GreenDigitalWolf: But the decision is random, isn't it? There's not really a way to lean towards a direction between "he wants me to think that, so I'll do the opposite" and "but maybe he wants me to do the opposite, so I won't do it"
On the first Day, it is pretty dang random. We have to parse every post, see if the wording seems odd, and/or look for common tells. Obviously, we can end up falling into the WIFOM trap a lot. "That sounds really scummy, but would a true scum
really be that blatant about it?" or "Wow, that was a really pro-town thing to say, but is it just scum trying to hide in plain sight?". In the end, I just end up going for the person that gives me the biggest scum-vibe, and hope for the best, because the only thing worse than lynching a townie is to lynch
nobody.
After the first Day, we will hopefully have more information to use to back up our scum-hunting. We will hopefully have a previous Day's lynch bandwagon to study, as well as, most likely, a new Nightkill. Also, town power-roles (I think it is only Seer in this game) will have had a chance to take action on our behalf during the Night. From those data points, patterns may start to emerge in previous posts. Each Day that goes by will add to the circumstantial (or concrete, in the case of Seer) evidence against scum.
So, it is essentially a race against time, and there is obviously a lot of strategy, as well as luck, involved. Such as, when should the Seer role-claim and reveal his/her evidence? Every Night that goes by adds more opportunity to investigate suspected scum, but it results in more dead townies (and worst of all, put the Seer at risk of death by Nightkill). I think a lot of that decision would be based on how successful the investigation have been in either confirming town or pegging scum.
