Newbie 1934: Tundra (Game Over)

[Flavor Leaf]

I like hammering

[Flavor Leaf]

And it was funny

[teacher]

Because you pushed the theory hard, and that counter-evidence was available to you at the time, so knowing why you discounted it reveals your thought process.

Again, answer 598.

Again, case me, or stop complaining about the lack of 1v1

Again, tell me who you think my partner is (because all of the candidates youve mentioned have the same voting issue, so you need to answer that question you keep avoiding regardless).

PEDIT: Or you wanted some credit for the hammer on what was an inevitable lynch. But I'll give you credit for answwering that.

[teacher]

Seriously, I cant wait to hear about why I am a better D1 lynch that Raya. If the answer is that Raya has been run up multiple times by many slots, and so cant be scum, you would still need to explain away Baezu. (obviously you also havent been willing to join this wagon, but Im assuming town!you for purposes of casing me).

Baezu has been unwilling to touch this wagon, and indeed joined the then-available counterwagon and flaked for days.

How is that more scummy than a slot you cant case other than by gamestate (which has changed).... Im waiting.

[Eggs]

I'm tying Raya to FL in my head. He's done more than anyone else to distract from this lynch, from the early day 'town who got run up' to the current bickering and teacher irrelephant wagons.

[teacher]

Well, I no longer think that it’s an all SE team, so why should I defend it?

Because you pushed the theory hard, and that counter-evidence was available to you at the time, so knowing why you discounted it reveals your thought process.

Also because the question was asked in real time, and you ignored it multiple times. 605, 616, 626, 718. Simply ignoring bad logic and then saying its not relevant because youve changed your mind doesn’t help your credibility.

And its not the first time you’ve done this. In 572 you refused to justify a townread of Eggs because you were “no longer there.” But you were there as of 463 and thought Eggs couldn’t be a partner with me (your strongest scumread) as of 503.

Bottom line, claiming fluidity and changing reads is fair, but you should be able to defend thoughts you voice even if they have been overtaken by events.

[teacher]

OK, so I've read back through. Here's what I will say:

First, I feel like the last 300 posts has just split up town. I attribute alot of that to Flavor's push on both Elephant and me -- it was gamechanging, a very different but logical take that forced many slots to react. It kicked Elephant out of what had been almost a mayor slot of leading the game, and forced many slots to reexamine their previous townleans. I also shared, at an emotional level, Elephant's reaction. It was a perplexing push to counter because it was so different than my take on the game (my readslist was essentially inverted to his) while also making sense on a surface level.

Second, I feel decently confident that Elephant is town. I have followed his play and his gambits all game, and think they are town advancing, More to the point, however, I have shared his mindset and vibes as well. We are approaching the game in a similar way, with similar reads and reactions. I know that reads does not equal alignment, but I do think reactions do.

Third, that does not mean I think Flavor is scum. I cant read him, and do not want him lynched at all. I can certainly see scum motivation for this play -- breaking up town, casting suspicion on SEs who are generally decent town players. But I can also see a hell of a lot of town motivation -- a logical read of the gamestate that generates quite a bit of informative reactions. I really do want Flavor to respond to Menalque's questions on his approach that I expanded upon in 722

Fourth, as the first point suggests, I think the main mission for the rest of the day has to be rebuilding a townblock. I would LOVE if people would get out of their scum!tunnels and started to try to actively find, bond, and form consensus with those who they think are town, and questioning others (even if not their desired lynch) for more information to analyze after flips. Here, for me, I town Menalque, Elephant, Eggs, and Cinnamon. I lean Voted that way, and am null on Flavor. I want Pine to eat rope, and lean scum on Baezu as well. I am still waiting for Baezu to answer 591. 592, and 595

Finally, I find the lack of attention given to Bob/Baezu's slot AND THE WAGONS perplexing. I just addressed Bob/Baezu. The wagon point is that almost all wagons this game have consisted of the same players:

• Navneet 65: Voted. Irrelephant11, Menalque, Eggs
  
• Voted 300: Irrelephant11, Menalque, Cinnamon, teacher [Eggs was here before me, but switched to me on what he said was a reaction test)
  
• Raya 407 and 604: Voted, Cinnamon, Irrelephant, and (Menalque/teacher, depending on when)
  
• Current raya (no VC): Irrelephant, Voted, Eggs, Cinnamon

I like lynching Raya, because I think a green flip clears Voted pretty hardcore given that the same wagon was on him, and definitely confirms scum in [Cinnamon/Elephant/me]. That is incredibly helpful to me because Im town on both of them right now, and would be forced to reexamine. It would also vault Flavor way up in my reads, both for pushing so hard and also for his theory being right. On the other hand, a red flip I think essentially clears all of Voted, Cinnamon, Irrelephant, and me, creating an incredibly strong townblock to win the game.

Intent -appr 24 hours

Pine, I appreciate the sub in and the read through effort. You cant be allowed to make it to D3, and I think you are the best lynch for the day for the reasons given. If Im off on you, Id love if youd continue the readthrough. But as I said on the last page, I think you should claim at some point tomorrow, and Im now making that point hard with the intent.

Pedit: It was not intentional, in the sense that I did not notice it. It is a genuine reaction to the wagon information, however, so Voted can join Menalque in my will lose at Lylo if scum pile.

Im signing off but I dont want these two to be buried in catchup.

Pine there is intent

, Baezu please answer questions, and I welcome any other reactions to the wall.

[Menalque]

Here’s where my head’s at: I think it’s either Pine/FL or Teacher/irrelephant.

I’m leaning towards the latter because I’m more persuaded all in all by FL, and so that’s where I’m putting my vote.

VOTE: teacher

Incidentally, I think FL is either the best or the worst mafia player I’ve ever been in a game with, depending on how this turns out.

[Eggs]

...a flaking Newbie is NAI.

Snipped quote obviously.

Ran the numbers so I can call bs on this.

Unfortunately the numbers don't agree. In games 1901 to 1930:

0.60 replacements per newbie scum
0.56 replacements per newbie town
0.40 replacements per SE scum (but 6/15 is a small sample)
0.14 replacements per SE town

NAI seems about right. If you go back only to 1903 not 1901 the newbie numbers flip. Well within variance.

There goes 80% of my case.

UNVOTE:

[Eggs]

This needs pushing

VOTE: Baezu

[Eggs]

Ugh I'm not liking this at all. Very conscious time is pressing but I'm not really satisfied with the pine lynch.

I think he could pretty much only be scum with either FL or Bob.

Cinnamon got pushed by Raya and is a (qualified) town read from pine
Voted and teacher are solidly on this lynch
Irrelephant is scum by pine and strong town by Raya.

And in either case I'm happy with the Bob lynch. Pine+FL is a worry true, but that's a pretty specific case.

[Eggs]

Sorry, Baezu. I mean you not Bob. Too much hurrying.

[Eggs]

Missed menalque there but I'm town-reading him.

[Pine]

Egg you’re going from replaced lurker newbie to replaced lurker newbie. Not saying you’re wrong, Baezu is in my unsorted list, but it’s lazy.

At work. Will post more at lunch

[teacher]

As someone who played in at least a quarter of the games involved, 758 seems off. Like I don’t doubt that it’s an accurate report of statistics on a cold read, but I think it probably suffers from a codin/category error. I assume it counts replacements like (Brass->flavor) in this game, since some mods (Incl. Plot, who nodded prob 35%+) list anyone who got a role pm in their first post, even if they never logged in/confirmed/responded/posted in thread.

I have not done a sit down study, but my lived experience tells me that players who post in thread, then replaceout when getting scumread, are more likely scum. It’s not just the flaking but the flaking when already under pressure.

Also, while I will definitely compromise on Baezu, it seems like that’s a contradiction with the reason to get of Pine/raya/Nav - the case on them is flaky non game advancing - is the same as what you’re getting off, other than the board has been ignoring it.

[Eggs]

Absolutely true about category errors. However I'm struggling to find such errors that won't be flat across newb!scum and newb!town.

The fact that you come back with anecdote doesn't help. I'm kinda ignoring the second paragraph unless you can nail down the situations where town is more likely to flake to make a reasonable hypothesis that there are times when scum are more likely to flake.

Also nav wasn't really scumread. He was kinda consensus town at that point. Also Nav has tended to flake more on his town games, came across a couple in my trawl but he didn't flake as scum unless I missed one.

Also if Baezu is the same argument as Navneet and has gone under the radar isn't that an indication that he is more likely to be being kept out of the firing line?

Also I didn't like Bob's tone and pine pinged me town on voted's thing.

[teacher]

To fix the error, or at least modulate it, only count replacement of slots that posted in the game. I would do it but in class all day. I do really appreciate the counting effort.

I hear where your coming from on Bae and being allowed to stay under the line. I’m willing to compromise there I think Pine is more game advancing since more slots have defined associatives. But that’s a difference not a disagreement if that makes sense.

Nav was not consensus town, as you can tell from the wagon, but I think that’s also pretty irrelevant.

[teacher]

To fix the error, or at least modulate it, only count replacement of slots that posted in the game

For example, this gets rid of 4 town replacements in Newb 1909 alone, where 2 slots were twice filled with people who joined the queue then forgot the site exists. Since replacements like those are both common (yet to play a game without one) and 78% likely to be town, including them skews your data to make replacement seem less scummy than it is.

[Pine]

Really not liking that selective interpretation from teacher. If you cherry-pick enough, you can get data to say almost anything you want.

As a general rule, newbies flaking is NAI. The sharp variance in SE flaking that Egg found is far more interesting, but as he said there’s not a large enough sample size.

[teacher]

I’m not encouraging cherry picking. I’m encouraging coding the same dataset similarly - don’t count replacements who never post. Count replacements that do.

Why I’m actually talking about it so much is that I found eggs data shocking. I do mechanically scumread flakes. I have encouraged others to do so. I do it always, across several games. If Eggs data is sound, I’m wrong and need to change my approach. I want to know that - what data says. Until I trust the data, though, I’m standing by my experience (however anecdotal it may be).

Here your slot has two mechanical scumreads for me - the replacement, AND twice greeting the thread. That, plus the rayas awkward actual post, plus the difficulty in getting a lurker wagon across the line makes me feel pretty good about a redflip.

[Eggs]

Agreed on need to do better analysis. But also counting on a flat e.g. 30% on both town and scum for genuine life/forget site replace-outs won't affect the numbers. Since I'm not doing more work and nobody is proposing a time when town flake more than scum, data would suggest there is no time when scum flake significantly more than town.

[Eggs]

I'm with you on the greeting tells. That was part of my remaining 20%. But from pine? Is that still a newb scum tell?

[teacher]

I will code tonight after school because this would really change my approach to the game.

I also didn’t like Rayas substantive post.

[teacher]

But also counting on a flat e.g. 30% on both town and scum for genuine life/forget site replace-outs won't affect the numbers.

This is error. Flat discount doesn’t make sense when 78% of slots are town. Town replaceouts should be discounted more heavily than scum. But like I said I will do the actual coding and share dataset and results tonight.

[Eggs]

Nav was not consensus town, as you can tell from the wagon, but I think that’s also pretty irrelevant.

118120125