2016 Mafia on The Syndicate: A Statistical Review
Posted: Sun Mar 05, 2017 10:39 pm
A few times in recent games I have referenced my "spreadsheets" containing some kind of statistical information about players here. I haven't said much about what that actually is, or why I would have such a thing, but I am going to share it now. It is deliberate that I am doing this now, after the Sockys have concluded -- we talked about it among the hosts and I didn't want these numbers affecting voting or anything like that. There's one main reason for that:
These statistics do not reflect the skill level of the players included. Mafia is a complex game which is difficult to reduce into a number crunch. The purpose of this information is only that some might find it interesting to look at.
NOTE: I made up quite a lot of jargon throughout the process of compiling this data, so at first glance it might look like total gibberish to you. In this post I will explain everything as well as I can. If anyone has any questions though, please ask.
Statistics for individual GAMES played on The Syndicate in 2016
Statistics for individual PLAYERS who played games on The Syndicate in 2016
IMPORTANT: It was impossible to compile all of the statistics for every game. This means that some games are less complete in their data than others, typically because of difficult-to-follow sock accounts which have morphed, inconsistencies in vote/poll tracking, or various other issues. The following spoiler details which games feature which statistical information:
These limitations are reflected in the individual player statistics too. You'll see some players who played in a game but have zero recorded posts -- this is why. For games where I could not ascertain the information for one player, I didn't accumulate it for any player.
REPLACEMENTS -- It was extremely difficult to keep track of all of these, but I believe that I was able to do so for all of the games with relevant statistical information.
WHAT DOES THIS STUFF MEAN?
Voting HIT and MISS -- A HIT reflects a vote placed for a member of a mafia team only. Any votes placed for any independent players are classified as a MISS. I understand that some independents are anti-civilian and that it's good to vote for them, but this made my life much easier as the statistician.
BUS votes -- These refer to deliberate votes by mafia members against their teammates. If a mafia member votes for a mafia member of a different team, it is counted as a HIT.
Cycles alive -- A "cycle" refers to the combined period of a Day phase and a Night phase. To be alive for "half" of a cycle usually implies a lynch occurred during a day phase (but not always).
PC² -- Post Count Per Cycle Alive -- This is my preferred metric for measuring the activity level of players in each game and overall. The higher this number, the more posts players made during their periods alive in games. This also means the statistic is vulnerable to outliers for players who are killed very early.
PC comb -- Post count when in multiple alignments -- This is a count of posts made by players in games where they've experienced a change of alignment (either because of a game mechanic or a replacement).
V-hit (good) -- Number of HIT votes placed while good-aligned
V-total (good) -- Total number of votes placed while good-aligned
The same applies for V-hit (bad), V-total (bad), V-bus, V-hit (indy), and V-total (indy). The "%" column next to each alignment section indicates the rate of HIT votes within that alignment by that player.
Other deaths -- This refers to all deaths that cannot adequately be explained by the other terms I used (lynched, killed, survived). This could refer to a mechanical death, a modkill, or being killed in an endgame scenario.
~~~
There is likely more I could/should explain, but this is all I can think of right now. Please ask me any questions you have; this stuff makes sense to me because I spent months staring at it. I understand everyone else doesn't have that luxury.
REQUESTS
If anyone would like a reduced view of their own personal statistics, let me know. I can try to get that for you and translate it a bit into more readable language. If there are a lot of requests, please bear with me while I find the time.
IDEAS
The beauty of statistics like these is that there are a ton of ways that they can be worked into interesting concepts and information. If you think of something that could be worked out with these, or you have another idea for statistical exploration, please let me know.
These statistics do not reflect the skill level of the players included. Mafia is a complex game which is difficult to reduce into a number crunch. The purpose of this information is only that some might find it interesting to look at.
NOTE: I made up quite a lot of jargon throughout the process of compiling this data, so at first glance it might look like total gibberish to you. In this post I will explain everything as well as I can. If anyone has any questions though, please ask.
Statistics for individual GAMES played on The Syndicate in 2016
Statistics for individual PLAYERS who played games on The Syndicate in 2016
IMPORTANT: It was impossible to compile all of the statistics for every game. This means that some games are less complete in their data than others, typically because of difficult-to-follow sock accounts which have morphed, inconsistencies in vote/poll tracking, or various other issues. The following spoiler details which games feature which statistical information:
Spoiler: show
REPLACEMENTS -- It was extremely difficult to keep track of all of these, but I believe that I was able to do so for all of the games with relevant statistical information.
WHAT DOES THIS STUFF MEAN?
Voting HIT and MISS -- A HIT reflects a vote placed for a member of a mafia team only. Any votes placed for any independent players are classified as a MISS. I understand that some independents are anti-civilian and that it's good to vote for them, but this made my life much easier as the statistician.
BUS votes -- These refer to deliberate votes by mafia members against their teammates. If a mafia member votes for a mafia member of a different team, it is counted as a HIT.
Cycles alive -- A "cycle" refers to the combined period of a Day phase and a Night phase. To be alive for "half" of a cycle usually implies a lynch occurred during a day phase (but not always).
PC² -- Post Count Per Cycle Alive -- This is my preferred metric for measuring the activity level of players in each game and overall. The higher this number, the more posts players made during their periods alive in games. This also means the statistic is vulnerable to outliers for players who are killed very early.
PC comb -- Post count when in multiple alignments -- This is a count of posts made by players in games where they've experienced a change of alignment (either because of a game mechanic or a replacement).
V-hit (good) -- Number of HIT votes placed while good-aligned
V-total (good) -- Total number of votes placed while good-aligned
The same applies for V-hit (bad), V-total (bad), V-bus, V-hit (indy), and V-total (indy). The "%" column next to each alignment section indicates the rate of HIT votes within that alignment by that player.
Other deaths -- This refers to all deaths that cannot adequately be explained by the other terms I used (lynched, killed, survived). This could refer to a mechanical death, a modkill, or being killed in an endgame scenario.
~~~
There is likely more I could/should explain, but this is all I can think of right now. Please ask me any questions you have; this stuff makes sense to me because I spent months staring at it. I understand everyone else doesn't have that luxury.
REQUESTS
If anyone would like a reduced view of their own personal statistics, let me know. I can try to get that for you and translate it a bit into more readable language. If there are a lot of requests, please bear with me while I find the time.
IDEAS
The beauty of statistics like these is that there are a ton of ways that they can be worked into interesting concepts and information. If you think of something that could be worked out with these, or you have another idea for statistical exploration, please let me know.
