]

Just dropping in to say that I've looked at this setup, but it requires a ton of thought. It's a very complex setup (probably more complex than I'd typically expect in a normal).

Quick initial thoughts: The PGO + Mafia Master Backup is both abNormal (2 roles that are greylist at best; I'm inclined to veto PGO as being too far outside the spirit of a Normal), and something of a balance disaster; if the PGO dies first, scum effectively retaliate to the many investigative roles around with a kill. In general, swing will be a huge problem in this setup, given that you're double-rewarding the Mafia when town power roles die (the town don't have the role, the scum do), and that most of the town power roles gain power as the game goes on and become disproportionately better as the scumteam gets smaller. Thus, early power role deaths will make a huge difference to the balance.

[callforjudgement]

Not really.

The big problem with this sort of setup is that two of the town investigatives are capable of catching the scum that does the kill, but not the other scum. As such, they're relatively weak when the scumteam is large (although town has

enough

investigatives that there's nonetheless a high chance of one of them getting useful information early), and much more powerful when the scumteam is smaller. No ability given to a single scum can counteract that, as there's (approximately, but pretty accurately) a 1 in 3 chance that that scum member dies first. Your ninja strongman actually probably makes the positive feedback effect worse, because the chance that that scumteam member might die early means that the scumteam may well use the shots in the early game (at least the ninja shot; strongman shots are normally saved to break up a potential follow the cop, rather than for JK immunity).

The Neapolitan is the odd one out here; they can't catch scum directly and have an incentive not to claim results (because it helps scum figure out where the town power roles are). As such, that role helps to hold the setup together, as opposed to the other roles that are trying to tear it apart. With 6 VTs (I'm assuming we're removing the PGO), it has a reasonable chance (about 50%) of confirming townies, and as such it's probably worth a little less than 2 confirmations if unlimited-shot (like the version you have there). (Confirmation count is something I use to try to balance setups; it's adjusted by subjective criteria a lot, such as who's doing the confirming or what sort of things can be confirmed, but overall something like 3 and a bit confirmations at Night is probably about the right place for a 10:3 setup.) However, a (reasonably) powerful unlimited-shot investigative like the Neapolitan is swingy in its own right, as town's chances will depend on how quickly they die.

I'm finding it hard to gauge the balance of this setup because so much of the action is in the late game, but feel that it probably has unacceptably high swing. Once that's fixed, it'll change the balance anyway, so we should probably try to get the swing under control first and then worry about the win chances for the various factions.

[callforjudgement]

JK→Doctor is a large improvement swing-wise. Doctors have no investigative ability and can help prevent town power roles dying, reducing one cause of swing.

For the Tracker role, Backup Neapolitan would probably give away too many details of the setup (although swing would indeed reduce as a result); it's not a role people think of adding to a game as a "fake backup". Motion Detector is basically useless except that it has a small chance of catching scum in a lie at massclaim; this basically means it only triggers in lylo. This would be swingy except that the role is so weak that it probably won't matter whether it dies or not. Night 3 Cop is fairly swingy in that its three possible results (dead on or before night 3, finds scum, finds a townie) are reasonably different in power level, and all of them have a reasonable chance of happening, but it probably has lower swing than Tracker just because it's one-shot.

A Mafia Tracker tends to cause town power roles to die early. Likewise, a Mafia Roleblocker gives another method via which the Mafia can nullify town roles. These make the worst case for scum less likely, but no less devastating if it does actually happen. Does this count as "good enough town play / bad enough scum play that town deserve the win"? Possibly it does, although I'm less sure than in some other setups I've reviewed, given that it mostly relies on scum failing. (The roleblocker might be necessary if we have a Doctor and a Neapolitan to prevent a Follow-the-Cop style strategy building up a critical mass of confirmed VTs. That said, this isn't an Open, and I don't think town have enough information to know that the strategy would work, meaning that they're unlikely to try it.)

How attached are you to the idea of the Mafia being able to inherit a choice of roles from town? It's an idea I haven't seen before, and we might be able to make it work. In such a case, I think a minimal start would be to give Mafia a Rolecop; this would increase the chance that they discovered a town power role fast enough to be able to have some impact on the game intentionally, rather than by chance. (Additionally, losing the Rolecop wouldn't be too much of a swing problem because it would be most useful early; lategame it would mostly only help scum by letting them get their fakeclaims in order, which tends to reduce swing.)

[callforjudgement]

Encryptor's pretty much null in terms of swing, as it places the setup between full-daytalk and no-daytalk, both of which are allowed in Normals anyway. (For what it's worth, far too many games I review include one, mostly because mods feel like they should have scum power roles.) Neighbourizer is relatively weak and thus something I consider later in setup analysis, when it's more finely balanced. (That said, it is probably negative utility for scum, especially because if scum want to fakeclaim universal backup then they'd probably have to use it rather than just save the shot.)

Commuter's a good example of a role that's useful for both town and scum (it allows scum to dodge investigatives), and it has no swing issues. Of course, scum aren't exactly going to be able to kill it unless there are limits on it.

Follow the Neapolitan isn't something I'm worried about here. It's one of the things you have to check for, but the checks look good.

If we're trying to get the setup to work as initially designed, it's worth checking which roles would be useful for scum to snag with the Backup, while avoiding having too many lategame-focused roles for town. (Note that having one such role isn't too terrible; it's when you have them in multiples that things start getting overly swingy.) Let's run through the list of explicitly-Normal-as-town roles (barring Masons and Neighbours who presumably wouldn't be stolen by a Backup, roles that confirm themselves as town because allowing them into scum hands would be nonsensical, and Universal Backup due to recursion):

Useless or negative utility for scum: Bodyguard, Cop, Doctor, Neighborizer
So weak it'd be unbalanced as a Backed Up role: Motion Detector, Voyeur
Potentially useful for scum, but not in this setup: Gunsmith,
Mildly useful for scum, nothing special: Commuter, Follower, Jailkeeper*, Neapolitan, Roleblocker*, Rolecop, Rolestopper, Tracker*, Vanilla Cop, Watcher**
Overly powerful in scum hands: Vigilante
Depends on how the role is defined: JOAT

In the "mildly useful" section I've placed one asterisk against the roles that get more useful for town in lategame (thus causing excess swing if there are too many), and two asterisks against the Watcher, who is so powerful in town hands that it distorts the rest of the setup.

The role that stands out to me most here is the Rolestopper. In town hands, it's basically a Doctor, which is good; we wanted something like that in the setup. In scum hands, on the other hand, it's actually useful, unlike the Doctor: you can use it to save a buddy from an investigation, or make your kill target immune to Doctoring and Jailkeeping. (The first method here is better in the setup, although scum wouldn't know that.) So if we're trying to use the Master Backup idea, I think we definitely want a full Rolestopper in the setup.

Neapolitan and Commuter are also fine on the town side, and putting that against a Rolecop and Master Backup on the scum side would give us something similar to the original idea of the setup, but with fewer swing issues. So at the moment, we have something like

[ilst]
[*]Town Neapolitan
[*]Town Commuter
[*]Town Rolestopper
[*]7 VT
[*]Mafia Rolecop
[*]Mafia Master Backup
[*]Mafia Goon
[/list]

Although I haven't run the balance numbers yet, I unfortunately suspect that this is scumsided. Before trying to balance it further, though, I'd like to see what you think of the general idea here, and whether it fits what you were aiming for.

[callforjudgement]

Ugh, screwed up the markup, and you can't edit in PTs. I meant this:

• Town Neapolitan
  
• Town Commuter
  
• Town Rolestopper
  
• 7 VT
  
• Mafia Rolecop
  
• Mafia Master Backup
  
• Mafia Goon

[callforjudgement]

Whoops, just realised that you can't rolestop the kill as you'd stop the kill from going through.

This is good for balance in this setup, as it means that Mafia would use the role in a way that would benefit them.

[callforjudgement]

The Commuter looks a bit out of place here. Its main goal is going to be to try to earn a free mislynch via stopping two kills (between Commuter and Rolestopper); it doesn't do anything else for the setup because there are no investigatives that depend on unusual numbers of kills (sometimes Commuter/Bulletproof can act as a false positive for these).

I'd say the town power is mostly in two places: the Neapolitan, and the chance of two no-kill nights. The worrying thing here is that if town

get

their extra mislynch, the setup is only slightly townsided (depending on how useful the Neapolitan results are, and if the Neighbourizer manages to do some convincing scumhunting in the QTs). If they don't, which is much more likely (they only get six tries to stop two kills, and scum probably won't repeat a target or aim for an obvious target once they see a failed kill), it's pretty scumsided, with town not having any roles that can save themselves by claiming. I think the town needs something extra here.

(Also, where did the other reviewers go to? I was hoping for input from someone else.)

[callforjudgement]

That seems like it could be balanced; it's stronger than Commuter (especially in this setup) and yet none of the roles produce firm investigation results because they interfere with each other, so there's less chance of a blowout. Another thing to look out for is if any of the roles could confirm themselves by claiming, but I doubt it; some of the roles are confirmable but even when confirmed, they're believable as the opposite alignment.

One thing I'm a little worried about is that the Neapolitan confirms a townie as town, then the townie gets rolestopped and scum's kill on the confirmed townie fails. That isn't really unbalanced but it would feel bad for the scum. (It does, however, give them enough information to probably play correctly from then on, in addition to confusing the roleblocker.) Another potential possibility that would cause problems would be if the Neapolitan survives until lylo and uses it to 1v1 a scum who claims VT but scans as not-VT. This is relatively unlikely (the Neapolitan would be 50% likely to survive until lylo given a 4:3 lylo if they were very good at crumbing and scum didn't have a rolecop, so the chance is much lower in practice), and also not really a problem; again, it's most likely just to be a case which leaves the scum player annoyed. I don't think either of these is a dealbreaker but I'm mentioning them in case you or one of the other reviewers have comments.

Anyway, I think this setup is within reasonable balance bounds; I normally judge 9ps, and the more players there are in a setup, the harder it is to judge balance, but also the less it matters (the larger the setup gets, the less impact any individual balance mistake has compared to random chance). The game seems about right in terms of confirmation count, and there doesn't seem to be a town or scum blowout situation (town blowout D1/N1 is to lynch scum, prevent the kill and not know why, and confirm a VT, which is actually pretty small as blowouts go; scum blowout is to kill the Neapolitan and discover the Roleblocker, and town still have a flghting chance if that happens, not to mention that the construction of the setup makes it less unlikely than the average scum blowout).

In terms of normalcy, things generally look pretty good. You haven't explained what voting does or how unvoting works; I'd recommend you throw a mention in because we sometimes get newbies in Mini Normals and they might not necessarily be able to guess. Obviously, you'll need to write up a new set of PMs and action resolutions as the setup has changed (it might not be worth bothering doing this until the other reviewers turn up, though, because they might well want changes to the setup), but the ones you had before show a good understanding of how action resolution in a Normal works. (You did miss the "you are neighborized" PM from the list of possible night results, though.) I'd recommend that the example VT PM in the rules is the same as the actual VT PM (you missed a line); it wouldn't be a balance issue in this case but it might lead to a VT trying something stupid, or possibly even an attempt to skirt the edges of rule 10. (I've seen it happen before.) I'd also recommend you explain the situation with activated Encryptor better in the other scums' role PMs; something like "you can talk during pregame and during the night; while the Encryptor action is active, you can also talk during the day". Some people are going to get confused at a role PM that says they can sometimes talk during the day, but not when; seeing an Encryptor on their team would probably clear it up, but it's best to be safe.

You also need to work out what happens if the Encryptor action gets roleblocked; this is a case that isn't covered by the normal action resolution because Encryptor is normally passive, but becomes active if made 1-shot. As far as I can tell, the roleblock works, and you lock the mafia PT so that they don't talk in there because they think they're allowed to; to avoid confusion, you should probably post something there like "The Encryptor action failed, so this thread is locked", because the failure is observable and would confuse the scum otherwise. (The Encryptor action is untargeted, so it can't be rolestopped.)

Prods of 48-with-substantial-content+24 are going to be hard for some players to keep up with; I fear that if you don't allow for prod-dodging, you're going to have to find a lot of replacements. (Of course, I might be completely wrong about this.) This is completely your call; I just thought you should be aware of the potential problem.

You might want to consider whether to place a statement about faction sizes in your game advertisement (if you do, it should also be in the rules). "This is a 10:3 game" is reassuring to people who don't like multiball. This is entirely optional – sometimes it's fun to have to figure out the faction sizes too – but most mods don't realise it's an option, so I'm bringing it to your attention.

Finally, I tend to get a little annoyed at seeing a "Modkilled" section in the first post, especially when it's nearly always empty; it sort-of gives off the impression of "this mod is triggerhappy". (I just have an alive and dead, and would put modkills in the dead section if I ever had to make one.) This is just a personal preference/peeve of mine and has nothing to do with the Normal guidelines, I just thought I'd point it out. (You are perfectly free to ignore me on this.)

[callforjudgement]

Oh, something else you didn't mention anywhere is what you do with confirmations and when you end pregame.

The "standard practice" for a 13p Mini would be to leave the main game thread locked until eleven players had confirmed. (The idea is to guarantee that at least one scum had confirmed before allowing the game to start.) This doesn't need to be mentioned in your rules, but would typically be mentioned in a post you made after sending out the role PMs. (Something like "All role PMs have now been sent out. The game will start when 11 of 13 players have confirmed.)

[callforjudgement]

@Cheery Dog: It's definitely pretty low on the swing. What do you think of the balance win-percentage-wise? To be balanced, a game both has to have non-excessive swing, and a win percentage around 50%.

[callforjudgement]

"Slightly better than a mountainous" is unfortunately too low. The town win rate in practice for 2:11 Mountainous stands at an enormous 0%.

That said, town have a

lot

of advantages here compared to a typical Mountainous.

Aeronaut, care to weigh in? I haven't seen you post here at all.

[callforjudgement]

1-shot Encryptor is something that isn't covered in the Normal rules (not in the sense of "this isn't Normal", but in the sense of "this is apparently Normal but there's no guidance to suggest what it actually does"). I guess that if we're taking an analogy with Bulletproof, it'd allow the scum to collectively send one message during the day (and be activated automatically when that happens), but that's clearly ridiculous, so one day of messages makes a lot more sense. Does that make it blockable? The only other role I can think of along these lines is activated PGO and I don't know whether that's blockable either.

I think this probably comes down to the category of "mod decides, but needs to have a plan confirmed in advance", like with roleblocker vs. jailkeeper. So making the role unblockable is entirely reasonable. Normally I'd write the role as "Strong Encryptor" to make the unblockability clear from the flip, but that's a bad idea in this case as it might be misinterpreted as being able to make an unblockable

kill

. I guess we're just going to have to live with the ambiguity.

[callforjudgement]

You need some rule along the lines of "If more than half the players are voting for another player, that player will be 'lynched' and removed from the game. If deadline is reached before a player is lynched, the day will end with no lynch.". There are plenty of ways to word it, but a) without that rule, the game doesn't make sense, b) people won't know whether you're using majority or plurality voting. (Basically everyone at mafiascum uses majority voting and you should, too, but there's enough doubt that it's worth making it clear.)

You need to make it clear (either in role PMs or in the rules) that players can't self-target. (Partly because that's what Normal rules say, partly because if you have a Rolestopper and Roleblocker, self-targeting can lead to situations that the normal rules don't resolve.)

Your Rolecop results are wrong. Here are the correct results:

"Vanilla" on a VT or Goon; "1-Shot Encryptor", "Rolestopper", "Roleblocker", "Neighborizer", "Neapolitan" on the power roles.

The point is that rolecopping someone doesn't give away their alignment. Of course, the scum already know what it is in this case, but you shouldn't make the role do something different just because it doesn't matter to the setup. (Note that it's not 100% useless for scum to rolecop their buddies; they might do it to give a false positive to a hypothetical Tracker. There isn't a Tracker in the setup, but

they

don't know that…)

There are a couple of other interactions I had to check to avoid that the (e.g. is it possible for the roleblocker to be considered to have lied about their role because they tried to block the flipped non-Strong Encryptor and didn't know they were blockproof?), but in each case, there turned out to be no problem (in this case, because only scum can observe the fact that daychat is open). So with the changes I suggested above, everything looks fine normalcy-wise.

In terms of balance, I'm personally fine with it, but I'm not confident enough (nor allowed by NRG rules) to sign off on the balance simply by myself. Given the situation, I'm going to ask the listmod for help.

[callforjudgement]

"50%+1" is misleading because with 9 players, the lynch threshold is 5, not 5½. "More than half" is a nice simple way to describe the actual rule that people use. ("A majority" would also be mathematically correct, but too many people don't know what that means.)

Mafia do know it's not multiball in this case (10:3 is in the setup advertisement), although I agree that we should write the roles as if they didn't.

[callforjudgement]

You forgot to fix the rolecop results. Other than that, the OP's looking good.

The town power here is hard to judge the balance of because it's a large set of relatively low-powered roles, which mean that there are a lot of possible scenarios. This isn't a bad thing for the setup at all; it just makes review more difficult.

Strangely enough, I think having two blocking roles actually weakens town a bit. The existence of the Rolestopper means that the Roleblocker is less useful as an investigative (you can't conclude much from the lack of a kill). The town roles aren't as powerful as they look, as they'll all tend to get in each others' way; town won't get much actual benefit from their blocking roles because they need to stop two kills to gain a mislynch (unlikely) and blocking non-kill actions will more likely hurt town than scum. I do have something of a concern based around scum not knowing what to claim in a setup like this, though (VT is the obvious claim but the existence of the Neapolitan makes it risky; most power role claims will be hard to believe or lead to a 1v1). That said, claims tend to come out at lylo and by that point, either it'll be the first investigative information of the game, or else scum will have a good idea of what's coming. (Possibly both!)

Aeronaut, if you're worried about breaking strategies, what do you think about limiting the shots on the Neapolitan (to maybe 2 or 3)? That way, if town get a Follow-the-Neapolitan set up, they'll pretty quickly just run out of investigations, and it doesn't interfere much with the balance of the setup otherwise.

[callforjudgement]

I personally think it's probably around OK as it is, but I was waiting to hear back from Aeronaut. Again. :-( If someone raises a balance objection, we should listen to it. I'll send another PM.

The situation with backup mods in Normals used to be that the listmod would back up all the games personally. It's a while since I've seen that happen, though, so it might have changed. A backup mod wouldn't need experience, though (because a mini normal doesn't need experience); the only restriction would be that they couldn't be backing up another game at the same time.

[callforjudgement]

OK, /approve, either with a 3-shot or an unlimited Neapolitan. (Something_Smart, you need approval from all three NRG members to be able to run the setup.)

[callforjudgement]

Cheery Dog?

[callforjudgement]

Right. I've let the listmod know, so the game's back in listmod hands for now. (You'll be added to the Normal queue at some point in the near future.)