Open 794 Pick Your Power X/Y (finished)

[Hoopla]

hello friends.

i have a game-winning plan right here:

we execute downwards from the top of the draft.

then, we hope scum NK from the bottom upwards.

this should end up winning us the game.

VOTE: drdolittle

[Hoopla]

VOTE: hoopla wonder if i shld yeet a policy vote onto the most famous player on the plist

i believe my celebrity status earns me a ticket to end-game.

or at the very least, a free club sandwich.

[Hoopla]

you can have a mafia celebrity loyalty status card. i'll put a couple of punches on it for you.



close.. but not this time

okay, just give me the bread.

[Hoopla]

ok now that i actually read the posts, here's a vote on someone that i think is slightly >rand (that symbol means more than right i haven't had my coffee yet) scum

VOTE: gypyx

discuss

i falsely townbinned scum-gypyx on D1 in a recent mini normal; chalked up his early game waffling + a strange "slip" as overeager newb-town behaviour.

little did i know...

[Hoopla]

interesting. are you noticing similiarities now?

i remember him being very active early, participating in multiple conversation threads. a lot of it surface level admittedly, but i think i was fooled by his early-game energy.

there wasn't much early-game joking around from him. seemed focused on his "scumhunting" busywork; asking questions, general commentary - you know, all the good stuff you're supposed to do.

if anything, his relaxed beginning in this game is different.

[Hoopla]

i don't think it's necessarily town!indicative this time around though. you?

may have rolled town.
may have figured out that as scum participating in the jokes & early game narratives is a safer opening than the forced scumhunting routine.

i perch happily on this fence.

[Hoopla]

hoopla, listen. i haven't properly played mafia in like 4-5 days now. withdrawl is getting to me. i am *itching* for a good argument. how are we supposed to have it when you're on the fence???

how about instead of an argument we do some number analysis?

~~

i've collected all the number selections scum have made in the

16

14 runs of this setup (one run omitted due to scum not being able to coordinate pregame, another game abandoned).

so, in those 14 games, how often do you think scum have gone for a WIFOM-play and selected the same number as a buddy? have a guess.

[Hoopla]

yesterday i was skimming the run of pyp where skitt was scum, and i think someone there said it's happened once? the numbers being 2/2/6

i dunno if it's happened since, but i'm too lazy to check.

fwiw my guess weould've been something like 3 so

none.

that game you're talking about was open 630. the mod made a mistake and didn't let scum talk pregame, which resulted in two scum picking the same number. so, never willingly.

scum simply don't like sacrificing good draft position for potential WIFOM trickery. well, it isn't really WIFOM since scum never do it. we can fairly confidently declare there is only 1 (at most) scum within paired and tripled numbers.

you can even see visually that scum end up disproportionately higher in the draft compared to the bottom half. it's actually quite glaring how much of an advantage knowing two other numbers and avoiding them can be;

draft positions of 42 scum in 14 PYP games:

1st - xxx
2nd - xxxxx
3rd - xxx
4th - xxx
5th - xxxxxx
6th - xxx
7th - xxx
8th - xxx
9th - xxxxx
10th - xx
11th - x
12th -x
13th - xx
14th - xx

we execute in the top half of the playerlist today, and preferably a singleton.

[Hoopla]

this made me worry that hoopla is the scum doubling up, then i looked at the draft and i see both she and ico picked 5 and i thought "pls no"

just for the sake of transparency, i have doubled up as scum in a team mafia PYP game many years ago. me/zito/faraday choose 6/6/9 i believe?

i believe that is the only time its happened if you include the other 4-5 non-central park PYP games.

[Hoopla]

it gets worse;

draft numbers selected by 42 scum across 14 games;

1 - xxxxxxxx
2 - xxx
3 - xxx
4 - xxxx
5 - xxxx
6 - xxxx
7 - x
8 - xxx
9 - x
10 - xx
11 - x
12 - x
13 - x
15 - x
20 - x
22 - x
42 - xx
413 - x

[Hoopla]

well then
guess it's over
VOTE: datisi
ggwp

i gave you a few pages to hang out and have a laugh.

i'm not a monster, you know.

UNVOTE:
VOTE: datisi

[Hoopla]

ok that's enough self-voting for this game, wouldn't want this to spiral into an actual serious wagon now would we

VOTE: gypyx

hoopla, out of curiosity, why wasn't your RVS vote on one of the 1's?

gypyx, do you have an opinion on like anything that's happened

in actuality, i'd probably prefer to wagon a player who hasn't been doubled/tripled up.

but i am amused at your slot hitting the jackpot on two separate metrics.

[Hoopla]

hoopla, I would like one essay on number analysis please.

in actuality, i'd probably prefer to wagon a player who hasn't been doubled/tripled up.

can you elaborate on why you'd prefer this when you can pls ty

it was late last night, so i didn't have enough time to finish collating the data.

one more set;

of the 196 slots across 14 games, 42 are scum, 154 are town. a given slot individually has a 21.42% chance of being scum.

how often does each slot fall on a singularly chosen number? a double, a triple etc.

singleton

mafia - 20 (29.85%)
town - 47

double

mafia - 12 (18.2%)
town - 54

triple

mafia - 6 (15.38%)
town - 33

quadruple

mafia - 1 (12.5%)
town - 7

quintuple

mafia - 1 (10%)
town - 9

sextuple

mafia - 0 (0%)
town - 6

[Hoopla]

to me, the frequency with which scum land on singular numbers is the most influential data point... if we're to look at the numbers.

in games gone past, i have probably relied on them more than necessary at the expense of scumhunting. i intend to use them as a guide this game, rather than an out and out solution - because it isn't that. we can't just rely on numbers. but they are a good framework to help visualise where scum are expected to be, and can help guide wagon choices.

as the game progresses i expect actual suspicions to arise more naturally, but the distribution of scum in the top half of the draft, as well as in the single numbers is too big to ignore on day 1, which is the lowest information phase.

Am I gonna have to bring up the

duck

dragon

elephant in the room regarding Hoopla's plan?

The people toward the top of the list are more likely to have PR's. And, uh, PR's are how you win games.

you win games by executing scum. allowing scum vigs/PGO's/blockers to roam free because they

might

be a town PR is a recipe for settling for a low% execution.

[Hoopla]

an example of how we should use number analysis;

- gypyx is 11th in the draft (a low frequency slot for scum)
- he also chose 11 (a low frequency selection for scum)
- he is also on a doubled up number (an average frequency outcome for scum)

i'd want to be quite confident gypyx is much more likely scum than random if we're to wagon there.

[Hoopla]

in the mean time;

VOTE: something_smart

[Hoopla]

Also, italicizing

might

is pretty disingenuous. According to your stats DrDolittle for instance has a 30% chance of being mafia. That puts him at a 70% chance of being a town PR, and probably a strong one as well.

in my eyes, the damage scum can do with vigs/PGO's/redirectors etc. is worse than the value a town PR can yield in slot number 1. because if they're town, they're likely to be blocked/NK'ed/redirected etc.

winning top of the draft actually kinda sucks as town since you're a magnet for interference. whereas scum in top slots can wreak havoc because town's tend to leave them unchecked out of fear.

sometimes you have to step up and execute PR's or PR claims to win games.

[Hoopla]

The biggest reason to keep the higher positions around are because some of them will self-resolve by virtue of being NKed.

^this is decent a counter-argument to leaving top slots alive.

anyways, this is probably enough naval-gazing theory pontification from me, so i'll let others jump in as i read the thread.

[Hoopla]

Hey everyone! Quick question, are wolves normally able to talk about what numbers they want to pick like on other sites? Because if that's the case it's probably 1 max in each group.

i agree with the grumpy cat re: clover.

questions like these, especially from new-ish players are a pretty reliable towntells. most mafia don't think to fabricate questions like this - or at the very least, it takes a cheeky scumbag to work this angle.

that was probably the most alignment-relevant thing that has happened all game. points for grumpy cat for picking it up and correctly interpreting it.

[Hoopla]

i'm unsure what to make of the gypyx fake-hammer and fake-reaction fiasco.

when you have players acting in a disingenuous way in order to harvest reactions, you risk players responding in kind (which is what gypyx did).

so, now the question shifts from "was gypyx's reaction genuine?", to "was the decision to fake a reaction to a fake hammer town?", which is just too flimsy of a premise to really parse anything meaningful imo.

if pressed, my intuition leans town on gypyx because it's kind of overt and bombastic to respond to a fake hammer with an equally facetious reaction test.

[Hoopla]

Like for example in this post. Hoopla could notice this is her scum PT as well and say guys let's both pick 5. I can use stats to defend us. Im largely uninterested in using these stats to figure out alignment, especially based on day 1

well yeah, obviously if i am scum i could rig the deck and get me and my buddies to pick unlikely combos i can defend with numbers. that goes without saying.

however, if i'm town, there are some serious inferences we can make throughout the game with the numbers.

[Hoopla]

even if you are town, how do you know scum players haven't thought about your line, and have players like you in the lobby to provide them with free defense?

independently, I don't buy this line of thinking, and I don't think you should be townread for suggesting it.

because across 14 games, not a single game has seen scum double-up on a teammate's number. why of all games would scum suddenly think to subvert number analysis in this 15th run specifically? i think we can infer based on history that scum won't double up. it's clear that the only realistic way it could have happened is if i'm scum.

but if i'm town, you must admit, these are some telling statistics.

the disproportionate distribution of scum at the top of the draft/on single numbers is jarring. the only argument against the numbers is how much we should be weighting them in our decision-making. if you think i'm scum obviously weight them less.

i'm not even suggesting we should blindly execute a singleton. i could settle for a doubled up player. as far as i'm concerned, the main players that should be off the table today is the trio of word/crayons/iamausername. the odds of one of these players being scum is so low we'd need overwhelming evidence to take a shot there.

[Hoopla]

Wrt to the latter quoted line, everyone is focusing on the shitpost and forgetting about the distinctly non-shitposty 67. Though it's obviously not a be-all end-all, it is a serious plan.

me stating my preference for pressuring the singletons first isn't a "plan" - it's just that: my preference.

make your own decisions.

[Hoopla]

I like, just don't agree with this, especially this line:
"the main players that should be off the table today is the trio of word/crayons/iamausername. the odds of one of these players being scum is so low we'd need overwhelming evidence to take a shot there."

why?

at most there is one scum in the triples. and across 14 games, there have been 13 triples; of which, only 6 have had scum. half the time all three are town. the other half it's 1 in 3.

they may as well be the basis of a townbloc in my eyes.

[Hoopla]

the disproportionate distribution of scum at the top of the draft/on single numbers is jarring.

and I think many people have pointed this out already but you need to do a similar distribution for town to compare...

for distribution of where scum land in the draft, you don't need town to compare, as there are a fixed amount of slots: 1st to 14th.

you need to compare for the breakdown of singletons vs. doubles vs. triples etc, as a game can have 10 singles or two, 5 doubles or none etc. for that data set i

did

list the distribution of town to compare.

[Hoopla]

do you think 6/39 is significantly different from 3/14?
regardless I'm just gonna take the worst's advice on this

it's slightly significant if you compare to the average frequency of all slots to be scum.

but why not compare the trios - 6/39 (~15%) to the singletons - 20/67 (~30%).

that's pretty saucy.

[Hoopla]

anyway, i'm also happy to pipe down about number theory.

for a while...

[Hoopla]

actually, a question for the worst;

fwiw i'd also fairly-strongly encourage my team to pair

why would you strongly encourage your team to pair your picks?

[Hoopla]

oh for a few reasons! it's exciting

keep talkin'

[Hoopla]

can we leave this for postgame?

no, i actually want to know.

as far as i can tell, the only reason to tank your team's picks as scum is if you're concerned about number analysis and want to subvert it.

the impression i get from your posting is that you don't think the numbers are relevant, so why bother subverting what you don't believe in?

to me, it seems more likely your claim of "i'd consider pairing numbers as scum", is some casual shade intended to cast doubt on said numbers.

~~

i just finished my post as you posted. in the interest of transparency, i'm going to post anyways it since it shows why i asked you this, but i believe my characterisation of your thought process may be wrong on first glance..

[Hoopla]

very well kind duckling.

lets transition this conversation into something less theoretical;

who are your top scum reads? any town reads?

[Hoopla]

I would seriously consider it, but only because hoopla is in the game.

(Huge fan btw)

^ clearly this cat knows how susceptible i am to shameless flattery.

[Hoopla]

ugh i feel like i have too many "null but kinda scummy" slots right now

hoopla, do you have any non-mech reads/hot takes to offer

this game needs more content

at the best of times, it takes me a while to really sink my teeth into a game (even one's without distracting mechanics). some of that stems from pessimism about D1 inaccuracy. like, i know you have to "fake it til you make it", so to speak, with early reads but ugh. townhunting i find somewhat easier d1, but like you, i tend to end up with a lot of null reads.

i've come to accept that that's okay on D1.

my de facto coping mechanism is aggressive bandwagoning on non-town reads. after a while, real information (wagons/claims etc) seep into the game and i find it easier.

that doesn't really answer your question, so to throw you a bone;

i like clover, grumpy cat, crayons for town. i've been flipflopping on the worst, but feel better about him based on our most recent interaction.

i think you've been a useful contributor too, though i admit i haven't seen your play before, so can't really trust my judgement there yet. i tend to favour utility executing D1 and like keeping around voices of reason like you - because in the event they are town, it will be +EV.

composition of the umlaut wagon looks decent, soooo

UNVOTE:
VOTE: umlaut

[Hoopla]

can you elaborate on this? like i'm not sure exactly what part of my play or posting so far feels like ~useful contribution~ (or what part you consider it to be that)

i liked our early interactions. 369 seemed a fair criticism of my number theory, and had other useful thoughts.

your "secret scumread" on me that you've been mulling over also seems a good sign. i can be a polarising character. one of the ways i rely on reading people early is how they react to my presence d1. quite often scum tend to uncritically suspect me or they'll placate me and throw me in the townbin me. less often, they'll fake a nuanced position on me. perhaps i'm a sucker for when people display they're *actually* thinking about my alignment, but it tends to be a useful tell from my experience.

i don't get your triangle diagram though. the vertical axis is just uncertainty?

[Hoopla]

why is the iamausername wagon taking off?

in my eyes, his posting seems in keeping with his usual d1 town meta; low-activity, but offers piecemeal observations, and interesting soundbites.

i suspect his posting style may not jive with this crowd.

[Hoopla]

hoopla my nuanced read on you is that you are a cake which has 1 candle and shows you are forever 1 year old. Since eternal youth hasn't been invented yet it means your account is controlled by a constantly shifting array of 1 year olds that each spend a year on the account. This also means you are breaking site rules

Reported

busted...

by a cat!

[Hoopla]

imusername's takes are too selective. Like, he's taking special care to criticise very specific things which he thinks look good to poke at or are reasonable to. The lack of ~feels~ or big picture takes is generally scum-indicative.

i think that's in keeping with his style. he tends to focus on unique/specific things instead of dropping recycled platitudes.

for example, his commentary (summarised in post 160) about something_smart's strange timing about criticising my "plan" was a good observation.

[Hoopla]

kind of but not really. i'm not feeling "certain" about any of my read so far. however, the closer someone is to the "ogofidk" point, the more their slot gives me a headache, is a pain to form a read on, makes me confused, makes me have paranoia, etc, etc. i'm more of an emotion/gut driven player, and it's slots that make me... uneasy? lemme know if i'm explaining it badly.

this sounds like a very mysterious triangle.

[Hoopla]

There is something I dislike about this but I'm not sure how much of it is just that the post ends with a vote on me.

hmm lets test it out;

...


at the best of times, it takes me a while to really sink my teeth into a game (even one's without distracting mechanics). some of that stems from pessimism about D1 inaccuracy. like, i know you have to "fake it til you make it", so to speak, with early reads but ugh. townhunting i find somewhat easier d1, but like you, i tend to end up with a lot of null reads.

i've come to accept that that's okay on D1.

my de facto coping mechanism is aggressive bandwagoning on non-town reads. after a while, real information (wagons/claims etc) seep into the game and i find it easier.

VOTE: iconeum

[Hoopla]

There is something I dislike about that but I'm not sure how much of it is just that the post ends with a vote on Iconeum.

i think my voting software has been hacked

...


at the best of times, it takes me a while to really sink my teeth into a game (even one's without distracting mechanics). some of that stems from pessimism about D1 inaccuracy. like, i know you have to "fake it til you make it", so to speak, with early reads but ugh. townhunting i find somewhat easier d1, but like you, i tend to end up with a lot of null reads.

i've come to accept that that's okay on D1.

my de facto coping mechanism is aggressive bandwagoning on non-town reads. after a while, real information (wagons/claims etc) seep into the game and i find it easier.

VOTE: drdolittle

[Hoopla]

don't like some pressure on iamausername, if that even is his real username?

i'm contractually obliged to offer at least one token defence per game for fellow 08'ers.

a union scheme to thwart the "kids these days".

[Hoopla]

do you have any numbers on how often these 08'ers roll scum?

i don't remember ever seeing it, but i'm sure it's happened once or twice.

[Hoopla]

UNVOTE:
VOTE: iconeum

[Hoopla]

how can you vote me after that pagetop

i'm getting in on the groundfloor while your wagon is fresh.

now that we have grumpy cat on board too...

i sense this ticket will be too hot to pass up.

[Hoopla]

Which posts suggest that hoopla has no agenda?

believe me, i have an agenda.

scum are going to HATE it.

[Hoopla]

looks like this is where the action is at;

UNVOTE:
VOTE: marquis

[Hoopla]

I'd also expect scum not to come in on page 5 with an explicit announcement that they're not going to play right now, when Things are already happening in the game.

contrary to popular belief, things aren't happening. this is quite possibly the lowest content 25+ page game i've been in.

it's been largely pointless waffling, people jostling for allies, and empty threats that don't carry the risk of execution.

lets actually put someone to L-1 and see what happens.

marquis will do.

[Hoopla]

finally, the first wagon battle.

umlaut vs. marquis

an intriguing contest. lets view their stats;

UMLAUT

- number pick - #21
- draft position - 3rd
- posting frequency - 31/744
- avatar cuteness - 8/10

MARQUIS

- number pick - #1
- draft position - 4th
- posting frequency - 9/744
- avatar cuteness - 9/10

~~

the town has narrowed things to two of the top four picks. both choices have remarkably similar cuteness and posting frequency, though you'd have to say umlaut edges marquis on content. are either of them actually scum? hard to say. number analysis views both these slots in the "hot zone™", but we perhaps haven't seen enough behaviourally to make this a slam dunk execution. role claims are looming to be a HUGE factor in determining which way this one will fall.

lets go to expert field commentator the worst for his thoughts.

[Hoopla]

Clarify “hot zone”.

a mysterious blend of the different numerical forces underpinning our game.

believe me, these puppies are hot!

[Hoopla]

I saw marquis picking on some primary schoolers in the mall the other day

he does have that look about him, doesn't he?

@marquis

CLAIM TIME!

[Hoopla]

ah yes,

the waiting game...

[Hoopla]

when in doubt, always bow to the altar of numbers.

[Hoopla]

am i correct to think (becuase i've fucked the math up a couple of times this game already) that the fact marquis picked number 1 in draft makes him more likely to be scum fmpov (similar to asking "does picking 1 make you more likely to be scum" but we're the only two 1-picks and i know i'm town so for me it's just him while to the rest of y'all it would be "is it more likely than random that there is scum in the 1-picks?")

which has nothing to do with his play so far.

from your perspective (if you're town), you'd put him in the same category as the other singletons. maybe even higher since he'd be pseudo-singleton + a highly picked scum-number.

[Hoopla]

We should lynch marquis

i want a roleclaim from him first.

[Hoopla]

hmm this is rough...

but lets look at the glass half full;

perhaps we're doing

so well

in this game that the scumteam is collectively panicking. this display of extreme stalling being a desperate ploy to get things back on their terms.

[Hoopla]

1-shot vig, you say?

even if we don't execute A50 today, i don't think we should try to direct his shot. if he's scum, he can easily ignore the direction if it doesn't suit them, and if he's town, it's better that scum don't know whether they should interfere with his shot or not.

i think i'm in agreement with this.

another reason to shoot is from the perspective of putting into odd-numbers for the rest of the game. his vig-shot - even if on town - doesn't come at the expense of a mis-execution, so it's essentially a free shot.

i am concerned about scum interference if the shot is declared, but i am equally concerned that A50 would be shooting with no reads.

perhaps a solution to ease both of these concerns is possible?

my idea: have the town select a pool of three top suspects. we don't need to full voting procedure, just get a rough idea and lock three people in.

A50 can ISO those three slots overnight and pick his top choice. this is much easier to manage than a full reread of the game. it also prevents him from shooting an obvtown mistakenly as town (or even brazenly as scum). this also creates enough doubt on who the vig-target is that scum may not know to waste an interfering role on A50.

if all three choices are town, scum can safely let the shot go through, but we should back the collective wisdom of the town and assume we nominate at least one scum, which leaves scum in a tricky situation laced with doubt.

and as i said before, worst case scenario, even if we lose a town player amongst (whether because of town inaccuracy/A50 being scum/scum redirecting), we still end up in an 11p game tomorrow instead of 12p. same amount of mis-executions.

thoughts?

[Hoopla]

Or just elim A50

we can do that tomorrow if need be.

there's a chance we learn more about his slot via N1's results, so we may not need to guess on his slot today.

and we're not burning a mis-execution by taking the chance.

[Hoopla]

i'm here and checking in.

a little behind but will properly catch up soon.

a quick comment: i think i now dislike my idea of leashing the A50 vig and prefer just actually eliminating him today. i found the worst's posts about his behaviour convincing.

my gut says the gypyx wagon is a red herring, who needs a stern lecturing to cut that nonsense out. it's suicidal as either alignment, and if he's town, i want him to pull his head in.

[Hoopla]

VOTE: a50
i repeat: they are both scum

i'm persuaded by the argument that gypyx-scum would have at least ~some~ level of guilt preventing him from self-destructing in such spectacular fashion. you're directly letting down two teammates who you've been in close quarters with - whereas, as town, you don't really know who you're letting down, so it's easier to not care.

i would bet good money on gypyx being town.

having said all that, i used similar logic to falsely clear him (in mini 2146) based on an action i didn't think scum would make, so he may just be one of those special characters. this one does feel different, though.

[Hoopla]

i wouldn't be surprised about his alignment either way. my call on him being townie is mostly gut. consider your points heard & noted.

i think we probably have a juicier fish on the line in A50, anyhow.

[Hoopla]

is it bad i think #1120 is towny for a generally noncommittal read on a slot who everyone seems to have a committal read on ....... ?

not bad at all.

in fact, i think you should be commended for your brave and insightful observation.

[Hoopla]

A50 first, then gypyx tomorrow...

well, unless he dazzles us with another gambit to save his skin.

i'm expecting big things!

[Hoopla]

we're going to look back on this moment with fondness once we realise we've been to-ing and fro-ing between two scum this whole time.

[Hoopla]

congrats to the town heros for restoring peace in this game thread.

and shame on scum for their dastardly ways. crime never pays, believe me.

[Hoopla]

okay so next time this setup is played all 3 scum need to pick the same number

this would be a power move for sure.

i see no downside.

[Hoopla]

yeah, exactly - it's very much an intent thing tho.
was it explicitly intentional in this setup? hard to say because it's modified over the years and the current iteration is published anonymously. i kinda doubt they thought about multitasking when they designed it tbh.

i designed the original PYP X/Y setup, and don't recall whether or not multitasking was standard. quite likely it wasn't. the design ethos at the time for opens/normals tended to favour mafia performing single actions - i believe one of the theories being that interactions with trackers/watchers/RB's etc were more elegant when each player can only go one place, or perhaps that interesting situations arise when scum have to forgo a PR use to make a kill... but don't quote me on that. it could well have just been a holdover from the oldschool flavour argument of "people can't be two places at once!"

i'm sure PYP mods have probably used a mixed approach over time re: multitasking, but i think overall this setup would benefit from the change. it helps minimise some edge case scenarios where scum fall behind early and simply can't catch up without confirming a bunch of townies along the way. probably a wise adjustment given where town winrates are at these days.

[Hoopla]

To boot, the sheer level of information available to town creates a huge potential for townsiding swing, depending on (to be brutal) the degree of skill of the scum players. There also isn't a whole lot of scope for manipulation which is, though common for an open setup, sad to see in such a spicy setup. (The greatest example of this I can think of is TheWizard's running of the other PYPX/Y setup; a couple of years ago, a scum 1s watcher claimed gunsmith and in the right era that could have swung endgame. Now we just wait until everyone is cleared and policy it )

after all this time, it may be taboo to change the role pairings of such an established setup, but i think it's become obvious that fruit vendor and possibly 1-shot PGO are wasted slots. you could throw something fresh in there to make those pregame choices more of a dilemma,

and

also as a way to introduce more scum manipulation into the fray.

even without throwing in new roles to those slots, something like a Godfather 1-Shot PGO and a Miller Neighbouriser would be a cheeky way to balance those two role pairings, while also nerfing the cop somewhat.

2042 sounded unintentionally snarky, sorry

i didn't detect snark;)