SUPP 2017 MAFIA: COMPLETE

[Commander Shepard]

VOTE: Shoshin

#sleepDeprivedCommandersIsBad

[Espeonage]

Ooooooooooooooooooooo. I think Vax is a friendly neighbour.

1) Unlikely. With 2 cops a tracker neighbor and my memory is shit at midnight other invests.
2) It would have hit me and thus been created
3) if true Unless scum rolestopped me but why?
4) Why the fuck you fishin?

5) sleep sounds amazing.

If he was a friendly neighbour maria wouldn't be a cop now would she.

idk I thought I had a brain wave.

[Dunnstral]

Active: Keeping Track -- You are in a neighborhood with TwoInAMillion. You know that TwoInAMillion is "Bins". You may speak during the night. Once per Night Phase, you may name a target in the neighborhood. You will learn who your target visited that night. You must both agree on the target. If one of you dies, you will lose this power.

Took a closer look at this - There's no TwoInAMillion or Bins in this game? What is this role talking about

[Dunnstral]

Nevermind, I see it now, it was Pine

[Punreader]

Punreader is confirmed not group scum unless they are BP as I am relatively confident they shot me last night.

Well, I

am

BP, but my BP has the inverse condition of my vig. If I vigged you successfully, I'd have no BP against you; if I didn't vig you because you're ranked higher than me, I'd be BP against you but also not be making an attempt on your life with my visit (I visit regardless of whether I am ranked higher or lower than my target).

Just caved and checked my own decision, I used pun and Dunn. Probably because they rhymed.

For the attentive, this means there is a maximum of one pun in ManWithNoName/Commander Shepherd. This doesn't guarantee one is pun (they both

could

be town), but it guarantees both aren't.

I can explain this, it would simply be anti-town of me to do so.

My main thought right now is that I'm back to being suspicious of Fire because firstly, he's alive, and secondly, why investigate Dunn? That doesn't seem incredibly useful.

Correct. He is, with a high degree of certainty, pun. However, correct play here is to lynch him

tomorrow

, not today.

Fire is fucking scum. He fucked up a fake claim guilty on me to buy himself a day, and I guess to that end he succeeded. So this needs to end now.

Vote: Fire

If the choices are between lynching Fire Assassin and lynching Commander Shepherd, I am choosing to lynch Fire Assassin 100% of the time.

However, I must insist we let him live for one more night and simply lynch him tomorrow, as that is the optimal play.

[MariaR]

MariaR, do you know if you were roleblocked?

I don't know. All I know was I was auto bp'd

[the worst]

mfw Pun confirmed SK but we already knew that

whatever my read on CShep slot isn't changing rn

[DeasVail]

However, I must insist we let him live for one more night and simply lynch him tomorrow, as that is the optimal play.

I understand you may have reasons for this that I am not privy to, but as the most likely person, in my mind, to be scum, I will likely be pushing for his lynch today unless this is explained.

Also Shepard, does the PGO element of your role only work with kill attempts or other actions as well?

[Punreader]

However, I must insist we let him live for one more night and simply lynch him tomorrow, as that is the optimal play.

I understand you may have reasons for this that I am not privy to, but as the most likely person, in my mind, to be scum, I will likely be pushing for his lynch today unless this is explained.

I don't have reasons you're not privy to in regards to Fire Assassin. (I do for Commander Shepherd, but the less said there, the better.)

However, my reasoning for this, I stated yesterday.

Pun

, can you give me your thoughts on

Fire/Maria

?

I could see any combination of town-town, pun-pun, town-pun, or pun-town as possible. I can tell you on play I am not punreading MariaR, which would make it slightly more likely it's town-town or pun-town, but this is not locked in. Regardless of their alignment, however, correct play is to let both live and gather more results.

After all, if both are town then pun can't let both go unchecked.

I'm highly suspect of

Fire

at this point

Suspect him all you want.

He is objectively a terrible lynch today because we have two cop claims; if both

are

town (and yes I realize people doubt this but it is not impossible), then pun's only option is to deal with them during the night, lest they both produce more results.

I am telling you now that if we lynch Fire Assassin today and he flips town the entire punteam will be on that wagon and I will lynch exclusively from the players on it, for precisely this reason.

Now I admit, there is a significant chance Fire Assassin is just pun, but this changes nothing.

If he is pun fakeclaiming cop, he still has to produce results every day, or be lynched when he fails to do so.

So we wait.

We give him, at minimum, one day phase.

Correct play in a theme game such as this is not to lynch a cop claim and if the cop flips town lynch the other cop claim; correct play in a theme game is to not lynch either cop claim and let them investigate for a night, and then off of those investigation results, revisit and reevaluate every day on whether to lynch one of the cops and if so which one.

This is unchanged. Reevaluation has determined he is significantly more likely to be pun than yesterday, but he is not absolutely certain to be pun. And on the off chance he is town (after all, our track record in nailing down pun thusfar has not been stellar so who's to say this read is right when prior punreads have been wrong?), then this will sort itself out overnight when the pun are put up against two cop investigations the same night.

MariaR could not investigate last night.
MariaR

can

investigate tonight.
And if Fire Assassin fails to produce an investigative result tomorrow for any reason, we lynch him for it on the spot.

It's simply a matter of optimal play.

[Commander Shepard]

However, I must insist we let him live for one more night and simply lynch him tomorrow, as that is the optimal play.

I understand you may have reasons for this that I am not privy to, but as the most likely person, in my mind, to be scum, I will likely be pushing for his lynch today unless this is explained.

Also Shepard, does the PGO element of your role only work with kill attempts or other actions as well?

FoS DeasVeil

Why the fuck would I tell scum whether it did or not? That seems to be recipe for dumb

If I am then scum can’t roleblock then kill me they have to mislynch me
If I am not then scum know if they can eliminate me.

[Punreader]

I would like to note in terms of the extra names for data crunching:

Spoiler: Predicted Rankings

1. Chara:
9 10s
1 9
1+1 8s
3 6s
1 5
1 3
1 0

90 + 9 + 16 + 18 + 5 + 3 = 141
9 + 1 + 2 + 3 + 1 + 1 + 1 = 18
141 / 18 = 7.83

2. the worst:
1 10
3 9s
1 8
5 7s
2+1 6s
3 5s
1 2

10 + 27 + 8 + 40 + 18 + 15 + 2 = 120
1 + 3 + 1 + 5 + 3 + 3 + 1 = 17
120 / 17 = 7.06

3. Taly:
7 10s
2 9s
1 8
1 5
1 4
1+1 3s
1 2
1 1
2 0s

70 + 18 + 8 + 5 + 4 + 6 + 2 + 1 = 114
7 + 2 + 1 + 1 + 1 + 2 + 1 + 1 + 2 = 18
114 / 18 = 6.33

4. TehBrawlGuy:
4 10s
3 9s
2 8s
1 6
1 5
1 4
1 3
2+1 2s
2 0s

40 + 27 + 16 + 6 + 5 + 4 + 3 + 6 = 107
4 + 3 + 2 + 1 + 1 + 1 + 1 + 3 + 2 = 18
107 / 18 = 5.94 (5.944)

5. Nahdia:
3 10s
1 9
1 8
1 7
2+1 6s
3 5s
3 4s
1 3
2 0s

30 + 9 + 8 + 7 + 18 + 15 + 12 + 3 = 102
3 + 1 + 1 + 1 + 3 + 3 + 3 + 1 + 2 = 18
102 / 18 = 5.67

6. Dunnstral:
3 10s
2 9s
1+1 7s
2 6s
2 5s
1.5 4s
4 2s

30 + 18 + 14 + 12 + 10 + 6 + 8 = 98
3 + 2 + 3 + 2 + 2 + 1.5 + 4 = 17.5
98 / 17.5 = 5.60

7. DeasVail:
1 10
1 8
5 6s
5 5s
3+1 4s
1 0

10 + 8 + 30 + 25 + 16 = 89
1 + 1 + 5 + 5 + 4 + 1 = 17
89 / 17 = 5.24 (5.244)

8. Espeonage:
3+1 10s
1 9
1 8
1 7
2 6s
2 4s
1 3
1 2
4 0s

40 + 9 + 8 + 7 + 12 + 8 + 3 + 2 = 89
4 + 1 + 1 + 1 + 2 + 2 + 1 + 1 + 4 = 17
89 / 17 = 5.24 (5.235)

9. Shoshin:
1 10
3 8s
2 7s
3 6s
3 4s
1 3
2 3s
1+1 0s

10 + 24 + 14 + 18 + 12 + 3 + 6 = 87
1 + 3 + 2 + 3 + 3 + 1 + 2 + 2 = 17
87 / 17 = 5.12

10. MariaR:
2 10s
1 9
2 8s
4 6s
1 5
1+1 4s
1 3
1 1
3 0s

20 + 9 + 16 + 24 + 5 + 8 + 3 + 1 = 86
2 + 1 + 2 + 4 + 1 + 2 + 1 + 1 + 3 = 17
86 / 17 = 5.06

11. Pine:
4+1 10s
1 5
5.5 4s
2 2s
1 1
3 0s

50 + 5 + 22 + 4 + 1 = 82
5 + 1 + 5.5 + 2 + 1 + 3 = 17.5
82 / 17.5 = 4.69 (4.685)

12. Punreader:
6 10s
1 8
2 7s
2 5s
2 4s
1 3
3+1 * 2 0s (8 0s)
60 + 8 + 14 + 10 + 8 + 3 = 103
6 + 1 + 2 + 2 + 2 + 1 + 8 = 22
103 / 22 = 4.68

13. Screenplay:
1 10
3 7s
4 6s
2+1 5s
1 3
1 2
1 1
3 0s

10 + 21 + 24 + 15 + 3 + 2 + 1 = 76
1 + 3 + 4 + 3 + 1 + 1 + 1 + 3 = 17
76 / 17 = 4.47

14. Fire Assassin:
2 10s
2 9s
1 8
3 5s
3 4s
1 3
1 2
2 1s
2+1 0s

20 + 18 + 8 + 15 + 12 + 3 + 2 + 2 = 80
2 + 2 + 1 + 3 + 3 + 1 + 1 + 2 + 3 = 18
80 / 18 = 4.44

15. Vaxkiller:
2 10s
3 6s
2+1 5s
3 4s
1 1
5 0s

20 + 18 + 15 + 12 + 1 = 66
2 + 3 + 3 + 3 + 1 + 5 = 17
66 / 17 = 3.88

16. PenguinPower:
1 10
1+1 9s
2 8s
1 7
1 6
1 5
.5 4s
1 2
1 1
7 0s

10 + 18 + 16 + 7 + 6 + 5 + 2 + 2 + 1 = 67
1 + 2 + 2 + 1 + 1 + 1 + .5 + 1 + 1 + 7 = 17.5
67 / 17.5 = 3.83

17. NicoRobin:
1+1 10s
2 4s
1 3
1 2
3 1s
8 0s

20 + 8 + 3 + 2 + 3 = 36
2 + 2 + 1 + 1 + 3 + 8 = 17
36 / 17 = 2.12

18. Lovebird:
3 5s
3 3s
1 2
5+1 1s
4 0s

15 + 9 + 2 + 6 = 32
3 + 3 + 1 + 6 + 4 = 17
32 / 17 = 1.88

NOTE: As I killed Pine, my score must be higher than this calculation would indicate.
I am not sure how much higher I am meant to be, but it is at least one position since Pine is confirmed as 12 and I shot him. (I'm not sure where I am precisely. I know my role doubles the effects of 0s, and I know Nico's role doubles the effect of scores she gives and she gave me a 0, so those

should

have stacked.)

PREDICTED DATA RESULTS:
18. Lovebird
1.88
0 10s; 4 0s

17. NicoRobin
2.12
1+1 10s; 8 0s

16. PenguinPower
3.83 (EXACTLY ACCURATE)
1 10; 7 0s

11. Pine (DISCREPENCY DUE TO ME BEING RANKED HIGHER)
4.69 (4.685)
4+1 10s; 3 0s

4. TehBrawlGuy
5.94 (5.944) (EXACTLY ACCURATE)
4 10s; 2 0s

3. Taly
6.33 (EXACTLY ACCURATE)
7 10s; 2 0s

ACTUAL DATA RESULTS:
18. Lovebird
1.78

(-.10 difference between prediction)

0 10s; 4 0s

17. NicoRobin
2.06

(-.06 difference between prediction)

1+1 10s; 8 0s

16. PenguinPower
3.83
1 10; 7 0s

12. Pine
4.67

(-.015 difference between prediction)

4+1 10s; 3 0s

4. TehBrawlGuy
5.94
4 10s; 2 0s

3. Taly
6.33
7 10s; 2 0s

Since we're seeing some data points that are precisely as indicated, it is a safe assumption those ratings are 100% accurate.
Notably, the differences in the prediction versus the reality are likely to lie in Pine's missing scores and in the discrepancy in Fire Assassin's rankings. (He submitted three players as 4, and it is unclear which of them is a 5.)

So from this, we can fairly assume, once more, that rankings were as indicated, with the possible exception of the players' ranking on me. (Though, if I had to guess a reason, it may simply be that Pine changed his mind on me and ranked me above 0 and above 1. 103 + [somesmallnumberrangingfrom2to10] / 20 = some number greater than 5.15. This would place me above Pine, but below Nahdia given the maximum rating for me would be 5.65, thus explaining a successful shot on Pine, his ranking, and an unsuccessful shot on Commander Shepherd.)

[Punreader]

So from this, we can fairly assume, once more, that rankings were as indicated, with the possible exception of the players' ranking on me.

This is useful, because if you can trust rankings to be accurate, then you can also trust the rankings to be done with an alignment associated with the picks.

If ranks are truthful (and we know they are), then players have accountability for what their rankings are. Certain decisions in their rankings then become more, or less, suspect. This will be most apparent once we get a pun flip; with even so much as a single pun dead, how players ranked that pun player will reveal a lot about everyones' alignments.

However, even prior to a pun flip, we can use rankings as a punhunting tool. If we have the spare time, it may be useful to compare stated D1 stances with the rankings; if there is an obvious disparity between the claimed rankings and the stated reads, that is automatically suspect, unless they can provide a

very

compelling reason for the dissonance.

[Punreader]

(For those who wish to see a quick reference, 824 is the end of D1. All posts with a number lower than 824, then, are part of what formed a player's D1 stances.)

[Punreader]

No surprises here, but my first iso was of Chara, and what I saw in Chara's posting more or less lines up with,

TehBrawlGuy (10)
Taly (10)
Chara (10)
Nahdia (9)
Punreader (8)
the worst (7)
MariaR (6)
Shoshin (6)
DeasVail (5)
Dunnstral (5)
Espeonage (4)
Fire Assassin (4)
Lovebird (3)
PenguinPower (2)
NicoRobin (1)
Pine (0)
Screenplay (0)
Vaxkiller (0)

These are fairly true to stances, so there is no disparity to be had. (But we all knew that anyway.)

[Commander Shepard]

So tldr scum lied as I said?

Like what are you saying with all this?

[Punreader]

The following people are some form of townread for me. No indication is made as to the strength of townread:
Dunnstral
Nahdia

Now here's some dissonance I have to ask about.

Taly (10)
Punreader (10)
DeasVail (10)
the worst (9)
Shoshin (8)
Screenplay (7)
Chara (6)
Vaxkiller (6)
Nahdia (5)
Fire Assassin (5)
Dunnstral (4)
Pine (4)
TehBrawlGuy (3)
Espeonage (2)
Lovebird (1)
MariaR (0)
PenguinPower (0)
NicoRobin (0)

So Taly, myself, the worst, Shoshin, Screenplay, and Chara were all townreads. That makes sense. I've removed those names from the original 652, because what I really want to focus on here is the remainder. Dunnstral is listed as a townread, but is below both Fire Assassin and Vaxkiller, who are not listed as townreads. Nahdia is listed as a townread, but is below Vaxkiller who is not listed as a townread and is equal to Fire Assassin who is not listed as a townread.

The rest seem to loosely match up, but those rankings in particular seem jarring.

Care explaining, DeasVail?

[Punreader]

So tldr scum lied as I said?

No, the opposite, pun told the truth. Meaning we can punhunt off of rankings versus reads (town shouldn't have dissonance between the two; pun aren't

guaranteed

to have dissonance between the two, but it is quite likely they will), precisely what I am doing right now with a simple fact checking process.

[MariaR]

Do the number of 10s and 0 match up with what's been said so far? Perhaps we can catch a liar

[ManWithNoName]

Why were people thinking the scum might have lied about their rankings anyway?

[Punreader]

Dunnstral may be conftown, but it never hurts to do the fact-check if for no other reason to see if it holds merit. And,

Dunnstral (10)
Fire Assassin (10)
Chara (10)
Taly (9)
Espeonage (8)
Punreader (7)
MariaR (6)
Screenplay (6)
Nahdia (5)
the worst (5)
DeasVail (4)
Pine (4)
Shoshin (3)
TehBrawlGuy (2)
PenguinPower (1)
Lovebird (0)
Vaxkiller (0)
NicoRobin (0)

Dunnstral didn't give much on D1, but these do loosely match up with what he said, though MariaR is a 6 instead of the promised 4. (Then again, Pine was "a little scummy", got a 4, and Shoshin was suspect, got a 3, there's hints of not being impressed with TehBrawlGuy, he promised to rank PenguinPower low, etc., so ranking MariaR higher than that is not too surprising.)

So overall, fairly neat match between reads and rankings; no real visible dissonance.

[Punreader]

Do the number of 10s and 0 match up with what's been said so far? Perhaps we can catch a liar

I checked.

PREDICTED DATA RESULTS:
18. Lovebird
0 10s; 4 0s

17. NicoRobin
1+1 10s; 8 0s

16. PenguinPower
1 10; 7 0s

11. Pine
4+1 10s; 3 0s

4. TehBrawlGuy
4 10s; 2 0s

3. Taly
7 10s; 2 0s

ACTUAL DATA RESULTS:
18. Lovebird
0 10s; 4 0s

17. NicoRobin
1+1 10s; 8 0s

16. PenguinPower
1 10; 7 0s

12. Pine
4+1 10s; 3 0s

4. TehBrawlGuy
4 10s; 2 0s

3. Taly
7 10s; 2 0s

The numbers match up.
As I said, evidence supports the pun having told the truth. Given that they told the truth, they are thus accountable for their rankings, which is why I am currently in the process of isoing players and reviewing what their stated reads were, and pulling up their rankings to compare the two.

[Errantparabola]

Votecount 3.1

Shoshin

(3): Punreader, the worst, Commander Shepard

Commander Shepard

(2): Fire Assassin, ManWithNoName

Fire Assassin

(1): Espeonage

Not Voting

(6): MariaR, Chara, DeasVail, Shoshin, Vaxkiller, Dunnstral

With 12 alive, it takes 7 to achieve a majority.
Deadline in (expired on 2018-07-21 13:08:00)

[Punreader]

So I realize he's conftown but I still need to ask,

Pine not feeling pressure to produce content right now also can be weak town.

This is the only difference between your given reads and,

Espeonage (10)
Dunnstral (10)
Vaxkiller (10)
TehBrawlGuy (9)
Taly (8)
the worst (7)
DeasVail (6)
PenguinPower (6)
MariaR (5)
Chara (5)
Punreader (4)
Nahdia (4)
Shoshin (3)
Screenplay (2)
Lovebird (1)
Fire Assassin (0)
Pine (0)
Nico Robin (0)

What changed your mind on Pine?

[Espeonage]

I didn't really have scum reads and lack of content got really negatively voted as a result.

That has since become hypocritical but w/e.

[Espeonage]

Why were people thinking the scum might have lied about their rankings anyway?

This tbh. Scum have to fabricate reads so why would they not just fabricate reads that make voting go well for them? If they are passing day play they pass vote scrutiny.