I'd like to run a more mature themed campaign, along the lines of game of thrones. In that story, several of the main characters have an evil alignment.
I'd like to do something similar to this in my campaign, but I'd like to know how to keep the good and evil PCs from constantly trying to kill each other? What techniques can I use to have them work together? What techniques can I use to have them work against each other without fatal consequences?
The characters in Game of Thrones are almost as complex as real people. Real people cannot be put in one of nine little boxes and call it done. Alignment in general is extremely problematic for a lot of games, but this one especially so. It’s just far too simplistic to handle a “mature” game with any depth. Remember, the system was designed for straight dungeon-delving with little thought for more complicated campaigns. As such, you should not be labeling characters with cute little two-word descriptions like “Lawful Good” or “Chaotic Evil.” You’ll pretty much automatically fail to achieve a complex world full of realistic people with deep motivations and reasons for their actions if you do that.
If you are running this in a system that, like Dungeons & Dragons, has mechanical alignment, you’ll need to handle that. I’ve discussed doing that a few times; this answer sums up the options I see available pretty nicely (and also goes into some more discussion of why alignment is limited).
That said, some thoughts on how those who care about morals can get along with those who don’t, or even those who actively flaunt them for the sake of flaunting them.
The classic reason: both Good and Evil face some serious external threat that they must face together. Plenty of Evil has no interest in seeing everything destroyed, or, for that matter, in seeing everything dominated by some other Evil force. Dungeons & Dragons has several examples of these, such as the Blood War (archons, creatures of pure lawful-good-ness, have at times fought alongside the lawful-evil devils against the chaotic-evil demons), and then when the illithids suddenly appeared (which caused the Blood War to stop and potentially had celestials, demons, and devils all fighting together, though that did devolve into bloodshed among the allies).
The Good guys tolerate the Evil ones on the basis that they can redeem the Evil ones. The Evil ones hang around because the Good guys’ quests (and their rewards) are valuable to the Evil characters. Order of the Stick uses this heavily for Belkar, who is not at all picky about who he cuts up, and the Order provides him with plenty of relatively consequence-free victims. The rest of the group is aware of this but they have some hope (not unfounded, either) of reforming him, or at least preventing him from getting worse.
An Evil character isn’t necessarily a sociopath with no regard for anyone but himself. He may have values and friendships and relationships. He may be Evil because of his methods and his intentions, but that doesn’t necessarily mean he’re a bad guy. This is especially true in the case of simplistic D&D alignment, where you can ding “Evil” without doing anything that strikes a lot of people as horribly evil.
Have you ever read Order of the Stick? I see that KRyan's used it in his answer, but it's a good example. However, in my games, I've used three different methods that work pretty well.
Don't get me wrong, I'm all for letting players choose their character roles and archetypes, in fact I think it's the best thing you can do to facilitate a good gaming experience. However, characters are a product of group consent even if they are individually created by each player. If the party has characters who absolutely will fight all the time without stopping, you need to just let them work out the violence. This requires players who are mature enough to be allowed to do such a thing, but I've never seen this turn into too much of a problem, both as a GM and a player in this situation.
Basically, let the players work out the direction in which they want to take the group by allowing them to fight and imposing consequences on individuals or the group as a whole as a result of the repercussions of their actions. This doesn't "answer" your question, but I think that you may be more concerned about this than is necessary.
"Everyone is the good guy in their own head." Consider that when you're dealing with an evil character in your games. If you're looking at a truly psychopathic chaotic evil murder machine, you may actually want to forbid the character because they're a disruption to play. However, I've found that there can be groups with alternatively aligned political extremists; any Shadowrun game I've run has an example of characters who clash from a variety of sources, but the truth is that at the end of the day they're working together for a common goal. Even if it's personal gain, they've realized that their best opportunity is to just work together in order to keep the group intact so that they can get an optimal return on the profits.
Another thing to consider is that the betrayal underlying the group's foundations can actually be a bond; they have to give each other signs of trust, and should both sides prove to be responsible they will then have a character relationship that transcends their alignments.
Running with the last concept, consider whether characters are supposed to be "evil" or just have no qualms with doing harm. For instance, Rorschach from Watchmen. Many characters are willing to do horrible things either for a perception of greater good or just an apathy for the general condition of mankind, but will not actually cause harm unprovoked. For instance, these characters will take out a guard by slitting his throat instead of knocking him unconscious (the medically dubious nature of reliably rendering someone unconscious aside), on account of the fact that it's a more certain solution. While we're on the topic of comics, The Punisher presents a protagonist who is very good at this. The secret to creating an evil aligned character who is more than just an alignment label is that they are opposed to something, not that they're opposed to everything.
Encourage players to break alignment. It's very hard to have a good, deep character using some of the alignments because they just aren't built to create deep characters; they're meant to create easy to roleplay characters. They're a general description, not the list of personal values and motivations.
List motivations. Eclipse Phase integrates this very well; a character may be +Anarchism and +Populism, and -Violence. Evil characters should have positive motivations, and shouldn't just be +Violence, +"Bad Things", -"Good Things". If your system has an Edge/Action Point/Moxie/Bennie system use the motivations to determine how they are earned.
Encourage evil to manifest in ways other than violence. It's okay to have a rapacious magnificent scheming evil villain as a character, but it's typically not a good idea to have evil characters who look more like the Psycho from Borderlands 2 than Hannibal Lector. Likewise, not every Malkavian is going to check the flavor of the nearest wall every fifteen seconds.
Good and evil is not always clear. Sometimes the good guys manage to cause massive harm by doing what they think is right, and sometimes those who don't care about the impacts their actions may have on others don't actually wind up harming those around them. Good guys who let a villain go may wind up allowing them to cause massive destruction, and bad guys who take down a supposed paragon of good may discover that their target was just as flawed as they are.
Give players obligations and encourage them to share character histories; they should have reasons not to attack each other if they're traveling partners, business partners, or the like. Their honor or reputation may be on the line, and it may be impractical for the good guys to take down the evil guys in the party just because then they wouldn't be able to keep doing as much good as they do.
Encourage evil characters to be passive. It's not necessary for evil characters to always run in and smash all the good characters' work "for the 'lol's", instead it should be a premeditated, scheming, "What do I get?" situation where the evil characters work with the good guys for the wrong reasons but wind up doing good. Jayne from Firefly looks like this at times, even though he's probably not.
Encourage good characters to be passive. Sure they're with a band of ravaging, bloodthirsty psychopaths, but so is everyone else. They can't change the way the world works, and if you're in a Game of Thrones style world then they'll be a solitary glimpse of light amidst a dark abyss of moral decay. They have to pick their battles, and this isn't it.
Besides of the ingame mechanisms that spring to mind, it actually helps to discuss expectations about your game in your group. After some of our campaigns failed miserably due to in-party conflicts, we talked about what we wanted in games – and subsequently all had more fun.
It's just a simple truth that different campaigns work better with different styles of gaming. Some might be build around players competing, while others might implode due to it. You can still play both with the same group by getting everyone on the same page: A while ago I discovered The Same Page Tool, which just might solve your problem in this case:
The Same Page Tool facilitates finding a consensus on what sort of behaviour is expected of each role in the game. It includes questions such as "Are the players supposed to work together?", "What is the stance on rules?", "How much freedom do people have in describing their surroundings?" etc. The important thing is that it is not a questionnaire to be filled out by each member, but something that is filled out together in the group.
DnD Good and Evil Don't Have To Be Enemies At All
The first thing to note is that there is DnD is not always about "Good Vs. Evil". Even in its extremely simplistic alignment system it is more complicated than that. Evil fights Evil all the time, Good can certainly fight good. This is especially true on the scale of national politics.
Imagine that both Kingdom A and Kingdom B are Good (in the Dnd alignment sense) and prefer to be peaceful. But imagine they both get their water from the same river and a famine causes a water shortage and people will die of thirst unavoidably and crops are failing. (Yes, there are magical solutions in DnD, but lets say we're in a setting where that doesn't work). They might start out looking for other sources of water and even cooperate, but if the famine keeps getting worse and they find no other solution there is a good chance a war will occurr eventually even though they are both "Good".
On the other hand, if they suddenly find Kingdom C with bountiful water and food, they may both trade with it even if C is "Evil". This is especially true if C's Evil is mostly in how it deals with its own citizens and it plays the part of a good trading partner.
A Common Cause
The best way to get people to work together is to give them a common cause. As KRyan mentioned, the most common source of a common cause is a common enemy. In other words, people will set aside their own differences to defeat a powerful enemy they both hate.
But this can be broader than that, any goal they are both interested in will do. Perhaps instead of a common enemy it is a common threat defined more broadly, such as a plague. Even the Evil person wants the plague cured because it is killing his people too.
Or perhaps they all want to build something. They might have different motivations, but they still have a common cause. Maybe an Evil person wants a bridge so he can make money from the people on the other side. The Good person wants the bridge so he can bring needed medicine to those same people. They both want the bridge, they have something in common to work towards.
Limiting the Conflict - A Higher Power
If you want to let them have conflict non-fatally, you can make killing against the law, literally. Historically, court intrigues happened because it was generally illegal to kill a noble, and that law was actually enforced. This didn't stop all murders of nobles of course, but it prevented most overt violence between them. It did not stop them from bickering and interfering with each other's business in non-lethal ways (or in ways that were highly lethal for the commoners in their employ). If your characters are less than Sovereigns you can easily have the Sovereign mete out punishment for killing each other and leave them to find more creative ways to hurt each other.
Even if the characters are sovereigns you can easily have consequences for killing each other. Perhaps the international community shuns those who go around killing the wrong people. Perhaps deities themselves get involved when the wrong people are murdered and this is common knowledge in your setting.
In short, if you want to keep it non-lethal, add in severe consequences for anyone that uses lethal force.
So, it's important to distinguish between these two goals (both of which are mentioned in the question):
This is when you want the PCs to form a stable party, or otherwise consistently be on the same side.
At its heart, this is pretty simple: think about why real people work together. Now make that a part of the characters. Most simply, you need to give the protagonists shared goals and a sincere mutual affinity for each other. If they're selfish bastards, that friendship may not go as far, but complementary goals will still move them to help each other and stay together.
Note that personal goals that line up well are usually better than a single overriding purpose. "We all want to defeat the evil overlord" might work in a pinch, but "I want to usurp my brother's throne and the best way to get there is to help my friend seize control of the assassin's guild" is much, much stronger.
Essentially, you can adapt the setup for "sticky antagonists" to make the PCs each others' rivals:
The conflict: why are they actually at odds? These can be outright inimical or goals that almost fit together but not quite.
Attraction: a reason they're continually interacting. You don't usually pal around with your enemies just because, after all. They have to want the same things, or know the same people, or otherwise be pulled towards the same situations consistently.
The foil: a significant personal cost connected to destroying your enemy or otherwise severing the relationship. You can't exactly just murder your rival at court, for example. You'll probably be executed or exiled. What good will that do? Ditto it's hard for a non-sociopath to kill his own brother even if they are outright enemies.
In general, this is easiest to do in a game where the murder isn't the default way of handling challenges and conflicts. Note that the "foil" isn't absolute. Part of the fun and tension of this setup is that the option for a (probably bloody) final resolution is nearly always on the table.
A lot of the groups in Game of Thrones are a mix of these two — consider Jamie and Brienne of Tarth as an example of personal goals temporary bringing rivals into cooperation. Note that Game of Thrones isn't the best inspirational material in other ways, however: the series is more focused on changing loyalties and bloody betrayals than enduring friendships or even stable rivalries.
Do you want to play a game about complex people with a mix of admirable qualities and terrible flaws — what some people might call "shades of grey?"
Throw out alignment altogether. You don't need it! Focus on giving the characters coherent individual motivations, as described above.
Do you want to play a game where Good and Evil are palpable cosmic forces, like in D&D, but you explore what it means to be Evil?
Focus on the fact that alignment is about metaphysics, not personality. D&D-style Good and Evil might imply something about how a person acts, but, in truth, they say much more about a character's relationship to gods, magic, and the afterlife than they do about a character's relationships with other people. The basis of D&D alignment is that actions have spiritual consequences, moral codes have magical power, the ends don't justify the means, et cetera, et cetera. So embrace that. Take the moral calculus of the universe and throw it in the PCs' faces: "You're damned — now what?"
I had a pretty successful D&D campaign based on the idea that Evil is Evil but some things were just worth selling your soul for.
Characters like this within a mixed-alignment party are often the most tragic and driven among the PCs. They can work well as "devil's advocates" bringing the other characters' morality into focus. The trick is to interrogate their choices rather than forcing conflict by going behind their backs to do something the rest of the group doesn't want to do.
Do you want to play an "Evil" game in the sense that you want the protagonists to do vile and horrible things?
Okay, the most important part is to be very clear about how vile the group is willing to go. A very useful concept here is "lines and veils" (I'm going to quote from someone's game setup that I found on the Internet, since that seems more useful than abstract academic discussion):
Line: Something that will never be hinted at or brought up in-game. Something that is off limits.
Veil: Something that can happen in a game, but when it does, we skip the details and fade to black. We "veil" it.
[...] The point of this is, during play, just yell it out. If you have a problem with something that is taking place, please just bring it up and call it either a Line or a Veil. Or don't worry about the buzz words if you don't want, we can figure that out together. Just point out that you're uncomfortable and we'll fix it...
This is supposed to be fun, not uncomfortable. But since this game has a lot of shock value, it can easily get that way.
Note that a lot of things that you might tolerate in the source material — like the rape and torture in Game of Thrones, for example — might be way, way more uncomfortable when you're actually playing them at the table. Tread carefully and respect each other's feelings and well-being.
Also, I really don't recommend vile for vileness' sake. It just turns the game into an oddball farce. For this kind of "mature" game to strike the right note, you need a sense of purpose behind the horrible stuff.
It's sorta possible to shoehorn a single character like this into an otherwise non-psychopathic party, but exceedingly difficult. A good model is Lord Bolton (NOT his son!) from Game of Thrones. The man is a sadist through and through — he even claims it as his cultural birthright — but he knows exactly how much he can get away with, always flying under the radar and only letting out his worst self when he knows his victims are powerless. The Starks absolutely know that Bolton is a monster but they tolerate him anyway because he comes off as their "Poisonous Friend."
Honestly, I've got to echo the sentiments others have given that the concept of alignment doesn't fit very well into A Song of Ice and Fire. It would be best to dispense with alignment entirely.
If you really want to continue down this road, though, then remember that Evil characters are not necessarily psychotic or stupid. They do form genuine friendships from time to time -people they wouldn't sacrifice at the drop of a hat, or even at all- and if this happens, then as long as you don't put too much pressure on the Evil characters to kill the others, they won't.
Even if that doesn't happen, though, Evil characters generally will still work with others as long as it makes sense to do so. If the Evil characters can see the other characters as valued assets, that's not a strong bond like real friendship, but it'll get the job done. Make sure that every character, Evil and otherwise, has a chance to really shine, because this will reinforce that opinion.
And lastly, make contingency plans. Sometimes PCs kill each other, especially when at least one of them is Evil. Obviously you shouldn't hope for this, but you should make plans for what to do if each one of them dies, so that it's not the end of the world.
You don't say how experienced your players are, but if you're looking for a 'more mature themed campaign' presumably they can distinguish what their characters want from what would make an interesting game. Why not have a bull session where anybody can put up ideas? You then have a skeleton of basic ideas that at least one person likes and everybody else can live with, and can put your own flesh onto it (so to speak).
For example; the characters are random passers-by who are accidentally witnesses to an atrocity. One may wish to bring the perpetrator to justice; another to blackmail him; a third to join him, thinking that provable discretion and lack of scruples will bring rapid promotion. But; the party could only identify the soldiers, not who they were working for, while the soldiers did see how many witnesses there were as they fled. Blackmail is pointless unless you can prove all the witnesses are in on it, and judges tend to frown on prosecution witnesses who murder other witnesses; so either way, no quiet assassinations. And they literally dare not let each other out of sight; if one of the others cuts a private deal before you can, it will involve a messy death for you. Meanwhile, it is quite important that they discover who exactly was responsible, which will neeed co-operation.
(That paragraph is not 'my answer'. I know the idea is overused and incomplete, but I myself would quite like to run a character in the situation. If two other people make comparable suggestions, the GM will have the basis of a good campaign in which three people already have an interest.)
Once, one of my friends tries to run an evil campaign where he complately lost control of the game...
Yeap, I was the players who got killed with torture and rape... I was a male, like the rest of the group.
Basic problem is D&D logic. If you are a good guy, then you are a good guy. If you are evil, then you are. Players evaluates D&D alignmets as strict lines (even the rules insist that they are not). That means, an evil guy feels himself that he can do anything because he is evil. They will not hesitate to kill an NPC that they need to talk or ignore a quest thay had took. So keeping the game control is really hard. Be sure, many things will happen beyound your guess. Because (as I said), being evil make players think thay can do anything.
As an example, I can summarize the game; We talked with the city mayor and got a quest, then go the to first person that is related to the quest and kill kim. After haf an hour, we were escaping from the city after raped some girls and burned down a part of the city. Then other player who burned the city with me killed my character since he do not want any wittness who saw his face. DM thought he got everything ready to keep his campaign control but campaign ended within few hours
So first of all, it is really hard to keep control of the game, bacause players will not act like villain NPCs that you play.
Second, you must take care of problem classes. Paladins and most of the good clerics do not get along with evil ones according to their ethics and ethos.
Fitting a Game of Thrones campaign within D&D rules is harder. Because you probably will label most families as not good and nearly everybody as not lawful but in fact those poeple are good and lawful in GoT aspect. They are not evil according to medival age conditions of thereal world (where we can say GoT is quite close to Medival Age). Fitting too many treson and opprtunism is not D&D style.
GoT folk who looks evil but in fact were good people probably be marked as evil and chaotic according to D&D logic.
If you want it to be D&D mechanics, then you must dismiss alignment, along with all alignment base abilities and spells and similar things from core mechanics.
