Matrix6 BP fix

[mhsmith0]

The easiest and simplest way to fix the BP issue that's popped up in matrix6 games is to just replace it with a different role. What are people's thoughts on replacing it with a backup (becomes tracker or jailkeeper if or when the first dies)?

That gets rid of the issue of incentivizing early bp claims (also makes it a risky scum fake claim since it's a role that can take actions and that can forcibly clear people), which I think is really the only current issue w the newbie setup.

[mhsmith0]

Newbie games should have sample role pms near the start, so it's easy enough to see what it means, plus presumably at least some ic/se players would be clued in enough to further explain.

[mhsmith0]

It's detrimental to development and interest, as claimed bps can get bored and lose interest. Fwiw I also think that it's a behavioral detriment to scumhunting, as people waste time with setup discussion and claim trains instead of working to determine alignment, and that this materially hurts town win rates over the long run.

Bottom line: it's negative for player development, and the mechanical benefit is much lower than people think.

Ps why is backup jk "really bad"? And how is backup tracker confusing to newbies?

[mhsmith0]

Backup is backup. So he'd inherit either tracker or jk ability depending on setup.

Late game pr is only useful if Roleblocker is dead; if Roleblocker lives town still up shit creek.

[mhsmith0]

If there is a better idea, then even better. What are you thinking of?

[mhsmith0]

That does nothing to solve the problem of bp being incentivized to claim d1 since it's still potentially useful in a tracker/bp setup. It also makes cop/doc row even more townsided.

[mhsmith0]

One potential wrinkle with a BP->Backup switch is that it lowers uncertainty of Tracker results. In Matrix6, if a Tracker sees a visit but there was no kill, it could mean:

1. He tracked scum to the Doctor's target. (setup C)
   
2. He tracked the Doctor to the scum's target. (setup C)
   
3. He tracked scum to the BP. (setup 3)

In "Backup6", #3 is no longer possible, and so getting a visit to a no-kill confirms the setup is C and enables Follow the Tracker. I feel like this isn't too big an issue since it's a corner case, FtT isn't too strong until one scum dies, and setup C doesn't exactly have a high town winrate already. But it's a consideration.

True, but it's IMO less of an unbalancing issue than BP auto claiming and then tracker knows that he's dealing with a doc if there's no BP claim. And just generally FTT is relatively weak barring an early scum death, and I think it's probably healthy to have a structure where early scum deaths are basically always really bad for scum team.

[mhsmith0]

It's a very low sample size thing, but IIRC it's not a major improvement. It ALSO creates a situation where a doctor fakeclaim becomes much more teneble in goon-goon, since once all claim BP or tracker, you have 50% odds of it being cop only, PLUS you have an easy built-in excuse for being alive after said fakeclaim ("I got roleblocked"). Amusingly, doctor fakeclaim was tried in 1774 and it was JK/goon/goon, and the fakeclaiming scum got autolynched

[mhsmith0]

If it doesn't improve winrates, why is used?

because people THINK it will improve win rates

[mhsmith0]

make osbp active instead of passive, thoughts

Would ruin town's already-shaky winrate in the BP setups, and newbie setups should probably keep to Normal roles for best education of site meta.

Also the problem is that BP roles are incentivized to claim D1 (and towns are debatably incentivized to push for their claim or no claim), and this change fails to solve that problem.

[mhsmith0]

UB only inherits roles from its own team. So can only inherit the JK role in column A, or tracker in row 3.

[mhsmith0]

In terms of helping scum (if you think A is now too strong for town, especially since row 2: cop/doc/rb is the biggest town win rate), it MIGHT be ok to consider turning the roleblocker into a roleblocker/encryptor. OTOH, town win rates in cop/doc/rb are already under 60%, and I'd be pretty surprised if they cleared 60% in JK/UB/rb, so I'm kind of thinking that daytalk for scum is a solution in search of a problem.

[mhsmith0]

It's actually not, for the important reason that interactions with a confirmed town will be non-indicative from everyone since all know the slot is essentially conftown. Objectively, you probably want a BP/not BP claim around the middle of D1, and then a tracker claim (if no BP) sometime before the end of the day phase (but super late in the day is perfectly fine).

Alternatively, you could just have a site meta where people don't policy lynch the BP claims later in the game, and then suddenly there's much less utility to the claim strategy in the first place. But since that's not really on the table, instead we are discussing swapping the BP with some other role that has strong incentive not to claim d1, eliminating the issue wholesale.

[mhsmith0]

I wonder if better ICs could in fact just fix the entire problem without having to alter the setup.

Better ICs fix a different problem tbh

[mhsmith0]

Been thinking on this more, and what about a 1-shot vig? They aren't gonna want to claim, and the role relies on their ability to develop good reads more than any other (Normal) role.

Possibly switch Doc and Tracker as well, so RBer can balance it in A, and Doc can balance it in 3, but that can be worked out later.

1) Vigs in newbies are bad, because you have solid odds of losing a mislynch due to a shot on town, and then town gets fewer chances to get it right. Vigs who take shitty shots are also liable to take a lot of postgame shit, and I REALLY don't think that newbie games are a good place to make "your awful night actions cost us the game" a potential postgame discussion template.

2) There's no real advantage to switching the doc and tracker, and a doctor/1-shot vig as a design (your new row 3) is super scum-sided for a newbie.

3) We're also trying to make as few changes as possible to a setup that essentially just works currently, with one very specific problem that we want to fix, as opposed to potentially creating all sorts of new problems by over-adjusting.

[mhsmith0]

Also Cop/Tracker would ruin many claim dynamics

and town tracker/scum roleblocker is kind of a mess too

[mhsmith0]

That's not the goal tho. The goal is to fix the broken experience of randing BP in current mayrix6 meta.

[mhsmith0]

Random lynching in 7v2 mountainous with a compulsive night kill by scum gets you 29.8% town win, 70.2% scum win.

Perfect scum win = 78% * 71% * 60% = 33%
Perfect town win = 22% * 14% = 3.2%
Town win w one mislynch = ... = 8.3%
Town win w two mislynches = ... = 18.4%
Scum win w one correct lynch = ... = 37%

And then to make an analogue to actual experience you account for town's ability to scum hunt, as well as scum's ability to influence the lynch. Of course, doing so is complicated so mainly you just acknowledge those things exist and that the random lynching model is illustrative more than precise.

[mhsmith0]

if you're random lynching, then day/night chat doesn't matter. because it's a model of a random lynch, and if the lynch is purely random, then you're presuming that the rest doesn't matter.

[mhsmith0]

jk/tracker/goon is horribly townsided, and bodyguard/goon/goon is horribly scumsided. Also bodyguard is just an annoying role in general.

[mhsmith0]

If a tracker knows that he's either with a doc or a bg, he's debatably incentivized to claim sometime day 1. The reason why tracker d1 claims are antitown currently is that there's a 50% chance that he dies immediately N1 if shot because it's a BP that can't actually do anything. The relatively nice aspect of a UB there is that tracker is a sitting duck if he claims d1 50% of the time. Also bodyguard still sucks as a role, and jk/bg/rb is probably even more scumsided than the current jk/bp/rb.

[mhsmith0]

In greys setup, a is townsided, 1 and c are ridiculously scumsided, 2 is mainly just weird, and I don't see the point of having an encryptor in general, much less in a manner that clarifies the setup needlessly. So basically I don't like any of his six possibilities.

[mhsmith0]

Pretty sure grey was just trolling with 154 fwiw.

[mhsmith0]

So IMO the basic downsides to a change like that are:

1) It's a major change when the current setup is roughly balanced (each row/column is between 40% and 60%, usually well between)

2) A PR-heavy setup where there are always two town PRs and always one scum PR would (IMO) put less emphasis on dayplay when I think that dayplay is something that should overall be getting more rather than less emphasis in newbies.

[mhsmith0]

Yeah but you're talking here about vanilla cop as secondary not primary town power; jk/cop/tracker are still there.

[mhsmith0]

Goon Doctor Tracker would certainly not be balanced.

A large sample of newbie game data suggests you're very wrong.

[mhsmith0]

As one of the only people posting in this thread who actually mods newbie games, I can say that, beyond the simple fact that trying to moderate player behavior in the way that LUV suggests is obnoxious to the players, it is EVEN MORE obnoxious to the person actually moderating the game. The way to fix behavior around a "broken" setup (and I still hold that there are substantial antitown effects of the strategy as currently pursued, as well as the fact that it makes a fake doctor claim MUCH more doable for scum) is to change the setup. Trying to mandate that trying to break a setup is a force-replacement or mod-kill offense is the sort of idea that would fairly rapidly lead to no experienced players wanting to play in that setup as well as no experienced moderators wanting to mod that setup.

[mhsmith0]

I think Titus means that mods are overpowering towns and "forcing" town wins through sheer weight of town power compared to numbers and scum power. Admittedly I could be wrong tho.

[mhsmith0]

Along these lines, I think we should automatically modkill anyone who votes anyone of their own faction. This will greatly improve games for some unspecified and unclear reason.

[mhsmith0]

The argument that is being presented is that experienced mods and players will not want to join or run a game if said ban is put in place. Essentially you're telling me that people are going to stop playing mafia on arguably the best site to play it because they want to use a strategy that hasn't proven to be effective.

Actually what I said was that people would go away from participating in NEWBIE games if such a rule was put into place (this has nothing whatsoever to do with normal games or most opens or most theme games). Which is certainly true for this newbie game mod, and I very very strongly suspect is also true for most other newbie game mods (though you'd have to actually ask them to know for sure). It's an obnoxious imposition to put on a mod to monitor whether people are pursuing their wincons in the *proper* way (and again, it's ALSO obnoxious for players - the retention rate for newbies who get kicked from games due to discussing a strategy they never fully understood including the reasons to discourage it is very likely to be pretty low, which then kind of defeats one of tte points of the newbie game queue in the first place). There are rules about trust tells and rules about toxicity for good reason; without them, games can be obnoxious and unpleasant for most to all of those involved. There are NOT rules about a shitload of other things that people can do, also for good reason; people are, in general, allowed to pursue their wincons as they see fit. We don't modkill or force replace for any number of sketchy things like:
- self-voting (or self-hammering) as town;
- fake claiming as town;
- bussing in a relatively obvious way;
etc.

[mhsmith0]

By the logic of the people saying "either stop using or ban the strategy", it's perfectly fine to have setups involving Cop+Doctor with no breaking-up factors as long as we either ban Follow the Cop or politely tell people to not invoke it. Nowhere else on-site has it been okay to continue running a busted setup while simply requesting players to not exploit it. You can't put the genie back in the bottle.

Now that said, I don't think this case is at all as urgent as people seem to be treating it as. Sure the current situation is not optimal, but it's hardly bad enough that we have to change the newbie setup yesterday. The strategy isn't even conclusively beneficial for town according to my statistics so far. (I want to say something else but don't know if it would involve ongoing games.)

In any case, it's clear this discussion has gone off the rails, and it's probably best for the listmods to decide what to do from here on out.

So because it's never been done before we must make sure to not do it.

Apart from a pointless appeal to tradition, is there a point to thinking like this?

FWIW, if we're going to pretend that rb is making a SERIOUS argument here (he's solidly implying that it is in fact reasonable to ban follow the cop as a corrolary to it being fine to ban the BP/tracker claim idea), I'd just point out how ridiculous it would be to ban discussion of Follow the Cop as a strategy. In a game with a cop and a doctor, does the cop just not claim a cop? Is the doctor banned from protecting the cop? Would it be anything other than patently ridiculous to ban people from discussing follow the cop, especially if a cop claims after a scum roleblocker (or MAYBE role cop) death? I mean hell, if it's a force-replacement penalty, then you could even get into an angleshooting situation where scum who are losing bring this up just to get force-replaced so that technically they don't lose (depending on how they count), or just so that they don't have to fight back against a tough situation.

I mean, that's on the face of it ridiculous, but if we're going to treat the argument seriously, there you go.

[mhsmith0]

[mhsmith0]

And ic doc goon as fairly obviously scumsided for a newbie game.

[mhsmith0]

The entire point of the discussion is to de-incentivize d1 claims. If that problem isn't solved, there's no reason to change the matrix.

[mhsmith0]

For a newbie game??? That's a bizarre set of non standard mechanics for a game that's supposed to be reasonably representative of a "typical" type of setup. And who even knows how much newbies would NOT enjoy being marked for death early on.

[mhsmith0]

Basically player X, if visited by a FN, announces that he was visited by the FN without saying who the FN is. The problem with it is that, once again, you're in a situation that potentially incentivizes early claiming (in this case, it'd be day 2 where the tracker is incentivized to claim if there's proof of no FN [other than the low chance the FN visited the person who died] ). I think that if we're trying to create better standards of play, it's good to have setups that punish and de-incentivize early claiming.

The idea of backup6 does this fairly nicely for a tracker; if there's a backup, then tracker gets shot with zero protection and there's no investigation that night, which means that a true backup claim is explicitly antitown in row 3, which means that scum are rewarded and town punished for a claim situation, as it should be (actually, even a "backup or doc" claim would then just get shot and die in goon-goon, which I think is also fine).

In column A, you get more town power compared to the JK/BP/RB setup, and you basically have a situation where it's super swingy on whether the roleblocker gets lynched early or not (just like cop-doc, since follow the cop is nearly a mechanically locked win for town given a d1 roleblocker lynch). And I think that's also fine, because it forces a game state to be highly incentivized on whether town can catch or scum can protect the roleblocker, which is kind of a normal outcome for any game with a key scum role (roleblocker, godfather, etc).

[mhsmith0]

I have to do the math but I'm not 100% convinced that is is anti town for the backup to claim early either.

In row 3, backup just dies night 1, which is obviously very bad for town (having a living tracker late is very good, and that makes 1 or 2 PRs dead).

In column A, jailkeeper is then forced to choose between using the JK defensively (preventing any implied guilty possibilities) or offensively (leaving the backup vulnerable to death). It also enables scum to narrow down the POE of who is jailkeeper (both from the claim directly and possibly via interactions with the claim), and it enables them to know it's not cop-doc, which helps them plan night actions a bit more effectively. It also means that if the jailkeeper is ever outed, it's basically a big turkey shoot for scum as long as the RB survives.

[mhsmith0]

Why would you want to make the setup more complicated? Is there an actual gain here? Is JK/UB/RB or Tracker/UB/goon somehow unbalanced or unhealthy? Because FN is a pretty weak role, to the point where I pretty strongly suspect that it'd lead to a meta where the two setups with FN (instead of BP) have the lowest town win rates by a decent margin, and I'd think some weird blend of FN/UB adds to the complication without creating any actual gain in terms of utility or fun.

[mhsmith0]

Discouraging the early claim strategy is literally the whole point of the excercise. So if you're not discouraging it, then why bother?

[mhsmith0]

Tracker UB is less gross than doc tracker tho I think, since if scum can hit tracker then UB, there are no investigative results. Agreed on rb day 1 lynch but a scum team who loses rb day one deserves to lose so I don't mind locking that result in even harder.

[mhsmith0]

It doesn't. It just makes it very very hard to win. N1 scum hits jk who doesn't announce target, and then it's n2 for backup creating one clear, and that's VERY winnable. But losing a major pr just for town cred SHOULD be a really rough road for a goon, to the point where he should need to run the table on great shots in order to win.

[mhsmith0]

Discouraging the early claim strategy is literally the whole point of the excercise. So if you're not discouraging it, then why bother?

I meant Backup6 doesn't (or shouldn't).

In either a or 3, backup is strongly deincentivized to claim early because either way he's a sitting duck, whereas if he lives a while he can become very strong. This is ESPECIALLY true in tracker but is pretty true in rb/jk too.

[mhsmith0]

Basically the problem is that whatever you put in that cell has to be stronger than (IC if claimed + ) to justify not following the early claim strategy.

That's not really correct. The current early claim strategy fundamentally rests on a few tenets:

1) If BP claims, he can't be quickly killed because he has a BP. This means that the mechanical cost to claiming as BP is limited (the cost to effective dayplay of an overly quick claim still exists, but people don't think about that so it doesn't really factor into the current meta even if it should). Also relevant is that BP can't actually DO anything active, so it's a role that scum might not shoot later in the game if they think they can get a mislynch on this slot for "fakeclaiming", whereas various other roles have active abilities that in general are self-resolving because scum CAN'T afford to allow them to continue living.
2) If you KNOW that you're in tracker/doctor (because there's no BP claim), then tracker can claim at his convenience knowing that there will be a doctor protecting him (whereas if it's tracker/bp he just dies)

OTOH, if there's a UB who claims D1 in tracker/ub/goon, then he just dies n1 most of the time, and if he doesn't (i.e. scum think they know who is tracker), then he just dies right after the tracker and he never gets an investigative shot off. This is MUCH worse for town than if the UB remains hidden and continues to exist as someone who can potentially investigate after the tracker dies.

Basically, an early claim is incentivized if the benefit to claiming (not getting lynched later in game for BP; automatically getting the doctor protect for a tracker after it's known there's no BP), is greater than the cost to claiming (no cost to BP since he can't do anything at night, other than messing with dayplay, but no one cares about that; cost to tracker is after the doc dies he's a sitting duck).

[mhsmith0]

I strongly do not support the Backup idea because it leaves town with absolutely no power if both of their PRs get outed.

To me that's a feature, not a bug. It being bad if both PRs are outed means that we go back to a structure where there's a meaningful cost to outing yourself as a PR (I'd also note that in JK/UB/RB, town still has power if the RB gets lynched before one or both of the JK/UB).

[mhsmith0]

Can you clarify why? I feel like you have two basically unlynchable PRs, with a nonzero (even if low) chance of them getting another guilty.

You also have some interesting interactions coming with the tracker; specifically, a tracker claim has 50% of getting doc'd, which means that scum have to roll the dice by either shooting a tracker or hunting the doc, which can make things tough on them potentially.

I'd expect tracker/backup/goon to be lower win rate than most other rows, but at the same time, a day 1 lynch is kind of crushing to scum, though not as much as in tracker/doc/goon. I also expect that town win rates in jk/ub/rb would go up from jk/bp/rb, which means that overall it ought to be a bit closer to 50% win rate.

[mhsmith0]

As much as I love a vig, they have no place in Newbie Games.

+100000000000000000000000000000 on both points

[mhsmith0]

If you think that things can get toxic towards people who fucked up normally in a game, just wait until you have a newbie take atrocious shots because he has no clue what he's doing in a newbie game and watch the results. Vigs are fine in games where's a presumption of at least slight competence among players, but in games where you presume that a majority of the people are completely new or pretty new, that's a recipe for disaster.

[mhsmith0]

I noted the issue with vigilante
Maybe make it a oneshot vigilante?

That doesn't fix the issues w vig at all. In fact it makes it WORSE to potentially have a newbie take a bad shot n1, hit MYLO, and on top of that never get to shoot again.

[mhsmith0]

With great power comes great

responsibility

blame

[mhsmith0]

Jailkeeper is a pretty awful role for newbie games as well. There's just too few players and the odds of blocking kills are just too high with that role.

And yet, in jk/goon/goon, even a block doesn't force a guilty result, as you reasonably have protected the target, or scum could have (suboptimally) no killed. I think it's a fine role for a newbie game; it gives the power holder a lot of potential strength, but it has to be played effectively to be good, and unlike cop it doesn't spoon feed information to the player who has the role.

I'd also note that JK/goon/goon has a town win rate under 50%, but not far under, which means that balance wise it's quite good.

[mhsmith0]

I mean, I'd say cop is a kind of shitty role because the game becomes more heavily focused on "did scum kill the cop or his target" and there's relatively less skill involved compared to figuring out who might be a jailkeeper, who said jailkeeper might target, and working around that (and the reverse for the jk) compared to cop which is basically just "don't die, don't cop someone who scum is very likely to shoot".

OTOH cop is pretty standard role so might as well get people used to it. But cop as a role is pretty lame IMO.

[mhsmith0]

Make it 1 shot then. Maybe make all the PR's 1 shot or 2 shot. Just throwing this out there.

Towns already have <50% win rates, in that world it'd go even lower.

[mhsmith0]

Problem with unlimited JK in the confines of a newbie game is that you start with a 25% chance of blocking a kill regardless of how it's done. The deeper you get into the game, the higher those odds get.

Why is this bad? Like, this is basically saying that the problem with a role whose primary function is to prevent kills it that it has realistic odds of randomly preventing kills? (I'd also note that odds are lower than that, since it ignores the case where scum shoot the JK, or where it's JK/BP/RB, which lessens the information that the JK can get from a stopped kill, which helps to nerf its OPness a bit)

[mhsmith0]

I would argue it's not 'good' to stop kills just for the sake of stopping them.

A jailkeeper makes a stopped kill in a game about twice as likely as one that has a doctor and two stopped kills = extra mislynched required for a scum win.

This scenario is pretty rare with a doctor but is probably more possible than desirable with a JK.

You're still failing to answer the fundamental question of why kill stopping is bad.

In 7v2 mountainous with mandatory NKs, towns have 30% odds from lynching randomly (and in newbie games, I'd venture that random lynching is probably a pretty good description of behavior). To improve those odds, you give town some kind of power. Said power can be investigative, protective, blocking, etc.

So in that context, I don't see the problem with "maybe scum will need an extra mislynch to win" (since achieving that point requires pretty effective play on the part of the jailkeeper, since just one kill stop doesn't create an extra mislynch). If town gets an extra mislynch via TWO stopped night kills, that's something that town has earned, and good for them if they then enjoy the reward. What's the problem with that?

[mhsmith0]

So scum are in trouble if they

1) Lose a scum member early
2) Get targeted by the town PR
3) Fail to kill the town PR by say n2

... ok? Like, this all seems fine enough to me? In jk/goon/goon, town doesn't get to a dominant position until the first scum lynch AND scum failing to hit the JK, so to avoid that, scum need to not let any of their team get lynched or nail the jailkeeper first thing after said scum lynch.

[mhsmith0]

I mean, one of the defining features of matrix6 as a setup is that scum are basically screwed after a day 1 scum lynch, unless they're in a roleblocker setup and the lynch is a goon (and even then it can be tough). The precise nature of that screwing is different in jk/goon/goon than it is in, say, tracker/doctor/goon, but you basically get to the same place more or less.

[mhsmith0]

I'd also say that if you generally expect quality town play, then towns are overpowered in nearly all of the matrix6 setups, but since you don't really get great play since it's a newbie game, I'm not sure it's super relevant. Basically, in any of the six setups, there are mechanical ways for towns to crush the scum given early scum lynches, or recover from early town lynches. I don't see why jailkeeper sticks out as more annoying than any of the other setups, and indeed I'd say it's probably less so, given how annoying cop is as a role, for instance.

[mhsmith0]

It's not fine. Being at the mercy of a roleblocking role as scum is literally the worst situation you can be placed in as a player. It's worse than follow the cop.

It's fine if say you're in that situation in a large game because the odds of it happening are much lower. But it's completely possible to commonly be stuck in that kind of situation in a newbie game through no fault of your own. Just because the game size is too small to handle the negative town side swings of a jailkeeper role.

I mean, in general if a goon is getting lynched day 1 they'll fake claim tracker to draw out the cop/jk/tracker CC, and if scum #1 can't manage to pull that off, or if scum #2 can't even dodge the n1 jailkeep shot after a scum lynch, then the scum team deserves to lose. Still not really seeing the issue tbh.

[mhsmith0]

It's skill based in the sense that the jailkeeper is likely to jail a scumread on n1 given a goon lynch, which means that the other goon needs to not be scum read by the jailkeeper, i.e. "dodge the jailkeep". I mean the JK COULD just roll a die and randomly jail, but that's fairly unusual I'd think. Which means that it is, in fact, skill based.

[mhsmith0]

I agree. It's relying on power roles a little too much. I am fine with a balance/compromise though.

Did you know that town has under a 50% win rate currently both overall and in 5/6 setups? So if your proposal boils down to "let's nerf town power", I fail to see what's driving it beyond a personal preference for power roles being nerfed. Because it certainly isn't a proposal that would make the games more balanced.

[mhsmith0]

Well I mean in the case of dodging a Jail you can be punished for looking too town as well. Say 5 players left and you're the sole remaining scum and the jailer targets you because they think you are town and think you're the most likely kill target. Now you've just lost the game because you looked too town.

Any jailkeeper who jails to save instead of to block when there are five players left is a jailkeeper who doesn't understand his role very well.

I'd also note that if you're at night 3 with the JK alive then, as scum, you've failed to PR hunt effectively, which is something that's on you and your team (I mean, if we're talking about random odds, random night kills have decent odds of hitting JK n1 or n2, even more so if scum have identified any players who are clearly vanilla townies)

[mhsmith0]

viewtopic.php?f=5&t=39739
45.8% town win rate overall

and apparently setup 1 (jk/goon/goon) is now back to 50%

[mhsmith0]

they'd be higher almost certainly. overall the setup is structured to be friendlier to newbies who don't really know what they're doing, kind of like training wheels.

[mhsmith0]

Teaching is (theoretically) why we have IC roles. I'd certainly agree there needs to be a better job done on that front, which is a separate topic of discussion entirely.

As far as win rates go, people tend to be pretty unhappy if they get stuck in a situation with a very low win rate and a "anyone can win" setup is inherently more interesting and fun (so you don't need exactly 50%, but it really ought to be pretty reasonably close, especially when you have enough of a large sample size to say that the win rates are legitimately representative). Part of the issue with BP setups is that both of them have town win rates close to 40%, which is about as low as I think is at all reasonable for this kind of setup. The more compelling reason why we're discussing changes to the structure is that a d1 BP claim strategy has messed with the learning experience to the point where a change is needed to eliminate that particular issue.

What we're not really doing, and I don't think that the people in charge (I'm not one of them fwiw) really want, is a major, radical departure from something that, as a whole, works. So, for instance, "I don't like JK setups because they can fuck over the scum team potentially" when the actual win rate is 50%, isn't really I think a valid objection. I'd even go further and say that JK as a role is more interesting and compelling than cop, and that it more directly rewards effective play and punishes ineffective play, both from the jk and from the scum team. Whereas cop is just "push a button, get a result unless you or your target dies", much less useful for learning IMO.

[mhsmith0]

a setup that is actually mathmatically 50/50 balanced

[mhsmith0]

IIRC the mathematical balance of a setup inherently relies on semi arbitrary assumptions, such as (for instance) random lynching, random night actions, etc. So when you say "mathematical balance" it is difficult to know exactly what you are trying to say.

[mhsmith0]

Does it? BP/Tracker is pretty nerfed for town, since BP can't actually do anything active, and you'll never stop two nightkills unless scum choose to no kill for whatever reason. Tracker/backup means that if scum ever blow a shot (missing PR) after first scum lynch, it's a guaranteed inno or guilty, which is somewhat strong I think, whereas in say tracker/doctor, if the tracker dies n1, then town has nearly zero power left.

[mhsmith0]

Yeah I want to attempt to start being an IC but this is holding me back. Do the cons outweigh the pros for just replacing BP with UB? Explain like I'm 5.

No they don't. Because I say so.

Now who wants candy!

[mhsmith0]

You have to get in my van before I can show you

[mhsmith0]

UB doing anything in row 3 is somewhat low odds, but IF he's able to get a shot off, it becomes super strong, with very solid odds of being a major difference maker.

Which I think is fine, since it's a supporting role, not the main role. The supporting role on the matrix SHOULDN'T be super strong, especially in a row or column without a roleblocker.

[mhsmith0]

[mhsmith0]

On row 3, it can, with low probability, counter a fake tracker claim that happens on day 2 or later (it can similarly confirm a tracker claim as real). But that's pretty much it. Overall pretty low utility.

And that ignores the realistic possibility that it will create a false guilty (tracker tracks role cop to the n1 death), which not only potentially screws town with a fake guilty, it just as importantly nerfs the tracker's ability to automatically know that a seeming guilty result is accurate, thus nerfing pretty severely the utility of the tracker (already a fairly weak role in matrix6)