My focus has changed over the years and I would still like to push out a large document that includes the information I want with a pragmatic focus but I may never make that happen. So this is 'hopefully' a rough draft and I should mention it doesn't fit the model I was working on in part 1.
II. Fundamentals of Scum Hunting
One of the biggest misconceptions about scum hunting is that there is a ‘right’ or ‘most effective’ way to do it. What works for someone else may not work for you at all. With this in mind its all about finding what works for you. -But- There are some basic core differences between the Town and the Mafia that every player should hold as true; how you go about identifying them but it is a very personal journey every player must embark upon. This will outline a large portion of accepted 'Tells' and explain why they are alignment indicative to help assist you in figuring out a strategy for your personal approach to scum hunting.
2. Types of Tells
When attempting to divine hidden information people look for ‘tells’ or indicators of alignment. My take on scum hunting is that all real ‘tells’ will draw back to a few very specific but very critical differences between the Mafia and the Town. Below are what I consider the core fundamental differences and some things I consider alignment indicators with reasoning.
• The Town players have to figure out who is on their team or who is their enemy
• The Mafia players already knows who is and is not on their team, and generally have a good idea of the alignment of the rest of the players
• The Town players are acting from a genuine and honest perspective
• The Mafia players have to fake their perspective and deceive the town
Knowledge of Impossible Information(TMI)
Occasionally you can peg a hard scum slip when a player has some information that should be impossible for them to know. If the player claimed that they are a doctor and somehow knows another players role there is a potential that they received the information from their mafia team. (Think before challenging a player on this as that you could out town power roles in the wrong situation)
Gap Between Confidence and Reads
Because they already know the players alignment they have to ‘make up reasoning’ to justify whatever position they take on the player they are talking about. This creates a situation where their conclusion is not based upon their reasoning. This gap can be identified by looking to see if their level of confidence matches their reasoning (Its often easier to identify this gap in their town reads, Presented absolute certainty is a good indicator)
Handing out easy Town credit/reads
Mafia players naturally have a tendency to throw out town credit with much less reasoning than they do as Town players. This is because there is a tendency to require less information to confirm a held belief than to establish a one. Additionally veteran scum players recognize the natural tendency for players to think along the same lines. The player who just got town read knows they are town and therefor they exercise the same bias when looking to see if the reasoning justified the conclusion. But as a 3rd party townie you can identify this because you do not know that either players alignment (or if your experienced enough and maintain the knowledge that their perspective shouldn’t just be confirming a known alignment you may be able to recognize it being done to you.) You can also factor in buddying/pocketing efforts although it is not the core reason this tell exists in my opinion.
Town reading those who scum read them (counter to natural OMGUS)
The natural subconscious reaction to a player accusing you of being scum when you are town is to view them as scum. This is because you know with certainty that you are Town and that if you were to be correctly read it should be identifiable. Your mind catastrophizes the interaction and jumps to the conclusion that the other player must therefore be scum. But this is not true for the Mafia alignment. They know that they are scum and they know that the player accusing them is town. This difference can be identified in the way they respond to pressure; particularly in the early game where a town player would have no significant read on others alignments or in the later game when the player has expressed they read the player accusing them as scum prior to the interaction. If their defense does not come from a perspective that you would identify as questioning their accusers alignment and they have not expressed a belief that the accusing player is town you should take notice. (This is more effective on newer players who’s reaction to pressure is less guided by experience, particularly if their meta has a ‘everyone is guilty until proven innocent’ nature to it.)
This is a commonly known indicator that is created as a result of the scum player not speaking from their own perspective. Scum players will consciously consider how a town player would say what they are looking to say and unintentionally slip, speaking from the perspective they are thinking from. This is easily noticeable when pronouns such as ‘They, you, your’ are used in place of ‘I, My, Me.’
Lack of Natural flow to speech
This is also a result of the scum player not speaking from their own perspective. The information they are communicating is transformed from their perspective to the perspective they are presenting from. You may notice that the communication is just kind of awkward or out of place. Maybe there’s broken thought processes or some things just don’t quite connect in a normal way. Regardless it’s a red flag you should take notice of.
Lack of consistency in reads
This is a pretty strong indicator that reflects a disconnect between what the player says and what they do. Town players are generally consistent because their actions come from an honest place. If they scum read a player they will be willing or even eager to lynch them; or if they town read a player they may make an effort to defend them. For scum it is much more difficult. They have to keep track of all the fake positions they take and stick with those positions. I don’t think I have ever read through a game and not found at least some inconsistency’s in a scum player. Some of it is a result of evolving conditions such as being forced to choose between a player they voiced that they read as town and their scum team mate. In those situations you can often see a gap between the voiced reasoning for them changing their perspective and what would be natural. Scum players also slip sometimes by forgetting what position they took earlier.
This occurs when a team scum specifically hunts for a neutral killing. Many players do it because the neutral really is a threat to them but also because it gives them something to genuinely hunt. They no longer have to fake reads or worry about speaking from a perspective that is not their own and can act somewhat naturally; but this can also be identified if you notice the player is very specifically jumping to the conclusion that their read indicates a neutral. The natural perspective on a read would be that the player is some kind of evil or even not town making this a decent tell.
This is more specific to turbos or EOD when chat is fast paced but scum can often freeze up in response to pressure. A scum called out on a slip may need some time to figure out how he can justify his slip while a town needs no time to think and can simply be honest.
Because scum have to make up their reads / reasoning they can be caught in a slip. Once caught they will have to justify their position which can result in flailing. You may notice them constantly adjusting their warrant or grounds saying that you do not understand or even adjusting their initial claim. This can often be identified by breaking down their argument. If they change their claim check their warrant to see if it supports the adjusted claim. This can also occur when they have taken a position with a lack of reasoning and are later forced to justify that position and fail to do so in a way that appears natural.
• The Town players objective is to figure out who is on their team or who their enemy is and eliminate those players
• The Mafia players objective is to avoid being identified/lynched, keep their team mates from being identified/lynched, and to kill town players through any means available.
• The Mafia has no obligation to do anything but survive the day cycle until night so they can kill another player.
Focus on Self Preservation
This is both extremely common as well as easy to identify. Scum players core motivation is to stay alive while towns is to eliminate scum- even if it means sacrificing themselves. You can often identify this behavior by looking at the players focus as pressure builds (especially if its nearing the end of the day and they recognize they could very easily be on the chopping block.) Experienced town players will generally focus on forwarding their win condition even if it means their death. They may do things like push their reads with the knowledge that their demise will reveal their role, making their words come from a ‘confirmed town’ and recognizing that they will be taken more seriously. You will also see experienced players scum hunt while they are defending reflecting that their focus is not exclusively on ‘staying alive.’ In contrast scum players (especially inexperienced ones) will only say and do things in an attempt to avoid the lynch. If they feel that they have been had they will sometimes just completely shut down and stop interacting all together.
Lack of Scum Hunting
Novice scum can often make the mistake of not pretending to solve the game. This can be a reflection of there being no motivation for the player to identify scum because they have no need to. Look for players who pop in and seem to kind of 'hang out' / interact as if waiting for someone to give them something to do.
Information Instead of Analysis (IIOA)
Similar to a lack of scum hunting some scum players will spend time and effort organizing information instead of analyzing information. This allows them to show effort with a focus on reducing entropy while also shifting information, perspective, and avoiding many of the tells that come with scum hunting. Favoring information will always be a pattern for deductive scum hunters especially in the PR setups hosted here but a lack of inductive reasoning should be a red flag.
Hyperawareness of others perception of them
Scum players are naturally hyperaware of how others view them. Everything they post can be taken from the mindset of ‘how will the town view this’ and they take special note when others voice concerns about them. This can sometimes come out as unnatural jumps to address any issue raised about them and can also be reflected in the way they speak about themselves.
Intent to lynch/kill
Because scum players want to eliminate players with little consideration to who they can often be too opportunistic. They also have incentive to avoid voting on a scum team mate and will instead have motivation to be on the counter wagon. This action pays off for them well if the lynch is avoided in that they can not even be seen doing it unless it fails. These behaviors are most easily identified with wagonomics although a lynch-happy attitude can sometimes be identified.
Focus on 1 player
Some scum just peg 1 player to scum read. This allows them to appear as if they are attempting to solve the game while also not dealing with the natural counter push / OMGUS that comes when you scum read a town. I most easily seen this in low activity scum.
I often see scum players pushing on a player until the first moment of resistance arises and then dropping the issue. This happens because their intention was never to push the player but rather to be seen pushing the player and when they are pressured on it they have no reasoning to create conviction as well as little motivation to draw attention to themselves.
Because scum players have to fake their scum hunting they often take cues from the town players reads. Novice/Lazy scum will often just grab the general positioning of read walls and present them as their own or just regurgitate what the rest of the town has voiced when asked for an opinion. This is easy to identify if your active enough to really understand the gamestate and much more incriminating as the game progresses. If you notice a player is not generating original thought step back and in-context iso them to check for this behavior.
-Note that things like in game shot mechanics very specifically reflect intent but must be seen in context of WIFOM
• The Mafia players have a known Team that they can work with and a shared chat to coordinate activities
• The Town players usually have no team they coordinate with
• The Mafia Players have a vested interest in keeping their team mates alive
Building a counter-train
Experienced scum players will sometimes deflate the pressure on their team mates by instigating pressure on another player to distract from the situation. This generally works well and is difficult to nail down until the mid-end game where a pattern of it occurring can be identified but is something that can be noticed and bring up the question of if you should look closer into this players associations, especially in post flip analysis.
Good Mafia teams will use their special chat to coordinate some basic outlines of intentions. Things like which players to focus on lynching are often established prior to or during the day. Fortunately for the town these types of coordinated actions can become identifiable after enough time has progressed. Vote pattern analysis such as Wagonomics can point to this.
Unnatural interactions(Puppet Show)
Anyone who has ever taken a theater class has seen some really terrible acting that is so awkward, it makes you feel awkward just watching it. Thats the kind of feeling I can get when watching 2 scum play out some rehearsed interaction.
What a player does not say is sometimes more important to notice than what they do. You will can notice scum players intentionally avoiding situations or discussions because they would be very uncomfortable to deal with. Examples could range from someone bringing up analysis of the votes in a lynch or maybe the entire train on one of their scum team mates. They may not want to get their scum buddy lynched but more than that they really don’t want to be associated with trying to prevent him from being lynched so they just avoid the situation entirely. This aversion is difficult to notice in the moment but can more easily be identified as the game progresses. I personally find aversion to be much more common than coordination and very particularly am critical of a Power Wolf who has massive interactions and reads yet strangely avoids talking to or about a few players. I believe this aversion comes from being forced to either bus or make up a fake read that ties them together when reading a team mate.
Pillow Pushing (Bussing)
Teamscum will often feel a need to bus for any number of reasons, but they have the conflict of having a vested interest in keeping their team mates alive. This creates a situation where they want to be seen pushing but do not want to achieve any results that is noticeable.
Ironman is the opposite of strawman in terms of presenting someones argument. Teamscum will sometimes present their team mates defense as stronger than it is when speaking about their thoughts on that players case. This allows a strong speaker or a townread speaker to deflate pressure on a team mate and can be noticed by comparing the discussed players argument with the teamscums and identifying creative contributions to the original defense.
Rarely teamscum will coach etch-other in their separate chat and create a noticeable difference in a weaker player. Changes in an individuals colloquialism, punctuation, or a sudden deep quality of reads would be examples of what to identify.
An ISO reads or Isolation read is used when you isolate a player or players posts in order to identify tells or patterns. There are a number of types of ISO reads you can use with each having specific advantages given their focus.
A standard ISO is usually done by using your hosting sites ISO function or using search parameters to isolate an individuals posts from the game. Many players find that in doing so, patterns can be seen that that were not apparent when playing the game. It is exelent at pulling any of the Intention tells I listed outside of mirroring and any of the perspective tells I mentioned besides freezing.
A Multi ISO can be done if your hosting site has the built in function, using search parameters, or just putting 2 standard ISO’s next to etch other and looking at post count. Its real advantage is recognizing any Association tells (Although aversion may require quantifying interactions)
An In-Context ISO usually means skimming through the thread enough to keep up with the general idea of whats going on while focusing on a specific players posts. This is specifically effective at all Perspective tells as well as Mirroring, Aversion and building counter-trains.
-As a foot note one of my favorite little tricks is using the search function within an ISO. Makes it quick and easy to quantify anything such as how often another player is referenced by that player.
+1 -1 Reads (And variants)
This is generally used with an iso in an attempt to quantify something, usually a ‘scale’ of scummyness, behavior or association. Some players go a simple +1 or -1 while others use a scale or even add points to a set of categories (like Town/Neutral/Mafia, Spam/Scum Hunting/Analysis, or ATE/ATA/ATL ect..) This can be useful in drawing out patterns although I personally use it mostly in my notes to get direction as opposed to posting some +1 -1 wall to the town as a method of reading a player.
I will also mention this tool can be extremely useful for scum aligned players who are having difficulty gauging how strong they should present their read. (See Gap Between Confidence and Reads)
This is the favored technique for many good players that can be extremely effective. On a basic level you are using the actual game mechanics and game feedback to analyze the game though things like Wagonomics, Theory Crafting, and Feedback Analysis.
Wagonomics is breaking down vote patterns with inductive logic to get an idea of who is scum. You can do this by looking back at a vote count and recognizing which players voted who relative to the now known information. You will see this every day when people talk about “Joe was Mafia and Fred did not vote Joe” but you can take it many steps further by tracking players votes throughout the day. The saying “Actions speak louder than words” comes to mind when expressing intent. Some of the common ideas are:
• Individual voting off wagon of scum lynch as building a counter train (Assuming trains are V/W)
• Early vote park on a scum lynch gives little town cred
• Last minute switch to a locked in train on a scum could be seen as bussing
• Last minute switch to even up/shift trains as alignment indicative or kingmaker (Depending on result)
• History of being on the right or wrong ‘train’ as alignment indicative
Feedback Analysis is using all the other known information such as role flips, reveals, shots in a mini-game, and claimed feedback. In complex/large PR heavy games you can run 2 charts with 1 mapping both the players with their associated claimed feedback/roles, and a 2nd one mapping out each known/suspected PR in the game with the information you know on them from flips/reveals ect. By comparing these two charts you can be able to catch people in lies and use theory crafting to identify information.
Theory Crafting is used by creating a theory and matching it to what you know/believe in the game. A classic example would be a player saying “I think Joe/Fred/Bobby could be the scum team.” This is often used in conjunction with Wagonomics, Anti-spew, and Association tells in an attempt to create viable models of the scum team. The 2nd level of this is when a player builds many or all possible potentials and then examines the information for a lynch target. For example- if a player assumes that the scum team must either be Joe/Fred/Bobby, Sally/Fred/Robert, or Tiffiny/Joe/Fred then that player may view Fred as the logical lynch choice given in every situation he would be scum; and Joe second most suspicious given in 2/3 of the situations he is scum.
This can also be used to PR hunt by looking at known/claimed actions against a player in order to identify who would have motivation to preform that action (or what alignment a PR may be).
It additionally works in some semi-open setups that provide potential pools of roles to sort through claims for PR heavy or Flavor revealing games.
Meta hunting could be summarized as getting to know a player's normal behavior in association to their alignment in that game, and comparing it to their current play in order to get an impression of what their current alignment is. This extremely popular technique can be very effective when used responsibly, but also can be very detrimental if used inappropriately. Some of the common things focused on are things like:
Post/Read style, Techniques/Tells used, Focus in hunting, Tendency to troll / role-play, Piecemeal answers, reaction to pressure, tunneling, amount of direct confrontation, availability in game, expressed certainty of reads, Post density, referencing their meta, and use of meta reading
In my opinion its best used to identify what tells will be most effective on a specific player as opposed to being considered a technique in itself. I would post a caution to those who would use it as a technique in itself. People have bad days where they are sad or angry, they have busy days and real life stuff that comes up that you have no idea about but will change their behavior. It also has the critical problem of cherry pickers fallacy making people overestimate its value. The natural drive to take the path of least resistance can push players away from using more difficult but more fact/skill based techniques; and the reads generated are hampered by the player’s inability to push their view to anyone who doesn’t also have meta on that player.
You should consider the implications of the setup on how you scum hunt. If the setup changes a core reason for a scum hunting tell you should take time to consider how that must change the way you scum hunt.
Above all else you need to honestly evaluate your own play to figure out what works for you. If you dont reread your games you will be prone to bias in what you think works for you. This often results in players using techniques that just do not work for them and can really submarine their impact on games. Practice only helps you improve if you are able to recognize your mistakes so you can learn from them.
I am not going to say that this list includes every tell but I will say that I believe any scum hunting tell should draw back to a core difference. If yours do not you may want to reconsider it. Even if you disagree with how I look at scum hunting I believe you should be able to justify the tells you use to others, so that you can push lynches. If your position is not tenable you may be doing something very wrong.