NinSheetMusic Forums

Please login or register.

Login with username, password and session length
Advanced search  

News:

Welcome to the updating team, Seb and Insig!

Author Topic: The Werewolf Game: Tactics, Strategy, and Game Theory  (Read 163 times)

NocturneOfShadow

  • Nintendo Nocturne
  • ********************************
  • Gender: Male
  • Posts: 10408
  • This is Halloween
    • View Profile
The Werewolf Game: Tactics, Strategy, and Game Theory
« on: August 26, 2017, 07:52:40 PM »

So you've learned the rules of TWG, maybe even played in a handful.  But you can't quite seem to get how those good players know just what they know.  Here I'm going to break down some TWG strategy that goes slightly beyond the basic rules.

General Strategy

Since the play of the game varies quite a bit depending on your role, I'm going to break up each section into 3 overarching sections- Vanilla Humans, Werewolves, and Special/Third Party Roles.

Werewolf Strategy

Playing the "perfect" game as a werewolf is nearly impossible for all but the best of the best.  The werewolves are the only players at the start of the game who (usually) have perfect information- so you have to be careful of what you say.  Keen players will pick up on any slip of information you shouldn't have, so it takes a well-trained tongue.  In addition to this, you have to attempt to play your usual game of oblivious human and pretend to forget all the information you know.  And, you have to quietly and persistently manipulate the entire human team down the wrong trail while making a good case for why that trail is very pleasant and inviting.  For these reasons I believe the werewolf team is the hardest role to learn and play.

Tools Available to Werewolves:
  • Typically the werewolves have some power in addition to their night kill.  Most common is a Seering.  You thus have the ability (in alliance games) to potentially fake-claim seer to the person setting up the alliance, but in most instances it isn't a good idea.  There will always be advantages and disadvantages, so you need to learn when the advantages outweigh the disadvantages.
  • The number of werewolves alive is incredibly important.  The more alive, the easier it is to reach parity.  As a general rule of thumb, it's a bad idea to make a play if that play will end up costing you one of your numbers.
  • It can be easy to gain some free town cred by doing something simple like pointing out a humany post or dissecting the game setup for potential human advantages.  That can backfire though if you notice something you don't want the humans to find out.
  • When one of your partners is thrown under suspicion, sometimes it's best to add to the pile and sometimes it's best to try to steer the humans off.  A few factors can help you determine which is better.  If players don't get cardflips, you should be a little leery of "bussing"- throwing your partner "under the bus".  If there is an unusually large number of wolves, bussing can also be more profitable.
  • Being overly confident about a player's role can be generally viewed as wolfy.  Humans usually can't know for sure what role someone is until they've died, and maybe even then.
  • Being inquisitive as to the identity of Special Roles in the game can be generally viewed as wolfy.  In most cases, the Specials can do their own thing without guidance from the Human mass, so discussion about Specials is frowned upon in general.

Human Strategy

Since Human is the most common role and therefore frequent, players tend to develop pretty good human skills.  Being Vanilla Human is difficult in its own right because you have no unique powers to help you gain information, and you start off with little to no information.  Being a good human is nearly as hard as being a good werewolf, but being an adequate human is much easier.

Tools Available to Humans:
  • It's the responsibility of the Humans to generate information.  If you ever find yourself thinking "I don't know anything", it's because you haven't tried to find anything out.
  • In games with OOTC (Out of Thread Contact/Communication), the humans are benefited because you can share information without letting the wolves know you've discovered this piece of information.
  • Inactivity benefits the wolves.  If you see nobody has posted in the thread lately, it's your responsibility to make a post.  Even something as simple as pointing out, "Hey.  I think this is wolfy." is enough to spark some conversation.
  • As you're looking for the wolves, pay careful attention to interactions between players.  Wolf partners will publicly interact in some ways that humans don't normally interact with each other.  Wolves will also generally have fewer logs to share between them because it can be difficult to feign lack of knowledge to each other.
  • There's no such thing as a "bad" wolf.  If someone seems quite obviously wolfy, you may have to look again.
  • On the flip side, sometimes you do play with new players, but they will still be trying their hardest to conceal their identity and are unlikely to try anything that could out them.
  • If anything jumps out to you on a specific player, you MUST share that information with at least one other player.  If anyone is confirmed human, always send them all your logs, all the time.  If nobody is confirmed human, just share that specific information with someone who is very unlikely to be a werewolf with that player.
  • If you're having a hard time finding out who could be a wolf, look for humans instead.  This helps narrow down who could be wolves tremendously.

Special/Third Party Strategy

These roles are typically very fun to play.  Special Roles are among the easiest roles to play well, and Third Party Roles can range from something as simple as Fool to something as hard-core as Serial Killer.  The big thing that characterizes Third Party Roles is their unique win condition.  As a result, they're considered a threat to the Human team.

Tools Available to Other Roles:
  • Most Third Party Roles won't want to claim any kind of Power Role at all.  When in doubt, if you have to claim to someone, claim human.
  • If you're the Seer, if you seer someone green and there's a Master Wolf in the game, you could wait until you get a second green seering and claim to both of them at the same time to let them know the other is the Master Wolf if you die.  Alternatively, the odds are so low you can just claim to them anyway with low chance of failure.
  • In most setups as the Seer, if you seer someone Red you should claim your role at that point.
  • The Vigilante should almost never claim.  If someone counterclaims Vigilante, the night actions will determine who was lying.  The best Vigi is the silent Vigi.
  • If you're not sure who to shoot overnight, the second highest voted player during the day is a safe bet.  This helps the humans focus on another target and prevents the game from stalling.
  • As the Guardian, you should try your best to guard the player most likely to die.  The Seer may sometimes claim publicly to set up an alliance, so you'll have to guard them.
  • As any Special Role, if you're about to die, you must claim publicly.
  • However, don't just claim because you feel like you're under pressure.  Wait until the human block is for sure going to try to kill you.

Theory

While no two players have the same playstyle, there's a lot of theory you can apply to a player without a specific metagame from reading/playing in lots of TWGs.  While no theory is foolproof, to paraphrase a departed player, this game amounts to yelling at someone 'till they've died.  So if you're wrong, sometimes you're wrong.  But certain archetypes crop up and being able to use them to your advantage will help you win more often than not.

Reading the Wagon

The theoretical "best" wagon at the end of a day is one with a very close number of voters on each person.  If the person who flips ends up being a wolf, it's pretty simple to assume everyone on that wagon is human- you essentially cut the number of possible wolves in half.  When this doesn't work is when bussing occurs, but if the two wagons are close enough, the wolves aren't likely to vote for their partner when they can vote for someone else without fear of suspicion.  The worst possible wagon is either a many-way "kitb"- "Knife-in-the-box" (a tie), or an abnormally large amount of voters on a single person.  If one person is an uncontested lynch candidate, that means there's almost certainly a wolf among them.  Other, lesser things to watch for in the votes is someone making a defensive vote without much context- they could be "a bad player", or they could be a wolf starting to panic.  A smarter wolf will vote for someone they can make a case for.

Emotion in TWG

While TWG is a game about deception, manipulation, and intellect, tempers can flare and emotions can get in the way.  From my experience, someone getting very emotional about the game is more likely to mean they're human.  Unfortunately, it often gets them lynched because they are viewed as obstructing the human game plan and potentially a wolf.  If you have a problem with ornery temper, consider listening to chill or up-tempo music while playing.  If you can bob your head to the beat, you can detach from what you feel more easily and think logically about scenarios.  Something else I see often in games (but not usually on NinSheetMusic) is two players going at it with each other.  This behavior is very anti-town, yet it usually ends up being two humans in the conflict.

Metagame, or the "FireArrow Effect"

A decently skilled TWG player named FireArrow considers himself to have "no metagame"- that is, he doesn't play a specific way depending on his role.  I firmly believe that only the best of the best players can remove themselves from metagame context; therefore you can apply "would X player do this as Y role?" to any argument you may have on this TWG forum.  Understanding metagame is important, and as Davy and I found from playing over at FlashFlashRevolution, metagames develop differently in different forums.  So while it's not always possible to rely on metagame, when you can it's one of the most useful tools you have.

Aggressive Bussing

While bussing isn't a common tactic on NinSheetMusic, there are absolutely do's and do not's to bussing.  It's generally accepted that if there are only 2 wolves, you shouldn't bus your partner at all.  I've played games with 4 wolves where they did nothing but bus each other, and it ended up working out for them, as one wolf died and the rest ended up living through heavy fire and won the game.  Another strong reason to not bus somebody is the presence of cardflips- usually bussing only works if you know the person who died was a wolf.  Note that bussing is not the same thing as bandwagoning.  As a wolf, if you're reasonably confident that your partner will die, staying off their wagon will only increase suspicion towards you.


Alliance Games

Alliance games are uncommon and are heavily setup-dependent.  They require a large number of Special Roles and a way to protect the alliance- some common examples of this are a Guardian or a Brutal Human.  Typically the designated public special will claim to the thread on the first day phase and the rest of the specials wait for potential counterclaims.  Once nobody has counterclaimed each player messages their role to the public special and that person decides who is and isn't in the alliance.  Sometimes a wolf may sneak into the alliance, but it's a tricky thing to do.  NSM has a lot of examples of alliance games done both well and poorly, so be sure to check them out.  The important thing about alliance games is they benefit the humans so the game is balanced by making the wolf team overpowered.  Even though it can be easy to win with a well-set-up alliance, getting to that point is more difficult than it sounds, and alliances more often than not only end up being 3-4 players (the specials and the person who got seered).


This version has only scratched the surface of what you can learn about TWG, but since I wrote it all in one sitting it's hard for me to come up with strategies that I consistently use all at once.  Hopefully you learned something from this and if there's anything I forgot or anything you think should be made known just shout it out.

« Last Edit: August 27, 2017, 09:33:34 PM by NocturneOfShadow »
Logged
time for a splatfest! Christmas vs Halloween!

only Christmas sucks so Halloween wins by default

davy

  • Moderator
  • Rupee Ragtime
  • *
  • Gender: Male
  • Posts: 1136
    • View Profile
Re: The Werewolf Game: Tactics, Strategy, and Game Theory
« Reply #1 on: August 26, 2017, 09:28:32 PM »

    This is a really good guide. I'm impressed.

    I've got a couple of things to say about it:

  • Being inquisitive as to the identity of Special Roles in the game can be generally viewed as wolfy.  In most cases, the Specials can do their own thing without guidance from the Human mass, so discussion about Specials is frowned upon in general.

I agree that the identitiy of the specials is something that humans (and therefore wolves), should not be talking about. However, discussion about the actions that specials should take is gennerally seen as a good thing on NSM. Specials might not always know what the best course of action is with their powers, and the discussion that is created by talking about specials leads to activity, which is always good for the human team.

  • There's no such thing as a "bad" wolf.  If someone seems quite obviously wolfy, you may have to look again.

There is such a thing as a bad wolf, just like there is such a thing as a bad TWG player. While I agree that obvious wolf behaviour doesn't always mean that player is a wolf, it should not give that player a free passage.

  • If anything jumps out to you on a specific player, you MUST share that information with at least one other player.  If anyone is confirmed human, always send them all your logs, all the time.  If nobody is confirmed human, just share that specific information with someone who is very unlikely to be a werewolf with that player.

Rather than sharing the info privately, most time it is better to share the information in the topic. Only if the information is hurtful to the humans should it be discussed privately. If there are no confirmed humans, you must weigh the risk of sharing the information with a potential wolf to the benifit of sharing the information with another human.

  • Most Third Party Roles won't want to claim any kind of Power Role at all.  When in doubt, if you have to claim to someone, claim human.

Special should always claim their own role to confirmed humans. If a special has claimed publicly and hasn't been counterclaimed, you can assume that player to be a confirmed human. If you claim normal human to such a confirmed human, you are giving the wolf team the opportunity to fake-claim.

  • If you're the Seer, if you seer someone green and there's a Master Wolf in the game, you could wait until you get a second green seering and claim to both of them at the same time to let them know the other is the Master Wolf if you die.  Alternatively, the odds are so low you can just claim to them anyway with low chance of failure.
  • In most setups as the Seer, if you seer someone Red you should claim your role at that point.

This is assuming that there is no human alliance. If there is a human alliance, red results should be posted immediately, green results should be posted when it is neccessairy to either safe that player from death to lynch or to lower the amount of potential wolves (while keeping in mind that there might be a master wolf).

  • The Vigilante should almost never claim.  If someone counterclaims Vigilante, the night actions will determine who was lying.  The best Vigi is the silent Vigi.

This is extremely dependent on the game. When there is a guardian, vigi is one of the best roles to claim publicly. If the vigi has no protection, it is indeed better to stay off the wolves' radar.

Metagame, or the "FireArrow Effect"

A decently skilled TWG player named FireArrow considers himself to have "no metagame"- that is, he doesn't play a specific way depending on his role.  I firmly believe that only the best of the best players can remove themselves from metagame context; therefore you can apply "would X player do this as Y role?" to any argument you may have on this TWG forum.  Understanding metagame is important, and as Davy and I found, metagames develop differently in different forums.  So while it's not always possible to rely on metagame, when you can it's one of the most useful tools you have.  Take for example BlackDragonSlayer: his humanity negatively correlates to the number of :P faces he makes during the game! (Not always, but he does use them a lot, doesn't he?)

Giving an example of a specific player has no place in a general guide. This is something that should be discussed in game topics, and with that player's involvement.

can a mod pin this

Can someone make me officially TWC so I can pin this, lock some of the older topics and make some edits to the Basic Guide and Official Rules?
Logged
Quote from: NocturneOfShadow
[davy]'s in a way different time zone so basically he pops in at like 2 AM and posts 500 words and wins the game

NocturneOfShadow

  • Nintendo Nocturne
  • ********************************
  • Gender: Male
  • Posts: 10408
  • This is Halloween
    • View Profile
Re: The Werewolf Game: Tactics, Strategy, and Game Theory
« Reply #2 on: August 26, 2017, 09:45:49 PM »

This is a really good guide. I'm impressed.
*leers suspiciously*

Quote
I've got a couple of things to say about it:

I agree that the identitiy of the specials is something that humans (and therefore wolves), should not be talking about. However, discussion about the actions that specials should take is gennerally seen as a good thing on NSM. Specials might not always know what the best course of action is with their powers, and the discussion that is created by talking about specials leads to activity, which is always good for the human team.
well, on NSM yes.  We have an influx of new players, but eventually they won't be new players, and at that point specials need to know how to do their job on their own.  Since this guide aims to get players to that point, I decided to look at it that way.
Quote
There is such a thing as a bad wolf, just like there is such a thing as a bad TWG player. While I agree that obvious wolf behaviour doesn't always mean that player is a wolf, it should not give that player a free passage.
You're right.  I should probably rephrase this one, but my point was that we often look at behaviors that are considered wolfy but we don't consider that it's not something a wolf would do, if that makes sense.
Quote
Rather than sharing the info privately, most time it is better to share the information in the topic. Only if the information is hurtful to the humans should it be discussed privately. If there are no confirmed humans, you must weigh the risk of sharing the information with a potential wolf to the benifit of sharing the information with another human.
There's definitely some information that shouldn't be made public.  I suppose it could have a dependence on playstyle.  BDS says he shares almost nothing to the thread (which I think is wrong) and I find it beneficial to constantly talk to everyone about other people and point out specific things that person did.  I'll have to give this point more thought.

Quote
Special should always claim their own role to confirmed humans. If a special has claimed publicly and hasn't been counterclaimed, you can assume that player to be a confirmed human. If you claim normal human to such a confirmed human, you are giving the wolf team the opportunity to fake-claim.
I think you misread me here- that's for third parties, which are usually anti-human.  In one game on LLF, I caught the third party out on an innocuous post and he tried to claim a special role.  That backfired on him pretty hard, and given the general anti-town natures of third parties it's best to just claim human and not get caught out.

Quote
This is assuming that there is no human alliance. If there is a human alliance, red results should be posted immediately, green results should be posted when it is neccessairy to either safe that player from death to lynch or to lower the amount of potential wolves (while keeping in mind that there might be a master wolf).
you're absolutely right.  I consider alliances outside the normal setup, but now that you mention it, I'll add a special section on alliances.

Quote
This is extremely dependent on the game. When there is a guardian, vigi is one of the best roles to claim publicly. If the vigi has no protection, it is indeed better to stay off the wolves' radar.
I disagree.  Even when there is a guardian, the vigi will usually get more done working in the shadows.  Looking at it, I think our most recent game was a good one for the vigi to claim because of how quickly it *should* have ended, but normally you want to get more than 2 vigi shots off.

Quote
Giving an example of a specific player has no place in a general guide. This is something that should be discussed in game topics, and with that player's involvement.
I think it's a fitting example that some players "don't believe in meta", and he's more or less notorious for it.  If you're adamant about it, I can strip that part though.
Quote

Can someone make me officially TWC so I can pin this, lock some of the older topics and make some edits to the Basic Guide and Official Rules?
A while ago I sent a PM to JaMaHa asking him to TWC some combination of you, me, and BDS, but never got a response back.  But a global mod should be able to pin it as well.  Making the edits was actually an idea I had too and I meant to discuss some potential official rules changes with you to help combat inactivity
« Last Edit: August 26, 2017, 09:47:39 PM by NocturneOfShadow »
Logged
time for a splatfest! Christmas vs Halloween!

only Christmas sucks so Halloween wins by default

Dudeman

  • NSM's Least Active Mod
  • Global Moderator
  • Nintendo Nocturne
  • *
  • Gender: Male
  • Posts: 4562
  • Resident Rip Fanatic
    • View Profile
    • Email
Re: The Werewolf Game: Tactics, Strategy, and Game Theory
« Reply #3 on: August 26, 2017, 09:59:26 PM »

I stickied the topic; that's about all anyone can do without an admin.
Logged
All Hail Lkjhgfdsa_77

[close]

NocturneOfShadow

  • Nintendo Nocturne
  • ********************************
  • Gender: Male
  • Posts: 10408
  • This is Halloween
    • View Profile
Re: The Werewolf Game: Tactics, Strategy, and Game Theory
« Reply #4 on: August 26, 2017, 10:14:46 PM »

if anyone ever comes across games that give good examples of theory, PM me the link and I'll find a place for it in the guide
Logged
time for a splatfest! Christmas vs Halloween!

only Christmas sucks so Halloween wins by default

davy

  • Moderator
  • Rupee Ragtime
  • *
  • Gender: Male
  • Posts: 1136
    • View Profile
Re: The Werewolf Game: Tactics, Strategy, and Game Theory
« Reply #5 on: August 26, 2017, 10:58:02 PM »

well, on NSM yes.  We have an influx of new players, but eventually they won't be new players, and at that point specials need to know how to do their job on their own.  Since this guide aims to get players to that point, I decided to look at it that way.

That is only true for simple games. There have been multiple games on NSM where the better players needed a lot of discussion to decide on the best course of action for the specials, and sometimes, they didn't even come to the same conclusion (like TWG 81). As with normal humans, playing an avarage special is easy, but playing optimally as a special depends on a deep knowledge of not only the game in the broad sense, but also the game in question that is being played.

I disagree.  Even when there is a guardian, the vigi will usually get more done working in the shadows.  Looking at it, I think our most recent game was a good one for the vigi to claim because of how quickly it *should* have ended, but normally you want to get more than 2 vigi shots off.

We had a vigi claim in TWG 50 and TWG 40 from the top of my head. First priority in a game with a possible alliance is creating said alliance, regardless of what role the special has. Only if alliances are impossible or if there is a choice between specials claiming, the vigi could decide to stay in the shadows.

I think it's a fitting example that some players "don't believe in meta", and he's more or less notorious for it.  If you're adamant about it, I can strip that part though.

I didn't mind your FA example that much, but I still fell it should not be included. The BDS example should definately be removed, I'd say.
Logged
Quote from: NocturneOfShadow
[davy]'s in a way different time zone so basically he pops in at like 2 AM and posts 500 words and wins the game

NocturneOfShadow

  • Nintendo Nocturne
  • ********************************
  • Gender: Male
  • Posts: 10408
  • This is Halloween
    • View Profile
Re: The Werewolf Game: Tactics, Strategy, and Game Theory
« Reply #6 on: August 26, 2017, 11:09:49 PM »

I believe the vigi is the single most powerful common special role, so when people suggest claiming or not using the vigi I get a bit nervous

By saying yes the vigi can sometimes claim I think that leads to claims more often than is correct

"Almost never" is the frequency I'm sticking to
Logged
time for a splatfest! Christmas vs Halloween!

only Christmas sucks so Halloween wins by default

FireArrow

  • Nintendo Nocturne
  • ********************************
  • Gender: Male
  • Posts: 3583
  • hi
    • View Profile
    • Email
Re: The Werewolf Game: Tactics, Strategy, and Game Theory
« Reply #7 on: September 30, 2017, 08:55:47 AM »

I didnt realize i was notorious for saying that but i dont really mind one way or the other. I think something important to note about metagaming is that you can't metagame yourself as a defense ("I'm always inactive day 1 as a human!") If youre aware of a habit, then you have the power to use it to your advantage.
Logged
straight from the department of redundancy department
 

Page created in 0.089 seconds with 24 queries.