Ok. So I have read a bit to ground myself, but I still have the majority of the game to read. I am on page like 20 and some random assortment of recent pages. I know, I know, you all are like "What the duck, Mister Turkey!? You spent all that time telling us to slow our roll and you only read
twenty
pages!?
This is the worst trade of all time!
"
My answer to that is that is half right. I have spent the majority of my time focusing on creating my (patent pending) activity flow charts. It is a very simple premise and it has been pretty reliable in the two games I have used it. Iconeum can attest to its success in our recent game together.
Basically, I make a chart that gives me a graphical representation of a player's activity in the game. I use it to see if a player's activity has trended towards lurking. I also use it to see if their activity corresponds to some other stimulus like posting more if they are currently being pressured. Why? I believe that universally, players do not like playing as mafia. I think the psychological reasoning stems from having the "puzzle" of the game solved for you already. And stress. Lots and lots of stress. Humans do not like stress. Why do you think procrastination is such an issue? Playing as mafia is stressful.
Either way, I think people
in general
dislike playing mafia which in turn causes people to be less active in games.
I know that some players like to play mafia, or prefer to play mafia, and those players will defeat this tool. I acknowledge that. I have started a little folder of players who are widely known to like playing mafia, or players that I have run into that I know like to play mafia, so that I know to adjust the weight I gave to this tool when examining their play. None of the players I currently have on my list are in this game.
I know that some players are not active players. My response to that is this. Those players may not have a positive trendline, but they will not have a hugely negative trendline either.
I also know that sometimes people get busy and just cannot be as active as they once were in the game. That is a given that there will be some false positives. But, that's just the rule of the game. Moreover, those types of activity inhibitors are usually announced by the players in the game. I find scum are less likely to publicly announce those types of inhibitors, for whatever reason.
Now, I know this tool isn't the end all, be all. But it has thus far pointed me in the right direction. It makes me a more efficient scumhunter by focusing my energy on players more likely to be scum. That's where I think its usefulness lies.
Now onto the analysis. I examined the activity flow of every slot that is currently alive, including replacements. I did everyone for full transparency. So you can better understand, I recorded the number of posts a player made from the start of the game (or their replacement into the game) until September 3 when I joined. I do not go beyond the date I joined because that may skew the data if players knew my schtick.The only slots with negative trendlines are {Billy/LUV, Datisi, Almost50}. Since I know I am town and the setup tells us I am town and from what I gathered Almost apparently ragequit, I am going to say that Almost's trendline is more due to emotions rather than his alignment. More on Billy/LUV and Datisi in a moment.
The players with the most positive trendlines are Tchill13 and Klick. This makes sense to me. I feel like out of the players recently talking, Tchill13 and Klick are probably the towniest. I have some reservations about Klick that I will get into later once I more thoroughly read, but right now, I am going to safely consider them more than likely town.
Iconeum and Luca Blight barely have positive trendlines. Since Iconeum is a claimed PR and my role's existence would be almost completely redundant without another PR, Iconeum is town and I will not go into further wasting effort trying to examine him. As I said earlier, I think he is town for setup reasons and he can have the win if he is scum (CFJ can have my scorn in that event, as well).
Back to Billy/LUV and Datisi. As I said, these two slots have negative trendlines. I examined why the trendline was negative and both players have a particular pattern. They posted a lot early in their time playing and have not (recently) matched that volume of posting.
- LUV posted a lot after he replaced in (26 posts in one day), but hasn't posted more than 8 times a single day after that. He has the most 0 post days of anyone in the game as well. With the majority of his 0-days coming immediately after his large post day. Caveat: The "0 posts day" metric is a useful metric as well because that can help show a low post count player is not very active despite having a neutral or positive activity flow by virtue of posting
slightly
more later in the game.
- Billy did not follow this pattern, but he did have a negative trendline. When two players in the same slot follow the same activity flow, that is generally an indication that slot is at the very least Not Fun. What would make a VT slot Not Fun? It being mafia (see reasons above why scum is Not Fun).
- Datisi's posted a lot during a 4 day period (Aug 20, 21, 22, 23) wherein she posted 26, 33, 21, and 34 posts respectively. Since that time, the highest post count that Datisi has reached was 12 posts on September 3. I looked into why Datisi would be more active specifically during that 4 day weekend because that is slightly different than LUV's replacement into the game and subsequent catch-up barrage that artificially inflated his post count. During that 4-day stretch, Datisi was being highly pressured. She was the leading wagon for a good portion of it and her wagon maintained presence for almost the entirety of that time period. Once that wagon dissipated, so did Datisi's activity. That is scummy to me because it looks like a scum player doing
just
enough to tread water so to speak.
For these reasons, I believe this should be our kill order {LUV, Datisi, Luca Blight, Klick, Tchill13} with the scum team more than likely being LUV/Datisi.
VOTE: LUV
