Two Plus Two Publishing LLC Two Plus Two Publishing LLC

Go Back   Two Plus Two Poker Forums > Sports and Games > Puzzles and Other Games

Puzzles and Other Games Discussions about Puzzles and other non-gambling games

Thread Tools Display Modes
Old 03-12-2009, 09:30 PM   #1
soah's Avatar
Join Date: Jun 2004
Posts: 58,651
Werewolf Strategy for Advanced Players (repost)

These threads got left on the old forum, so I am going to repost them here in their entirety to enable further discussion. Please do not discuss strategy in this thread while you are currently alive in an ongoing game. You can view the original discussion here:




Advanced Wolf Strategy

The sticky has threads for beginner wolf and villager strategy, so I decided to go a step further.

First of all, what is your goal as a wolf? For a long time I thought that the goal of a wolf is to act exactly the same as I would as a villager. I usually did a pretty good job of sounding villagery at the start of the game, but I was losing most of my wolf games because playing to sound like a villager isn't the way to win.

To win a game as a wolf you need to get X villagers lynched. This is what everything comes down to in the end. Finding X is pretty easy. First you see how many lynches the village gets - divide the total number of players by 2 and round down. Village must lynch all wolves to win, so subtract the number of wolves. The number you are left with is the number of mistakes the village can make and still win. Your target is to get one more mistake than what they can make. This is X. In a 9 player turbo, X is 3. As a general rule of thumb, a village is allowed one mislynch for each wolf that is in the game.

So why is it that simply trying to play like a villager won't work? Because playing like a villager doesn't always get X villagers lynched. Sometimes you have the power to decide whether someone lives or dies, especially when the number of players in the game is small. You can argue to save them, or you can argue to lynch them, and the village opinions are mixed enough that what you choose is what will end up happening. In situations like this, you should not save the villager simply because that's what you would do if you were a villager. Getting that villager lynched counts towards X, and that is what matters. Going back to the general rule of thumb, a village can afford one mislynch for each wolf in the game... so if each wolf is able to get two villagers lynched before getting himself lynched, the wolves will win.

Another reason that you should not simply try to mimic your villager game is that even when you are a villager, you probably get lynched somewhat often. Most people aren't that good at telling apart your wolf game from your villager game. Villages often do stupid things. Take advantage of this by playing the way the villagers expect villagers to play, even if it's not what you would do in those specific spots as a villager.

Now, this is not to say that you should try to act different than usual... you obviously need people to think you are a villager, or you will get lynched. But the point of this is that acting like a villager is a method that you use to get X villagers lynched. Acting like a villager just for the sake of acting like a villager without a plan will not go very far.

As you get deep into the game, sometimes you must stray far from your usual game. I've been burned sometimes by getting to must-lynch and then trying to pretend I'm a villager and make honest assessments about the remaining players. You can't do this. Your goal here is to get that final villager lynched. If there's a villager that's totally misguided, just give them the chance to screw it up. As long as you get that one person to screw up, it doesn't matter if the guy you're lynching plus every single dead player realizes you're a wolf. It's too late for them. It's the results that matter. Play for the results.

And how about the topic of confirmed/clear villagers? These are absolutely your enemy as a wolf. You are greatly limited in how many players you can eat at night. If you can't get rid of the clear villagers by endgame, then you will almost certainly lose. No matter how good of a job you do of getting people to think you are a villager, you can do nothing to prevent other people from being seen as being even more clear than you. Again, this ties into your goal of getting X villagers lynched. For each villager left in the game, you can eat some of them, one will be alive at the end, and the rest must get lynched for you to win.

As a wolf, you have the power to limit the number of cleared villagers there are. Your night kills are very important. Find the seer immediately. If you aren't sure about the seer, kill people who are likely to have already been peeked, people that are very unlikely to be lynched, and people whose death will seem to incriminate someone else who wasn't previously in danger of being lynched.

The night kill stuff is pretty obvious, but even more important are your actions. Some villagers will end up clearing themselves because of things they say about one of the wolves, but many more villagers are cleared by the things that wolves say about them. Part of your wolf metagame should be to include fellow wolves amongst your suspects. You don't actually want to get them lynched most of the time, but you need to make sure that after you die, the village can't look back and immediately clear all of the people that you tried to get lynched. Ideally you would like to push for people that aren't in any immediate danger of being lynched, so that it seems like a spot where you could safely push for a wolf, and people that are highly likely to get lynched, so that they will die before you do. This leaves the village with very little information after they learn your role.

And while you can clear people by gunning for them too hard, you can also clear people by insisting too heavily that they are a villager. Most wolves are aware of how much heat they will take by heavily defending a wolf, and won't do it. You are going to have to post some villager lists at some point in order to fit in and not get yourself lynched, but be careful about how you do it. Now, I'm not saying that you should definitely not single someone out as being a definite villager in a manner that will clear them - but be aware that once you do that, it will be exceedingly hard for them to be lynched. You will be stuck defending them the rest of the game, and if you die before them then your interactions will be seen as wolf-villager. So the key here is to only do that with people that you were probably going to nightkill anyway. That way you don't give up anything.

In the same way that you can clear people by having wolf-villager interactions with them, you can also set people up by faking wolf-wolf interactions. As a wolf, it is nearly always correct to virtually ignore several of the players in the game. If they are ignoring you, it's even better. If there are two villagers getting bandwagoned, don't act ambivalent. Pick one of them and make a case for them being a wolf. Ignore the other or act non-committal. After your death, your silence will incriminate that player if they are still alive.

Once you take all of this stuff into account, it's time to make a plan. I've spoken in generalities, but when you're playing in a game you have to be specific about it. Figure out what X is in your game. Figure out who each of those villagers will be that get lynched.

As a wolf, you have an enormous advantage in that you can see the future. You know that the guy gunning hard for several wolves is likely to seem extremely clear later. You know who is going to look very clear later because a bunch of wolves are pushing his wagon. Can you get those guys lynched BEFORE the wolves get lynched? Order here is vital. If getting them lynched is out of the question, then find a reason to add them to your villager list. If someone is dropping clear seer hints and you're planning to eat him, go ahead and piggyback off of his reads first (add his peeked villagers to your villager list, try to get his peeked wolf lynched). Villagers are results-oriented, and "being right" about a lot of things before it becomes public consensus will keep heat off of you for a long time. And being right about these things doesn't cost you anything because the village was going to figure it out on its own eventually anyway. You pretty much earn villager points for free.

Sometimes, seeing the future tells you that you aren't going to be able to live until the end. Once you make this realization, try to take out as many villagers as possible before you go down. Suppose that someone is about to be lynched and claims a power role. What would happen if you counterclaimed in that spot, when there is not much heat on you? You would certainly be believed and the power villager would be lynched. You have given your team an extra villager lynch for free, and given them an extra night kill to work with as well. This is worth far more to your team than staying alive for a couple more days. You may not get such an obvious opportunity as that, but remember the golden rule of getting villagers lynched. Don't jump on a fellow wolf's bandwagon unless it's very likely to clear you (this is often a transparent move), or if the wagon is likely to stall and you might help clear them after your role is revealed. (again, dangerous). The best thing to do is to just keep voting for villagers that are highly likely to be lynched. You know that people will be rereading your posts in the near future to glean information, so try to minimize how useful it is. Better yet, make it misleading. Incriminating someone so that they get lynched after you is just as good as getting them lynched ahead of you.

So, by seeing the future you can usually get a decent idea of who is getting lynched. From previous rounds, you can take a good guess at who will be lynched in the upcoming days, and you can see who that will clear and who that will incriminate. As you get closer to the end of the game, you will know how many villagers aren't clear, and how many are clear or will become clear by the end. Is the number of uncleared villagers greater than or equal to X? And will they all be lynched before the final wolf? If the answer is yes, then make sure you don't [censored] it up. Make sure you can close it out and actually get them all lynched. If one wolf is going down, he needs to make sure that he doesn't have wolf-villager interactions with one of the remaining suspects. Sometimes it is better to simply go down without a big fight if you are confident that your partner can win the game, and if by trying to save yourself you might end up incriminating your partner or clearing a villager. On the other hand, if it looks like you are going to lose the game, then you need to try something extreme. The worst that can happen is that whatever you try will fail... but if you don't try anything then you will lose anyway, so you have to try it. Do something completely over the top, either gunning for someone much much harder than it seems like you should, or emphatically stating that someone is a villager. If your antics cause any villager to change their mind about something, then you've given yourself a chance to win a game. Don't continue down a "standard" course of action that leaves you no way to win.

So, in summary, it's all about getting X villagers lynched. The goal isn't to look more villagery than everyone else, the goal is to make sure that everyone else looks more suspicious than you. If everyone seems villagery, the village will probably start lynching people that are known as being convincing wolves, which is probably bad for you unless it's your first good wolf game. Avoid nightkilling people that have a good shot at being lynched, and leave a trail of death (both day and night) that continues to incriminate more and more villagers. Some villagers are going to get themselves lynched even without your help, so as long as each wolf that goes down takes some type of action that takes out one additional villager with him, then the wolves win easily.

Advanced Seer Strategy

This also sorta covers basic seer strategy, but I'm keeping the titles the same for this series.

Upon discovering that you are a seer, the first thing that you want to do is look at the game setup and estimate when you will probably be coming out as the seer. If you come out too soon, you won't have enough useful information - the wolves will kill you and your peeks. If you wait too long, your peeks will start dying and you risk getting eaten and having the village miss some of the clues that you have left. As a general rule of thumb, the seer should expect to be coming out approximately halfway through the game in a standard setup. In special games, this might be different. If the evil faction has a seer, then they will be able to seer-hunt twice as fast and you should expect to die much earlier. If you are a high-profile player and tend to get eaten early, then you should take this into account. If you tend to get lynched a lot early, then you should take this into account.

Knowing when you expect to come out is important because it plays a huge role in determining who you should peek in the early rounds. Having your peeks stay alive is crucial. Any peek that dies before you come out is pretty much useless. Peeked villagers are crucial, and in many cases are more valuable than peeked wolves. We'll get back to this in a minute, but first I want to get this important concept out:

Peeking wolves is NOT the primary goal of the seer. Your job as a seer is to give INFORMATION to the village. If you have not read my thread on Advanced Wolf Strategy, you should check that out now, as it contains important game theory information which you should be aware of as any role. Essentially, the goal of the wolves is to get a certain number of villagers lynched. Saving villagers from being lynched is huge. If you can save enough villagers from being lynched, the wolves will wind up being lynched on their own. Peeking a wolf who was going to get lynched soon anyway is almost useless.

OK, so let's go back to the start of the game. Right now you have little or no information about anyone's role. The most important thing here is that you peek someone who is very likely to be alive when you come out. From past experience, you should have an idea of which players are rarely lynched early and which players don't get nightkilled much on the first couple nights. There are secondary considerations as well, but I will cover those later.

As you get closer to the time that you expect to come out, you will have to adjust your plans based on the events that have occured. If you have peeked villagers so far and they have stayed alive, then the best thing you can do is to continue peeking more villagers (unless of course there is only one wolf left). If you find enough villagers, you have a chance to get a lock on the game, or at least get it so close that the village has a great chance to win. Peeking a wolf at this stage will only provide an immediate benefit - but once that wolf has been lynched, it may be difficult to find the remaining wolf or wolves.

On the other hand, if you do not have the ability to get the game close to a lock, then peeking more villagers may still be good, or it may be terrible... depending on which ones you peek. If you don't have the game close to a lock, then the wolves will eat you and your peeks before endgame, at which point the village will have none of your info left. The best thing to do at this point is to peek players who are likely to get lynched at some later date, but not immediately. This gives the village a lot of info. This player has obviously had some heat on them, and people are suspicious of them. Learning their role will immediately give the village better information about the players that wanted to lynch him and which ones defended him. Had you not peeked them, they would not have gotten this information until several days later. If you peek a wolf, they get lynched sooner than planned and the village can learn a lot from their role to use in the next couple of lynches - info that they otherwise would not have had. If they are a villager, this still provides info about which people were attacking or defending him, and most importantly you force the wolves to eat him rather than having the village use one of its mislynches on him near the end.

Now as I've alluded to in the last paragraph, the information from peeks goes beyond learning someone's role. The village also gains secondary information about all of the other players by analyzing their interactions - even on villager peeks. It may seem like a great idea to peek the UTR guy who just votes once a day and does nothing else, but it might be an even better idea to peek another guy who has a lot of strong opinions about people and sits atop several player's wolf lists.

This brings us to another crucial point: It is almost never correct to come out at the start of the day. (Unless you have a lock on the game, of course.) When you do this, you deprive the village of all the information that could have come later. Generally you will want to wait until as late in the day as possible. If you have a wolf peek, just go ahead and vote for him and see what happens. Sometimes they will end up getting lynched and you might decide to wait another day to come out. If things are not going well, then come out. Wolf bandwagons provide more information to the village than nearly anything else. One of the worst things you can do is to peek a wolf that was under little heat, and then out them right away. They will be lynched but the village is left with no information. In the case of a two-wolf game, the remaining lynches would be determined practically at random, as there would be no wolf interactions to judge at all. If you don't have any wolf peeks, you can go ahead and vote for your best guess, and as long as your villagers are not in danger there is no need to come out. If the next round is potentially must-lynch, then you will generally want to come out because at must-lynch you will likely face a wolf counterclaim. But if the next round is not must-lynch, then you should simply make sure that your peeks are extremely obvious and then just pray that the wolves don't eat you at night. As long as the village will understand your peeks, then you are freerolling here.

On the subject of leaving hints, though, your clever hints are never as clear as you think they are. If you want to make sure your hints are understood properly, you need to make sure that you speak confidently about your peeks, and don't speak in absolute terms about your non-peeks. Phrases like "X is a villager" and "I'm never voting for X this game" cannot be misunderstood. There is still room for creativity. "The last wolf is one of {list}" will be understood upon death, and may be overlooked by the wolves. Making wolf peeks clear seems a bit trickier - you *have* to vote for people, even if you didn't peek them, but if you get killed the villagers will often think that you've peeked someone that you really want lynched. Fortunately though, villagers rarely have a lot of certainty about who is a wolf. Anything you say that expresses less than certainty will make it clear that you did not peek them. On the other hand, if you have peeked a wolf, don't go berserk over it. This will just draw the wrong type of attention to yourself. A simple "X is a wolf" thrown in somewhere will suffice, provided that your other posts won't cause any confusion about your peeks.

Sometimes, though, it's best not to leave any hints. As I've mentioned previously, if you don't survive far enough into the game then the wolves will be able to kill all of your villager peeks before endgame, which sucks. If your experience has been that you get eaten a lot when leaving obvious hints (as a plain villager, hopefully), then don't leave hints for a while. If the villager you've peeked hasn't done anything villagery, it might be best to wait so that you don't make yourself too obvious. As you get further into the game, you will eventually have to start making sure that your peeks are very clear, even at the risk of exposing yourself. One thing that will work to your advantage, though, is that wolves are usually scared to eat players that are leaving extremely clear hints. These hints often come from villagers who are acting as bait... and eating these players and not hitting the seer is pretty devastating as the village can conclude that all of their hints must have been correct. Eventually, though, the wolves will catch you once everyone else has been eliminated as a possibility due to being wrong or waffling on their reads.

So you can leave hints, or you can not leave hints, but absolutely the worst thing that you can do is leave hints that the wolves will understand, but the village won't. It may seem clever to disguise a peek by saying "I think that X is a villager due to [something]... I've never seen him do that as a wolf." When the village sees this, it looks like you are just making a read... which happens to look the same as the posts you made about 3 other people that you didn't peek. Another clever thing that seers like to do is make wolf and villager lists with the peeks at the top, and their reads at the bottom. Well guess what... no one knows how many of your peeks are dead, or how many villagers vs how many wolves you've peeked, or even that you've put your peeks at the tops of the lists. Remember how the wolves don't like to eat people leaving really obvious hints unless they are sure they are hitting the seer? Well that means that wolves love to eat people like you, who seem to be right about a lot of stuff but haven't left any clear clues to be interpreted. The wolves have perfect info so they know which stuff you're right about. Villagers don't. They know you must have some stuff right, but unless they know for sure which stuff it is, it's meaningless.

If things go really bad, the wolves might realize you're the seer during the day before the village has any clue at all. This is especially likely to happen if you're pushing a wolf peek too hard. This gives the wolves tons of options - there is a good chance the wolf you peeked will claim seer before you do, and possibly even get you lynched. Barring that, the other wolves may wise up to you and start throwing your peeked wolf under the bus, thereby contaminating any bandwagon analysis done later. One game I was the seer with a wolf peek and he was on the verge of getting lynched, but a villager wagon swung out in front... I got desperate and started shouting for people to switch, and the worst thing happened - the villager still got lynched, and the wolves realized I was the seer. Once I made it obvious to the wolves, it would have been better just coming out to save the villager.

Your goal is for the opposite to happen - your peeks are perfectly clear once you die, but the wolves won't realize that you're the seer. I won't lie - pulling this off is super tough. It requires a lot of creativity. But it can be done. Suppose you've peeked two villagers who both voted for X last round, and a third player voted for X and was killed at night. If you say "no wolf would ever vote for X last round", the village will be able to figure out what you meant (as long as your other posts won't be construed as seer hints)... but if the wolves are just glossing over your posts, they may not catch the significance. Tying your hints to events/posts like that, while still making your meaning clear, is likely to buy you some cover. Remember that obscuring your peeks from the wolves does not automatically translate to making your peeks ambiguous to the villagers... just do your best with it.

So in summary:
1) Peek people that will be alive when you come out
2) Peek people that give secondary info about others due to interactions
3) Wait as long as possible to come out, and don't actually come out unless you need to divert the lynch in progress or if the next round is must-lynch. Otherwise, just make your peeks completely blatant. Any chance >0% that you survive the next night is just a freeroll for you.
4) Speak of your peeks differently than your non-peeks so that there is no confusion about what you meant if you are killed at night. If you feel that you need cover, then make sure NOTHING you post seems seery which will confuse the village and waste lots of their time and energy.

Advanced Villager Strategy

As a villager, your goal is obviously to lynch all of the wolves. To get this done, you need to examine each of the other players and determine which ones should be lynched, and which ones shouldn't. The village is limited in how many players they can kill, and to win the game they will generally need about a 50% success rate on their lynches. The order of the lynches is largely irrelevant. Sometimes someone will say "We should lynch X this round, because if we're wrong we get another chance next round... but if we lynch a villager this round and lynch X next round, then we lose if we're wrong." This line of thinking is silly. If you're planning to lynch two players, then which one of them is lynched first is irrelevant. The end result is that they are both dead and everyone else is still alive.

When selecting who to lynch, you generally want to lynch the player who you feel is most likely to be a wolf. Learning a wolf's identity gives you lots of information to use to determine your future lynches. There are some exceptions though. If a player has claimed a specific role, then it is generally better to leave them alive until you are certain that they are lying. If they are telling the truth, the wolves will have to kill them at night. This takes away one kill from the wolves and gives one additional lynch to the village. Sometimes people say "Well they will die tonight anyway if they are telling the truth, so we might as well kill them now since they are probably lying anyway." This logic is terrible, because it ignores the fact that the village is limited in how many players it can lynch, and wasting one of those lynches hurts a lot. Whether the player gets lynched as a wolf immediately or a day later makes little difference, but if they are not a wolf then it is a great boost to not lynch them at all. (In special games where the wolves have their own powers, it may be correct to lynch a player who has claimed a role if you believe they are lying, since allowing them to live an extra day will let them use their powers again at night.) Another time where you might not lynch the player who has the greatest chance of being a wolf is if lynching another player could give you enough information to change your decision about which other players you would like to lynch. You should rarely, if ever, lynch a player solely for information, but when the decision is close, you should consider what you will learn from knowing their role. This is the counterpoint to the section of the Advanced Wolf Strategy where I explain that if one player could possibly become cleared for some reason at a future time, you should try to get them lynched quickly before you miss your chance. As a villager, you should take the opposite approach and choose to keep a player alive for a while longer if you might later determine that they are a villager and should not be lynched.

Of course, the big question is how to find the wolves. Well, it's not really just about finding wolves. It's really about learning everyone's role. Finding villagers helps to narrow your focus, and if you find enough villagers, then you can simply lynch all of your suspects and win the game. Learning someone's role has a lot of similarities to playing poker. No one wants to be easy to read, but the fact remains that it is impossible to play well as both a villager and a wolf without having some differences in your games. You just have to know what sort of things to look for.

Wolves differ from villagers in two key ways: They already know who all of the wolves are, and they are trying to get villagers lynched instead of wolves.

Wolf Tells Derived From Having to Think Consciously:

The subconscious mind is extremely powerful. It is impossible to turn it off, or ignore it. Have you ever looked at text and tried not to read it? It's impossible. Your brain does it for you. When villagers read posts, their brains make judgements. It's automatic. This doesn't happen for wolves, because they already know the answers. There's nothing to figure out. They have to use their conscious mind to try to determine how they would react to each post as a villager. It's tough work. And to make matters even worse, they aren't trying to make cases for why wolves are wolfy, or why villagers are villagery. They want to do the opposite. Trying to fabricate believable arguments takes time. In general, a player's wolf game is going to be several levels behind their villager game, in terms of amount of content and quality of logic. The gap is a result of the lack of subconscious help, and the need to consciously fabricate arguments.

Some potential tells include:

* A player who normally posts lots of short messages will instead post long, but infrequent analysis.

* A player posts less content than usual.

* A player asks for other people's opinions, while not stating their own. These questions sometimes appear to be rhetorical. For example: "Does Wolf X's post clear Player Y?" "Is there anything to be learned from those bandwagons?" Of course, villagers ask questions too, so don't get carried away with this tell. Look at the big picture, not a single post in isolation.

* A player posts entirely objective analysis, as opposed to subjective analysis - they will restate other people's opinions, summarize events, but they will generally avoid making judgements about it. I've named this "commentator mode" because the wolves that have done this sound like play-by-play announcers on TV that just tell you what's going on, which you already knew anyway.

* Similar to the last point, a player will come across as wishy washy because they do not make judgements about the other players. Villagers can be wishy washy too, but there are differences. For instance, villager wishy-washiness could be a result of a gut read contradicting the other "evidence" about a player. Wolf wishy-washiness is generally devoid of gut instincts and instead a result of simply not knowing how they would react as a villager. For example: "X voted for the wolf d2, but on d3 they voted to save that other wolf, so I don't know what to think of them."

* A player who is normally very excited about finding the wolves comes across as being disinterested.

* A player overexplains things. Villagers generally don't explain things unless they think it is necessary. Wolves are trying to emulate a villager's thinking and may post their thought process when there really wasn't any need to. Or, the wolf might realize that it's obvious, and actually state that it's obvious in place of posting an explanation. Wolves know that their "reads" are wrong and they are generally aware that if they were a villager, they would be reacting differently. They want to make sure that everyone is aware that they aren't just making their "reads" at random and that there is actually logic behind them. Villagers know their thoughts are pure and don't feel as much need to justify themselves. This goes along with my first point about players that post lengthy analysis which is out of character for them.

* A player's posts do not maintain internal consistency. Wolves sometimes forget about one of the phony reads that they have posted, and contradict it later. Sometimes a wolf will simply forget about one of their suspects because no one seems to be interested in lynching them. Villagers will change their mind a lot, so make sure that what you're finding are real contradictions, and not just changes of opinion. Along similar lines, if you are dealing with a player who is less experienced than yourself, you should be able to make a timeline of how their reads have changed, and then locate specific reasons for them to have changed their mind. If you can find a reason for all of their changes, there is a good chance they are a villager. If their opinion changes seem to be strongly correlated to which players have a good chance of being lynched, then there's a problem.

* A player posts things which simply makes no sense - not to be confused with logic that you personally disagree with, unless you happen to know that the other person does not believe it either.

* A player overreacts to bad news. He's actually happy to get that news, but has to pretend that he's unhappy, and doesn't do a good job of it.

* A player does not know how to react when encountering an unusual circumstance. Wolves are always asking themselves how they would be acting as a villager, and most of the time they have enough past experience to come up with the right answer. For many players, just living until endgame is a somewhat unique experience. There will also be some times where there will be a truly bizarre string of posts, or some funky roleclaims, and the wolves won't know how to react. Look for players that clam up in these spots (or players that say something really stupid). When it's must-lynch with 5 players left in a turbo, it's not uncommon for only the villagers to be talking.

* A player mostly follows the majority consensus, and rehashes the same arguments presented first by others, rarely contributing unique thoughts. Players that seem to be on a completely different page than everyone else are usually villagers, even if their thinking makes no sense (exceptions occur if the wolves are in bad shape and need to shake things up in order to have a chance at winning).

Wolf Tells Derived From Having Different Goals:

Most of the above tells relate mostly to how wolves have a hard time blending in due to the way they have to think about what they should be thinking. The other thing that makes wolves different from villagers is that they are actively trying to get villagers lynched.

* I once heard the saying that professional gamblers consider what is probable, while suckers consider what is possible. Villagers are the pros, and wolves are the suckers. Villagers have no issue with deciding that someone is a villager - they are after the truth. Wolves hate to clear villagers, because they want to lynch villagers. They will be looking for reasons to lynch someone, while discounting or ignoring competing evidence. This is often manifested in the form of quotes such as "he could definitely fake that as a wolf" rather than "he's probably a villager, although it doesn't clear him."

* Along similar lines, lesser experienced wolves often don't bother at all with identifying villagers - they are looking solely at who they can get lynched, and ignoring the extraneous stuff. As villagers, they will honestly assess people and will conclude that some of them are villagers. More experienced wolves know that they need to clear villagers in order to blend in, but they still hate to do it too much and will often fall into the traps mentioned in the previous point.

* Villagers will reevaluate constantly. Wolves go into cruise control once they have a villager lined up to be lynched.

Tells Derived From Who Knows How They Will Look Tomorrow:

My first group of tells was derived from the fact that the wolves know who all the wolves are because it causes wolves to use their brains differently, but there is another group of tells which come as a result of the wolves using this knowledge to try to avoid making themselves look any worse than they have to while getting villagers lynched. These types of tells rely on the fact that wolves are not going to take an action which increases their chances of getting lynched unless they have something to gain from it as well. Going into specifics is somewhat futile because what villagers consider wolfy is constantly changing over time, and wolves will continue to adapt to this to avoid being lynched. However, there are some general tendencies that will generally hold true unless all of the players in the game are very experienced and aware of all of these tells. Just like using tells in poker, you need to know if your target is aware of these tells and could be reversing them on you.

* Wolves and villagers both want to get "credit" for driving the lynch of a wolf... but villagers can't ever be sure that they are right. Anyone that seems to be trying hard to take credit for a villager who gets lynched is probably a villager. Wolves will put in the minimum amount of effort necessary to get a villager lynched, because they don't want "credit" for it. For the next month after I've posted this essay, this tell will likely be meaningless, since it is so easy to fake.

* Conversely, if a wolf gets lynched, any fellow wolves on the bandwagon want to make sure they get their credit for it, and are unlikely to post last-second messages expressing doubt about their vote.

* If the two leading candidates at the end of the day are both villagers, wolves are more likely than villagers to leave their vote on a meaningless third party. This doesn't apply to people that aren't around near the deadline and left their vote on a reasonable candidate at the time of their final post.

* If one of the two leading candidates is a wolf, the fellow wolves who are around at the end will generally either try to save him, or pile on him to look good, but are unlikely to make a throwaway vote at the end.

* When a wolf is very obviously getting lynched, his fellow wolves are unlikely to continue defending him once things appear hopeless. People that continue to insist he is a villager or make absolutely futile attempts to save him are villagers.

* If both of the leading bandwagons are wolves, the fellow wolves will almost certainly pile onto one of them to make themselves look good in the short term, until it is revealed that the other wagon was also a wolf. It would be very strange for a wolf to make a throwaway vote here.

* Wolves are unlikely to assume a leadership role and try hard to get people to follow their vote. This always has a bad outcome: either they lead a villager lynch and take the blame for it the next day, or they lead the lynch of a wolf who might not have been lynched had they not pushed so hard. Wolves would much prefer to work outside of the spotlight, and have villagers lead lynches of villagers.

* Somewhat related to this is that wolves are unlikely to argue with each other at great length. People get into arguments in order to get the other person to change their mind... why bother getting into an argument where both people already know who is right?

Tone Reads, Player-Specific Tells, and Miscellaneous:

There really is no such thing as a generic villager tone. Many players consider it villagery to post sound, logical arguments and wolfy to post emotional, strange arguments. This is often backwards. Clearing anyone based on their tone without solid previous history with them is certainly a mistake. While there may not be such a thing as a generic villager tone, there are some tone things that can be associated pretty closely with wolves. The biggest one is unprovoked, random hostility. Villagers tend to see the game as a puzzle, while wolves are more likely to take the "us vs them" mentality as if there were some sort of heated rivalry. Naturally, villagers will react if you push their buttons, but people that seem to lash out for no reason are often wolves. Very similar to this is the use of insinuation and innuendo in place of logical attack, which comes off as being pretty aggressive. Sometimes, wolves will mistakenly show their happiness when things are going well for them. This is not to be confused with sarcastic type of remarks... but if someone seems genuinely happy when the village is in bad shape, you should be wondering why they are happy. And as a general rule of thumb, most players are more serious as wolves than they are as villagers.

For the most part though, analyzing tone comes down to comparing how a player acts from one game to the next. Three big things to look for are level of seriousness, level of interest, and level of confidence. When someone has played several games in a row claiming to have minimal reads and then all the sudden the next game he acts like he has the game cracked wide open halfway through day 1, he's a wolf. But for other players, that's their standard villager game.

Somewhat different than player-specific tone are player-specific tells. Just like tells in poker, player-specific tells are almost always generic tells which apply very strongly to specific people. Seeing someone eat an oreo when he is or isn't bluffing is useless. If you figure out that he's eating an oreo because he's calm, then you've got something solid to work with, and in the future you can associate that tell with other things that he does when he's calm. In werewolf, people are going to do things that give away their role, and they are likely to continue repeating the same mistakes in future games. If their actions are the result of some sort of psychological drive that you can identify, then you have a solid tell. If your tell is something random, then there is a good chance it is just a coincidence and it's not reliable.

And finally, there are still some other wolf tells which I haven't really covered, however, I believe that most of them are subsets of the things that I've already gone over. Trying to enumerate everything that I find wolfy would be a futile task, as much of it depends heavily on timing and context. At this time, there are only two other things that I'd like to add which don't seem to fit in well with the other sections. The first is that wolves will often make excuses for themselves preemtively. Villagers tend not to defend themselves until they are actually under attack. The second thing is that villagers tend to be more honest about posting things that seem wolfy on the surface. For example, a villager will generally be willing to admit that they really had no reason for their vote switch, when a wolf would invent a reason for it.

So there you have it.
soah is offline   Reply With Quote
Old 03-12-2009, 10:30 PM   #2
Carpal \'Tunnel
sopoRific's Avatar
Join Date: Sep 2005
Location: Corrupting Thetans
Posts: 7,224
Re: Werewolf Strategy for Advanced Players (repost)

Is it okay to post "what's my play?" questions about specific spots in this thread?

I was in an odd spot in a turbo a few weeks ago where I wasn't sure what to do, and it's slightly bothered me since. I was the seer, and soah was my n0 random peek on this turbo. Soah's wagon starts around post 76 and day ends at post 125.

Near the end of the day, soah is convinced I'm a wolf and passionately trying to wagon me. However, others are confusing his passion for wolfiness and are voting for him. Still others are apathetic about the exchange and are voting for others.

Being both inexperienced and someone who seems to naturally draw attention to myself as a wolf early in games (though this latter part could be selective thinking), I am rarely a target for the night kill and there should be a good chance I can last a night or two of seer hunting.

The votes stand 2 votes for me, 2 for soah, and a few other people with 1 vote. What is the correct play?
  1. Vote for someone with a single vote and hope to win the 3-way tie. This is what I ended up doing.
  2. Vote for soah to save myself, and either lie about my n0 peek or try to explain why I voted soah when I knew he was a villager when I claim.
  3. Claim seer and hope something good happens when the votes on soah and I scatter.

Nothing seems really good here. At the time, I figured I would be good for two more peeks, and if I claimed day 1 and the villaged lynched one of its own, me and soah would be unceremoniously killed over the next two nights, and I thought that it wouldn't help the village as much as getting an extra peek or two would.

Lynching soah and lying about my n0 peek at least sounds fancier. I didn't think about the lying part during the game.
sopoRific is offline   Reply With Quote
Old 03-12-2009, 10:41 PM   #3
Sun Tzu's Avatar
Join Date: Jun 2008
Location: the bog
Posts: 73,655
Re: Werewolf Strategy for Advanced Players (repost)

Lynch soah and let leaving one less peek than expected help you stay undetected.

Letting it flip is risky.
Sun Tzu is offline   Reply With Quote
Old 03-13-2009, 09:41 AM   #4
Carpal \'Tunnel
BombayBadboy's Avatar
Join Date: Oct 2006
Location: CRAI me a river.
Posts: 10,125
Re: Werewolf Strategy for Advanced Players (repost)

"No one wants to be easy to read, but the fact remains that it is impossible to play well as both a villager and a wolf without having some differences in your games."

Im beginning to think that balancing is extremely important. As a villa my game was wide open, basically just posting whatever came on my (subconcious?) mind. I still was called wolfy for it, though. As a villa I was very aggressive in pursuing some players, which is alot harder to do for me as a wolf. I think I should balance by tweaking my villa game, not my wolf game. Thoughts?

Bonusquestion: is it possible to lynch a wolf day1 when the wolves dont want you to?
BombayBadboy is offline   Reply With Quote
Old 03-13-2009, 09:48 AM   #5
Zurvan's Avatar
Join Date: Apr 2005
Location: On the front porch, yelling at kids
Posts: 32,281
Re: Werewolf Strategy for Advanced Players (repost)

I think I should balance by tweaking my villa game, not my wolf game. Thoughts?
Make your wolf game better, don't make your village game worse. You're a villager 3x as often as you're a wolf.
Bonusquestion: is it possible to lynch a wolf day1 when the wolves dont want you to?

OK, so that first sentence isn't very illuminating. Figure out what makes someone villagery. Do that as wolf AND villager. Profit.
Zurvan is offline   Reply With Quote
Old 03-13-2009, 09:51 AM   #6
Carpal \'Tunnel
Join Date: Aug 2005
Posts: 36,469
Re: Werewolf Strategy for Advanced Players (repost)

Optimal way to play to maximize wins for both teams is to play the best villager game you can at all times, and as a wolf to emulate your best villager game to the best of your abilities (or, if you see an opportunity to do something else worth possibly seeming wolfy but will gain your team an advantage, go ahead and take that)

Never dilute your village game imo

for the bonus question, I'm going to say if the wolves really don't want you to lynch a wolf day1 they can pretty much always present it from happening if they want such that they all band together and vote for one person

that may not be very good for the wolf team though. So to answer the question, if the wolves thought it was the end of the world if someone was lynched, they could definitely prevent that from happening d1 almost certainly. But they'd look so bad that they don't so lynching wolves d1 is definitely possible even if not 'planned' by wolves
LuckayLuck is offline   Reply With Quote
Old 03-13-2009, 11:41 AM   #7
Carpal \'Tunnel
gusmahler's Avatar
Join Date: Jul 2005
Location: Creepin'
Posts: 27,131
Re: Werewolf Strategy for Advanced Players (repost)

Originally Posted by LuckayLuck View Post
that may not be very good for the wolf team though. So to answer the question, if the wolves thought it was the end of the world if someone was lynched, they could definitely prevent that from happening d1 almost certainly. But they'd look so bad that they don't so lynching wolves d1 is definitely possible even if not 'planned' by wolves
In Dr. Horrible, the most powerful wolf was lynched on D1. But it also "cleared" two wolves who pushed his wagon.
gusmahler is offline   Reply With Quote
Old 03-13-2009, 12:40 PM   #8
Sun Tzu's Avatar
Join Date: Jun 2008
Location: the bog
Posts: 73,655
Originally Posted by gusmahler View Post
In Dr. Horrible, the most powerful wolf was lynched on D1. But it also "cleared" two wolves who pushed his wagon.
Clearing them for that was pretty bad, especially since they were both reluctant.
Sun Tzu is offline   Reply With Quote
Old 03-13-2009, 12:44 PM   #9
Sun Tzu's Avatar
Join Date: Jun 2008
Location: the bog
Posts: 73,655
Originally Posted by Zurvan View Post
Make your wolf game better, don't make your village game worse. You're a villager 3x as often as you're a wolf.
Originally Posted by LuckayLuck View Post
Optimal way to play to maximize wins for both teams is to play the best villager game you can at all times, and as a wolf to emulate your best villager game to the best of your abilities (or, if you see an opportunity to do something else worth possibly seeming wolfy but will gain your team an advantage, go ahead and take that)

Never dilute your village game imo

You have to base your wolf game on your village game but your village game shouldn't even take your wolf game into account.
Sun Tzu is offline   Reply With Quote
Old 03-13-2009, 05:00 PM   #10
Hitch's Avatar
Join Date: Aug 2007
Location: Delivering
Posts: 18,023
Re: Werewolf Strategy for Advanced Players (repost)

you should play every game as if it were your last game imo.

do everything you can to maximize your shot at winning the current game, everything else takes care of itself.
Hitch is offline   Reply With Quote
Old 03-13-2009, 05:06 PM   #11
Carpal \'Tunnel
DustinG's Avatar
Join Date: Nov 2004
Location: bubbling
Posts: 29,998
Re: Werewolf Strategy for Advanced Players (repost)

eventually you'll start to just care less and metagame will take care of it's self

if you play enough you'll have lazy village games, lazy wolf games, super villagery active village games, and super villagerly active wolf games

the key is making people believe that you're capable of anything
DustinG is offline   Reply With Quote
Old 03-13-2009, 05:18 PM   #12
Carpal \'Tunnel
DustinG's Avatar
Join Date: Nov 2004
Location: bubbling
Posts: 29,998
Re: Werewolf Strategy for Advanced Players (repost)


for real FPS awesomeness. You should take someone you think is a wolf, and claim you peeked that person as a villager. In that spot that Soah wagon should cease since if you are a wolf faking there then Soah would have to be a villager, and thus a third person would get wagoned.

If you are correct about that person being a wolf, then the wolves would have to put you on faking seer as the seer, which isn't going to happen.

In fact, just pick your best guess for a wolf and claim you peeked him wolf.

If he reacts in a super vilagery way then you should unclaim and now claim soah as your village peek all along, wolves will be so confused they'll never eat you and it would take a retarded village to lynch you there.

Last edited by DustinG; 03-13-2009 at 05:39 PM.
DustinG is offline   Reply With Quote
Old 04-08-2009, 12:10 AM   #13
cake eater
metsandfinsfan's Avatar
Join Date: Jan 2006
Location: waiting for well named
Posts: 143,713
Re: Werewolf Strategy for Advanced Players (repost)

id like to discuss whether villagers should leave fake peeks day 1 to try to get nked.

i was discussing it with dangeraw over PM but would rather discuss it here

in general, i hate fakeseering and fps but i think saying "xx is a villager" is a provillage thing to do day 1 as a villager

metsandfinsfan is offline   Reply With Quote
Old 04-08-2009, 12:12 AM   #14
soah's Avatar
Join Date: Jun 2004
Posts: 58,651
Re: Werewolf Strategy for Advanced Players (repost)

yeah that should be pretty much automatic
soah is offline   Reply With Quote
Old 04-08-2009, 12:15 AM   #15
chuckleslovakian's Avatar
Join Date: Aug 2005
Location: Saving Sinners
Posts: 30,425
Re: Werewolf Strategy for Advanced Players (repost)

I believe I was the first person to make a long post on why I thought that was a good idea
chuckleslovakian is offline   Reply With Quote


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

All times are GMT -4. The time now is 11:40 AM.

Powered by vBulletin®
Copyright ©2000 - 2016, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO 3.6.0 ©2011, Crawlability, Inc.
Copyright 2008-2010, Two Plus Two Interactive
Poker Players - Streaming Live Online