How to Catch Wolves (Newbie Version)

How to Catch Wolves

You’re playing your first game of forum mafia. The game topic has a hundred or so posts, and you’re not sure how you’re supposed to go about finding scum in all those words. How would scum players play differently from town players, anyway? People are unique individuals with their own ways of expressing themselves. How can you tell if they drew a scum role card?

_______________________________________________________________________

What are the differences between town and scum?

The defining difference between a town player and a scum player is:

  1. Town players do not know who else is town, and also do not know who the scum are. All they know is their own alignment and role and whatever the host has shared publicly about the setup.
  2. Scum players know who all the town are, and also, barring certain exceptions, who all their scum partners are.

As a result, scum play to hide their knowledge, including their knowledge of who the rest of the scum are! Meanwhile, town play to increase their knowledge of the game, and try their best to figure out the identities of the scum.

The two teams have different win conditions (often abbreviated as wincons):

  1. The Town team wins by lynching all the scum players.
  2. The Scum team wins by surviving until the end of the game as they outnumber the Town players.

This may sound like the Town’s win condition is more difficult than the Scum win condition, but the playing field is often level because the Town outnumbers the Scum team by a ratio of 3:1 or 4:1 in most games. A game with 9 players usually has 2 scum. A game with 17 players usually has 4 scum, etc.

Your friend may be town in a game, and be scum in another game. These roles are randomly assigned, and do not change the person who is behind the screen playing them. Individual players will approach the game in different ways, so how can we possibly use a blanket, one-size-fits-all method to figure out who the scum are? Well, players’ motivations always revolve around their win condition and their knowledge (or lack of knowledge!) about the game. So, they will change the person’s tactics as they play the game.

Let’s look at how the tactics differ.

  • Town need to find and lynch scum. They must also figure out who else is part of the town and work together with them to get enough votes to lynch scum.
  • Scum do not need to find scum or town, they already know. They do need to lynch town.

This basic difference can help you find the scum in your game and also identify who is town. Town can be identified via genuine scum-hunting – attempts to identify scum. And scum are required to falsify or pretend to scum-hunt, in order to blend in with the town, while at the same time getting town players lynched.

If you can identify who is genuinely scum-hunting and who isn’t, you can solve the game.

As you play mafia, you’ll hear the term “scumtell”. A scumtell is a behavior that a scum player is far more likely to exhibit than a town player. Here are some technical examples.

  1. A player is exhibiting fewer behaviors of scumhunting than either a) other players in the game, or b) their previous performances as town in other games.
  2. A player exhibits behaviors that suggests that they already know certain players alignments in a way that is excessive compared to a) other players in the game, or b) their previous performances as town in other games.

That sounds great, but what does it actually mean? Let us explain a little further.

_______________________________________________________________________

A player is exhibiting fewer behaviors of scumhunting than either a) other players in the game, or b) their previous performances as town in other games.

So, what are scumhunting behaviors?

  1. Asking other players questions whose answers could reveal the other players’ alignment.
  2. Sharing reads on other players to communicate where they stand in the game. Reads can be shared in many forms and levels of detail. The reads will indicate how the player sees other players in the game from most town-like to most scum-like, and why.
  3. Making arguments about players who they believe are scum in attempts to get their scumreads lynched.
  4. Making arguments about players they believe are town in attempts to prevent their townreads from getting lynched.

In contrast, scum players may go through some of the same motions, but their questions, reads, and arguments are informed by their extra knowledge about player alignments, and are shaped by their objective of getting town players lynched.

How will scum play the game differently compared to town players? Here are some examples.

Case Study 1
Player X is in a game, asking many questions to other players, such as “What did you for breakfast this morning?”, “How much mafia experience do you have?” and “How many scum do you think there are in the game?”. He also often gives reads on other players along the lines of “Player A could be town, but he also could be scum.” He also always votes on the biggest and most popular wagons in the game, but never explains why.
Is Player X more likely to be town or scum?

Analysis: Player X is a pretty good bet for scum. Although he is asking many questions, those questions show a pattern of being irrelevant to the game. Knowing the answer to what a certain player ate for breakfast would not tell us any information as to that player’s alignment! This does not appear to be genuine scumhunting! Also, the reads that he had given on other players appear to be fabricated: Of course they could be either town or scum, what else could they be? This further hints at Player X simply going through the motions of scumhunting, instead of truly trying to figure out the game. Finally, he always votes on the largest wagons in the game without explanation. Does he not care if the target flips town or scum? This is another red flag that Player X is not interested in scumhunting, or to find out who the scum are in the game. However, even though he appears to be uninterested in scumhunting, we can also rule out the fact that he is absolutely uninterested in the game, since he is still participating in the game by asking questions and giving reads, even if they are meaningless. One may be a false positive, but three is not a coincidence! All of these actions suggest that he is more interested in pretending to look for scum, rather than actually finding scum, which, as we have discussed earlier, is strong evidence that points to Player X being scum himself! Since he is scum, he does not need to look for the rest of the scum, because he already knows who they are! He isn’t doing a really good job with his act, though!

Case Study 2
Player Y is in a game, expressing a strong scumread on Player C. Player Y asks questions to other players, such as “Who do you think are the scum in this game?”, “What are your thoughts on Player A, B, and C?”, and “Why do you think Player C is town?”. He strongly makes arguments for Player C being scum, and achieves a lynch on Player C, but eventually Player C flips town.
Is Player Y more likely to be town or scum?

Analysis: Good evidence points to Player Y being town. Despite the fact that he was likely a major factor for the lynch of a town player C, we can see that he is trying to figure out the motivations of other players, including those of which who believe that Player C is town; Player Y could be possibly trying to learn if those other players are scum along with Player C and trying to protect Player C. This suggests genuine scumhunting. Player Y also displayed strong conviction and motivation in getting his scumreads lynched, which is an excellent example of putting the “hunt” in “scumhunting”, even if it turned out wrong in the end, he did try to advance the town’s goal in his own way by lynching someone whom he really thought was scum. Good people make mistakes all the time, and they are not necessarily scum for it! Although it is possible for a scum player to fake this level of scumhunting, they must be extremely skilled in order to do so: they must understand in detail the concepts put forth by this article, and be able to execute their act of deception extremely well, which few players are capable of. It would be a safe bet to consider this player as town, as even though he has made a mistake, he genuinely appears to have the town’s best interests at heart.

_______________________________________________________________________

What other behaviors could indicate a player is scum? Here are a few.

A player’s behaviors suggest that they already know certain players alignments in a way that looks excessive compared to a) other players in the game, or b) their previous performances as town in other games.

Scum already know who their partners are, as well as who all the members of the town are. They basically have the answers to the whole game, and have to hide this information. Town gets the short end of the stick, knowing none of the players’ alignments or power role information. Town’s strength is in numbers.

So, scum must do their best to hide their knowledge of other players’ alignments, and follow their win condition by surviving as long as possible. However, they want to discourage the deaths of their teammates because losing teammates will make the game last longer which gives them more time to make mistakes and gives Town more time and information to solve the game with.

So how does this information difference translate into different behaviors? They could be scum if:

  • The player acts or exhibits behaviors that indicate that they know a player’s alignment is town.
  • The player acts or exhibits behaviors that indicate that they know a players alignment is scum.

These behaviors made by a player may not always be apparent on first sight, but most of them happen to be glaringly obvious after players have been lynched and their alignments made public to all. For instance, a certain player who shows a consistent behavior of protecting every revealed scum will likely be put on the suspect list – is he protecting those scum because he knows through special information sources that they are scum, or does his actions and behaviors in his posts indicate that he really doesn’t know? That’s for you to find out.

Here are more examples that may conceivably happen in a real game.

Case Study 1
In some mafia game, the town seems to be in a disaster. The collective town has already made three consecutive mislynches in a row, and another mistaken lynch would bring the town into a state of LYLO - the last chance for the town - any subsequent mislynch would result in the game being decided in scum’s favor. One of the players on the suspect list today is Player A, who has not voted for any of the three mislynches, always voting elsewhere, and correctly called the three mislynches town at some point.
Should Player A be lynched?

Analysis: Lynching Player A would be a good call, yes. The assumption here is that scum are playing well, having decent capabilities to convincingly fake the behaviors of scumhunting, and as a result are managing to evade three whole days of lynches without a single casualty. However, Player A may be scum who have made a mistake – unless he is Gandalf, Player A is probably not getting all those correct reads without some help – perhaps it’s because he has the answer key from being scum! A majority of players is needed for a lynch: how likely is it that the town supporting those lynches was wrong all the time, while Player A was correct all of the time? Furthermore, another point against Player A would be the fact that he did not put in enough effort to prevent his townreads from being lynched in the first place – if he really thought they were town, wouldn’t he have tried with all his might to prevent them from getting lynched? One failure may be acceptable, but three failures in succession definitely smell fishy. It appears that Player A is not very interested in protecting his townreads or in helping the town to avoid losing, which is a scumtell that was mentioned in the previous chapter. So, despite the fact that his answers were all correct, it’s more likely that Player A is an evil saboteur rather than a mafia prodigy.

Case Study 2
In another mafia game, Player B has been lynched, and has been revealed to be scum. The collective town tries to figure out who could be partners with Player B. For Player C, they find evidence of the following on the previous day: Player C asks questions to every other player in the game, except for Player B. Players B and C do talk to each other, but less frequently as compared to with other players during the game’s day phase. When Player B was close to being lynched, Player C made the correct finishing vote on Player B, sealing his lynch and revealing Player B as scum. Player C made no indication that he was scumreading B prior to that vote.
Is Player C a good lynch?

Analysis: Assuming that there are no more better options than Player C in the game, where nobody else has violated the principles in the earlier chapter enough to be obviously distinguished as scum, Player C would be a good lynch. Player C’s special treatment of Player B suggests that Player C knows something that the rest of the town did not, even before Player B was lynched and was revealed to be guilty. Furthermore, as Player B and C have talked to each other in the game, it may be assumed the cause of Player C and Player B’s unusual interactions is not due to personal reasons. Did Player C find it unnecessary to ask questions to Player B to figure out Player B’s alignment, because Player C already knew that Player B was scum? Or could it be that Player C did not want to draw attention to Player B because it may have resulted in his teammate, Player B, being more obvious as being scum? Player C’s hammer vote on Player B may point further to Player C being partners with Player B: Not only is Player C now showing that he magically knows that Player B is the correct lynch, he is also failed to display an aspect of genuine scumhunting, from the fact that he has failed to communicate to the rest of the town his read on Player B for the entirety of the day. It is difficult to believe that Player C was unable to make a definitive opinion on Player B during that day, when there was likely generous discussion going on regarding the topic of whether to lynch Player B. If there is more evidence to be found in the game that supports Player C having fabricated or faked scumhunting, which are scumtells discussed in the earlier chapter, then Player C would make a very fine lynch.

_______________________________________________________________________

These broad kinds of tells apply to the player’s body of work in the game thread, not necessarily to single posts. Although sometimes a single post will give a scum player away, it’s far more common for the lack of town behaviors or presence of scum behaviors to become apparent over time. As the game progresses, you can examine all of a player’s posts, and, observe if their play fits into one of the above categories. If they do fit, they might be scum.

These are skills you can practice by playing games, but you can also practice by reading completed games and seeing how well you do at identifying player alignments. With practice comes confidence, and you’ll soon find yourself developing accurate reads, making strong arguments, and leading lynches on scum players.

Go catch them all!

_______________________________________________________________________

Note: This primer article is based on a Mafia451 forum topic by Metal Sonic that is geared toward experienced players as well as newer players. Metal Sonic and fferyllt worked together on this version, which is designed to be helpful for new players. Although written for forum mafia, the same principles apply to face-to-face mafia and other game formats.

My revision will be in the next post. I’ve tried to remove some of the clauses and phrases that are basically filler, switch from passive to active verb forms, remove unneeded adverbs (this is one of my own weaknesses in writing – adverbs are shiny!).

There is a lot of parallelism in the construction of the article, which is good if not overdone. Otherwise it gets very redundant and can make the article less readable. I also removed most of the exclamation marks. They’re a little jarring in an informational essay format.

So basically, this is a “less is more” revision.

How to Catch Wolves

You’re playing your first game of forum mafia. The game topic has a hundred or so posts, and you’re not sure how you’re supposed to go about finding scum in all those words. How would scum players play differently from town players, anyway? People are unique individuals with their own ways of expressing themselves. How can you tell if they drew a scum role card?

_______________________________________________________________________

What are the differences between town and scum?

The defining difference between a town player and a scum player is:

  • Town players don’t know who else is town, and also don’t know who the scum are. All they know is their own alignment and role and whatever the host has shared publicly about the setup.
  • Scum players know who all the town are, and also, barring certain exceptions, who all their scum partners are.

As a result, scum players play to hide their knowledge, including who the rest of the scum are. And, town play to increase their knowledge of the game, and try their best to figure out the identities of the scum.

The two teams have different win conditions (often abbreviated as wincons):

  • The Town team wins by lynching all the scum players.
  • The Scum team wins by surviving until the end of the game as they outnumber the Town players.

This may sound like the Town win condition is more difficult than the Scum win condition, but the playing field is often level because the Town outnumbers the Scum team by a ratio of 3:1 or 4:1 in most games. A game with 9 players usually has 2 scum. A game with 17 players usually has 4 scum, etc.

Your friend may be town in one game, and be scum in another game. These roles are arbitrarily assigned, and do not change the person who is behind the screen playing them. Individual players will approach the game in different ways, so how can we possibly use a blanket, one-size-fits-all yardstick to figure out who the scum are?

Players’ motivations always revolve around their win condition and their knowledge (or lack of knowledge) about the game. Motivation will change the person’s tactics as they play the game.

Let’s look at how the tactics differ.

  • Town need to find and lynch scum. They must also figure out who else is part of the town and work together with other town players to get enough votes to lynch scum.
  • Scum do not need to find town, they already know. They do need to lynch town, and they need to persuade enough town players to vote another townie that a lynch can be accomplished.

This basic difference can help you find the scum in your game and also help you identify who is town. Town can be identified via genuine scum-hunting – attempts to find scum. And scum are required to falsify or pretend to scum-hunt, in order to blend in with the town, while at the same time getting town players lynched.

If you can figure out who is genuinely scum-hunting and who isn’t, you can solve the game.

As you play mafia, you’ll hear the term “scumtell”. A scumtell is a behavior that a scum player is far more likely to exhibit than a town player. Here are some examples.

  1. A player exhibits fewer behaviors of scumhunting than either a) other players in the game, or b) their previous performances as town in other games.
  2. A player exhibits behaviors that suggest that they already know certain players alignments sooner than or more confidently than a) other players in the game, or b) their previous performances as town in other games.

That sounds great, but what does it actually mean? Read on, and find out.

_______________________________________________________________________

A player exhibits fewer behaviors of scumhunting than either a) other players in the game, or b) their previous performances as town in other games.

So, what are scumhunting behaviors?

  1. Asking other players questions whose answers could reveal the other players’ alignment.
  2. Sharing reads of other players to communicate where they stand in the game. Reads can be shared in many forms and levels of detail. The reads will indicate how the player sees other players in the game from most town-like to most scum-like, and why.
  3. Making arguments about other players, who they believe are scum in attempts to get their scumreads lynched.
  4. Making arguments about players they believe are town in attempts to prevent their townreads from getting lynched.

In contrast, scum players may go through some of the same motions, but their questions, reads, and arguments are informed by their extra knowledge about player alignments, and are shaped by their objective of getting town players lynched.

How will scum play the game differently compared to town players? Here are some examples.

Case Study 1
Player X is in a game, asking many questions to other players, such as “What did you do for breakfast this morning?”, “How much mafia experience do you have?” and “How many scum do you think there are in the game?”. He also often gives reads on other players along the lines of “Player A could be town, but he also could be scum.” He also always votes for the biggest and most popular wagons in the game, but never explains why. Is this player more likely to be town or scum?

Analysis: Player X is a pretty good bet for scum. Although he is asking many questions, most are irrelevant to the game. Knowing the answer to what a certain player ate for breakfast would not tell us any information about that player’s alignment. This does not appear to be genuine scumhunting. Also, the reads he gave of other players appear to be fabricated: Of course the other players could be either town or scum, what else could they be? Player X appears to be simply going through the motions of scumhunting, instead of trying to figure out the game. Finally, he always votes on the largest wagons in the game without explanation. Does he genuinely care if the target flips town or scum? This is another red flag that Player X is not interested in finding the scum in the game. But, he’s showing up and making posts with at least some content, so he is interested in the game. He’s still participating by asking questions and giving reads, even if they are meaningless. One may be a false positive, but three red flags are probably not coincidence. All of these actions suggest that he is more interested in pretending to look for scum than in actually finding scum, and that is strong evidence suggesting Player X, himself, is scum. If he’s scum, he doesn’t need to look for the rest of the scum, because he already knows who they are. His efforts look like an act, not like actual scumhunting.

Case Study 2
Player Y is in a game, expressing a strong scumread on Player C. Player Y asks questions to other players, such as “Who do you think are the scum in this game?”, “What are your thoughts on Player A, B, and C?”, and “Why do you think Player C is town?”. He strongly makes arguments for Player C being scum, but eventually Player C flips town._ Is this player more likely to be town or scum?_

Analysis: His questions are evidence that point to Player Y being town. Despite the fact that he was likely a major factor for the lynch of a town player C, we can see that he is trying to figure out the motivations of other players, including those of which who believe that Player C is town; Player Y could be trying to learn if those other players are scum along with Player C and are trying to protect Player C. This looks like genuine scumhunting. Player Y also displayed strong conviction and motivation in getting his scumreads lynched – an excellent example of putting the “hunt” in “scumhunting”. Even though he turned out wrong in the end, he did try to advance the town’s goal by lynching someone whom he claimed he thought was scum. Town players make mistakes all the time, so it’s important to look beyond the results (a townie getting lynched) to Player Y’s motivations. Although it is possible for a scum player to fake this level of scumhunting, it would take a high level of skill to do so. They would understand in detail the concepts put forth by this article, and be able to fake scumhunting quite well. It would be a safe bet to consider this player to be town. Even though he has made a mistake, he appears to have the town’s best interests at heart.

_______________________________________________________________________

What other behaviors could indicate a player is scum? Here are a few.

A player’s behaviors suggest that they already know certain players alignments in a way that looks excessive compared to a) other players in the game, or b) their previous performances as town in other games.

Scum already know who their partners are, as well as who all the members of the town are. They have most of the answers to the whole game, and have to hide this information. Town gets the short end of the stick in terms of knowledge. They know none of the players’ alignments nor any power role information (Special powers given to some town players. Most town players have only their vote and their wits). Town’s strength is in numbers.

So, scum must do their best to hide their knowledge of other players’ alignments, and follow their win condition by surviving as long as possible. They want to discourage the deaths of their teammates because losing teammates will make the game last longer. The longer the game lasts, the more time and opportunities scum have to make mistakes, and the more time and information Town have for solving the game.

So how does this information difference translate into different behaviors? They could be scum if:

  • The player acts or exhibits behaviors that indicate that they know a player’s alignment is town.
  • The player acts or exhibits behaviors that indicate that they know a players alignment is scum.

These behaviors may not always be apparent on first sight, but they can become glaringly obvious after some players have been lynched and their alignments made public to all. For instance, a certain player who shows a consistent behavior of protecting every revealed scum will likely be put on the suspect list – is he protecting those scum because he knows from his role pm or other information sources that they are scum, or do his actions and behaviors indicate that he really doesn’t know? That is your task – to figure him out.

Here are more examples that can happen in a real game.

Case Study 3
In some mafia game, the town seems to be in a disaster. The collective town has already made three consecutive mislynches in a row, and another mistaken lynch would bring the town into a state of LYLO (Lynch or Lose) - the final chance for the town where any subsequent mislynch or no-lynch would result a scum win. One of the players on the suspect list today is Player A, who has not voted for any of the three mislynches, always voting elsewhere, and correctly called the three mislynches town at some point. Should Player A be lynched?

Analysis: Lynching Player A could be a good call. Given the game state, where the town team is against the ropes, you can assume that scum are playing well, and have decently faked the behaviors of scumhunting. They have managed to evade whole days of lynches without a single casualty. However, Player A could be scum who has made the mistake of having too-accurate town reads while at the same time not pushing his reads enough to prevent mislynches. Player A may be so accurate due to help – maybe he has the answer key to the game because he’s scum. A majority of players is needed for a lynch: how likely is it that the town supporting those lynches was wrong all the time, while Player A was correct all of the time? Another point against Player A would be the fact that he did not put in enough effort to prevent his townreads from being lynched in the first place – if he really thought they were town, wouldn’t he have tried harder to prevent them from getting lynched? One failure may be acceptable, but three failures could be more than coincidence. Even though his reads were all correct, it’s possible that Player A is a scum saboteur rather than a town prodigy.

Case Study 3
Player B has been lynched, and has been revealed to be scum. The collective town tries to figure out who could be partners with the lynched scum player. Player C asked questions to every other player in the game, except for Player B. Players B and C did talk to each other, but less frequently compared to with other players. When Player B was close to being lynched, Player C made the correct finishing vote on Player B, sealing his lynch and revealing Player B as scum. But, Player C gave no indication that he was scumreading B prior to that vote. Is Player C a good lynch?

Analysis: Assuming that there are no better options than Player C in the game, where nobody has been more obvious about not scumhunting or about having information that town players wouldn’t have, then Player C could be a good lynch. Player C’s special treatment of Player B suggests that Player C knew something that the rest of the town did not. Did Player C not need to ask questions of Player B because Player C already knew that Player B was scum? Was Player C trying to avoid bringing attention to Player B because it would highlight that Player B could be scum? Was Player C avoiding interactions with Player B because he expected Player B to get lynched and didn’t want to look like a teammate? It may not be possible to nail down precisely why Player C’s hammer vote on Player B makes them possibly partners, but the interactions raise several red flags. Player C seemed to magically know that Player B was the correct lynch, and he didn’t communicate his apparent scum read of Player C at any point before voting him. If there is even more evidence similar to the earlier cases that supports Player C having fabricated or faked scumhunting, then Player C could make a very fine lynch choice.

_______________________________________________________________________

These broad kinds of tells apply to the player’s body of work in the game thread, not necessarily to single posts. Although sometimes a single post will give a scum player away, it’s far more common for the lack of town behaviors or presence of scum behaviors to become apparent over time. As the game progresses, you can examine all of a player’s posts, and, observe if their play fits into one or more of the above categories. If they do fit, they might be scum.

These are skills you can practice by playing games, but you can also practice by reading completed games and seeing how well you do at identifying player alignments. With practice comes confidence, and you’ll soon find yourself developing accurate reads, making strong arguments, and leading lynches on scum players.

Go catch them all!

_______________________________________________________________________

Note: This primer article is based on a Mafia451 forum topic by Metal Sonic that is geared toward experienced players as well as newer players. Metal Sonic and fferyllt worked together on this version, which is designed to be helpful for new players. Although written for forum mafia, the same principles apply to face-to-face mafia and other game formats.

It’s still a fairly long article.>2500 words.

And what if you are the Wolf and don’t want to get caught? :smiling_imp:

stay tuned for another episode to find out how!

1 Like

how do I in a game