Sammy, what do you think would be better advice to give planeswalkers? I don't think "charge day 1" is correct, because that in expectation decreases the amount of useful things a planeswalker can do before they die. Perhaps a mixed strategy would be better, but in a game this size (with about 1.1 expected planeswalker), I don't think that matters (plus I expect many people either didn't see my post before day 1, or just went with their own ideas).
I admit it would have been useful to mention that desperadoes activating for days 1 and 2 might run into this issue. But I can't remember to cover everything. *shrug* It's not like I'm the sole person responsible for giving strategy advice; the thread was, as named, meant to cause discussion of people's opinions.
I will also argue that if I were mafia, I would *not* intentionally try for a no-kill day. As I've said, making kills over IAP is harder, and removing leeway is risky, especially when waiting a day replaces a weekday (generally better for making kills) with a weekend day (generally worse for making kills). Plus, waiting a day means that town can be more coordinated beforehand (e.g. items transferred, hymnal rituals performed, etc). Also, from an OOG perspective: being mafia stresses me out, and I want to minimize the game-time I spend as mafia; no-kill days are less fun for game, and I care enough about the game meta to not do annoying things unless it's clearly better. BUT MOST IMPORTANTLY, it's really really bad if mafia/SK miscoordinate and have a 1-kill day (particularly if the 1 kill was by the serial killer; SK not killing just makes it very unlikely they'll live to make a second kill, but that's already unlikely). It would be particularly bad if I were mafia / SK and SK / mafia made a kill while I was at Squares; then I can't make a decent kill after that, and even if I'm mafia and my fellow mafia isn't at Squares, they'd be making a kill when 5-6 people could solidly alibi each other, which is not good.
Going off of this, here are my assessments of people's mafia / SK likelihood, as a result of day 1, conditioned on mafia/SK NOT having hit the other: Dylanhen - Less likely. Dylan had a good excuse to be on campus day 1, which he won't have day 2. I also think Dylan would try hard to make sure a kill was made. Sammy - Less likely. Although Sammy is often reluctant to make kills himself, I think he would not want to essentially cause a delay to the start of game, because he is leaving on Thursday. Me - Less likely. See argument above. Lucy - Slightly less likely. I think Lucy is likely to try to make kills happen. She did text the chat saying she was going to kill someone at Squares, but this seems roughly 50x more likely to be a joke than a mistext. m2w - Slightly less likely. I think m2w is likely to try to make kills happen. Julian - Very slightly less likely. I think he would try hard to make kills happen, but idk. Pravinas - - Very slightly less likely. I think she would try hard to make kills happen, but idk, she has been somewhat less active than I would expect. Rusch - Neutral. I vaguely feel like he's likely to push for kills, but I have no data on him being mafia / SK. Margarita - Neutral. Margarita is an experienced player, but I have no data on her being mafia / SK, I think? And I feel like Margarita isn't a fan of making kills. brunnerj - Neutral. I think brunnerj is likely to try to make kills happen, but he didn't have many good opportunities on day 1, and he may have expected to have an opportunity before Squares, which I ruined by also going to Squares early. Chiller - Slightly more likely. Less experienced of a mafia than average. Natalyat - Slightly more likely. Less experienced of a mafia than average. Guilld - Slightly more likely. New player. Jason - More likely, because he was roleblocked. (I had pointed out to Sammy day 1 that Jason was sketchier for not having any good opportunity to kill yesterday, but this sketchiness is a subset of the sketchiness of having been roleblocked.)
I think that if mafia or SK hit the other, it's very reasonable for the other to not make a kill. If mafia hit the SK, then waiting until day 2 for kills to happen basically gains a kill for mafia. If SK hit mafia, then waiting until day 2 somewhat improves the chances of SK living to use their kill day 4. Conditioned on mafia/SK having hit the other: dylanhen - sketchier; good kill target, good opportunity to make kills Lucy - sketchier; good kill target, fairly likely to make kills Julian - sketchier; good kill target, fairly likely to make kills Sammy - slightly sketchier; what mafia wouldn't want to kill Sammy day 1? m2w - slightly sketchier; good kill target, idk how kill opportunities go (it really depends on the UROP) me - neutral; what mafia wouldn't want to kill me day 1? On the other hand, mildly annoying to kill. chiller - neutral; good kill target, but less experienced making kills ritagram - very slightly less sketchy; mildly annoying to kill Rusch - very slightly less sketchy; mildly annoying to kill pravinas - very slightly less sketchy; mildly annoying to kill natalyat - slightly less sketchy; mildly annoying to kill, less experienced making kills. Guilld - moderately less sketchy; no one likes killing new players on day 1, not experienced at making kills. Josh - moderately less sketchy; mildly annoying to kill, few opportunities to make kills Jason - moderately less sketchy; very little opportunity to kill him, no ability to make kills
(Oops, I just realized I flipped the order of these lists between cases.)
I think that the first list has a greater impact than the second list, because in the second list, the sketchiness update only applies if they were the target or the killer; there is one mafia who is neither. Whereas in the first list, my comments apply to both mafia and the SK. I think "kill plans fell through" is a priori more likely than "mafia or SK hit the other," but the latter better explains why there are 0 kills rather than 1. So, I think my overall sketchiness ordering after day 1 is: Jason - slightly sketchy chiller - very slightly sketchy Julian - very slightly sketchy Lucy - neutral Natalyat - neutral Dylanhen - neutral Sammyluo - neutral m2w - neutral ritagram - neutral rusch - neutral Guilld - very slightly unsketchy pravinas - very slightly unsketchy me - very slightly unsketchy brunnerj - very slightly unsketchy
Note that the combined sketchiness calculation is very weak in magnitude! This is because the first and second cases have their sketchiness inversely related, for the most part. If we see evidence that one case is true over the other (e.g. the presence or absence of an SK kill), then we should use that case. However, note that the presence of a SK kill could be faked by mafia using scheme. Mafia just prefer to scheme near the end of the game, rather than near the beginning.
|