The Ten Commandments of Gamemastering

Yes, I am a giant geek. Deal.

I’ve been running role playing games for my friends for about 22 years or so at this point. I have created and run over a dozen campaigns and innumerable one-shot adventures. I’ve played RPGs with rank newbies and grizzled veterans and everybody in-between. I’ve lost count of the number of game-systems knocking around in my head (and, indeed, as any of my long-time players can attest, I sometimes get them all tangled up together.). What I lay out here are the strictures by which I try to run what I consider to be a good game. They may work for you or not; all I’m expressing is my experience. I might also add that I don’t always live up to these commandments myself, as much as I try. When they all work, though, something golden is bound to happen.

 

#1: Thou Shalt Not Take Thy Game Too Seriously

Yes, I realize you spent hours and hours prepping this adventure. Yes, I know your players are being paranoid weirdoes and metagaming this thing to hell and back. You are absolutely right that your players aren’t taking the Black Dragon of Immortal Dread seriously. I get it. Your players are being dicks. Now, pay attention: Get Over Yourself. Repeat back to me: This is a game. It is not all that important. I will not pout or complain or storm off or exact my vengeance on my friends in the form of falling rocks. The idea in an RPG, hell, the idea in every single game, is for people to have fun. Are your friends having fun? Yes? Then shut up. Roll with it. Improvise.

#2: Thou Shalt Clearly Explain The Game’s Goals

If these dudes show up to your ‘historically accurate barbarian campaign’, it’s your own damned fault.

The easiest way for players to ‘ruin’ your game (and please keep in mind that RPGs can’t actually be ruined if your go with it) is for them to have no idea what you want out of the game. Your fun is every bit as important as their fun and, if they are your friends, they should care about that. So explain to them in clear terms what you want the game to be. “This is a dark, gritty crime thriller game set in a dystopian future” or “This is a horror game where you guys are monster slayers” or “This game is supposed to be high adventure space opera with laser beams and talking robots and everything.”

This gives your players the choice: do you want to play this game that I have described, or do you not? If they don’t, no harm/no foul. If they do, they should be willing to buy into the whole thing. If you and the players work together, things will get awesome fast. If you and the players aren’t on the same page, it isn’t going to work.

 

#3: Thou Shalt Hold Story Above All Else

RPGs are not strategy games. They aren’t card games. They are Role Playing Games, which means people play roles (characters) and you, as Gamemaster, put them in situations where they can play those roles. They are, in essence, highly collaborative storytelling games. As such, the story needs to get top billing. It isn’t just that the players are doing a certain thing, there needs to be some understanding of why they are doing this thing and what the stakes are and what happens next if they succeed and so on and so forth. Now, this doesn’t mean a fair amount of strategy can’t be involved (particularly if the game’s concept calls for it), but any RPG that degenerates to mere dice rolling and accounting is going to be dull. Victory is far more sweet when it means something, and it won’t mean anything without a story attached to it.

Yes, even the dwarf samurai.

#4: Thou Shalt Find Every Player Character Cool

Yes, all of them. Even the ones that are very much not cool, like the pacifist half-orc accountant or that one guy who wants to play a wandering poet named Dweeber who keeps a dead fairy in a pouch in hopes it will produce pixie dust someday. It doesn’t matter if the player has come up with a character you think is the dumbest, least interesting, least appropriate or useful character ever devised by man, it is your requirement, as GM, to love him and try and make him look cool.

Why is this your job? Because the GM is supposed to make the adventure fun, and a player who finds his character useless and ignored (no matter how rightly) isn’t going to have much fun. Now, making them work may take some doing here, but try and figure out what circumstances will make that character shine and make a point to give them that opportunity as often as is feasible within the bounds of the story. Again, if you’re obeying Commandment #2, ideally this shouldn’t be much of a problem very often, but stranger stuff can happen. Anyway, if somebody shows up to the table with a stupid character, you need to forget all about that character’s stupidity and work your ass off to make them awesome. The game will be much better off for it, trust me.

 

#5: Thou Shalt Improvise

You cannot plan for everything the players are going to want to do. There are two solutions to this problem: First, you can just not let them do things that don’t fit inside your plans (this is called ‘railroading’) or, second, you can just say “Yes, and…”

“Yes, and” is an improv term, and it is an important one. It means you agree to what the players want, and you use this new input on their part to make more interesting and fun things happen. This requires you, as a GM, to think on your feet a lot, but that’s okay – you should prep with the possibility of improvisation in mind. When I design adventures, I typically devise a series of NPCs the players will or may interact with and then devise the plots these NPCs have going on their own. As the PCs bump into each of these NPCs, this creates a story that spreads throughout the little universe I’ve created. If I need to create new elements, I do so, and they continue to influence the setting of the game, creating more conflict, more story, and more fun stuff. It really works, and it beats the hell out of the whole ‘you can’t do that’ angle.

 

They will do this for six hours straight. Trust me.

#6: Thou Shalt Keep It Moving

Sometimes, players will start planning something (a raid on a castle, a dungeon exploration, an elaborate jail break), and then they will never stop. Not ever. They will spend hours and hours and hours sitting around and arguing with each other and then, when they finally get down to doing something, it’s already midnight and everybody needs to go home.

You job, as GM, is to cut that shit out. This is a really hard one, believe me, since a lot of players have an intrinsic distrust of you (which ought to be wildly misplaced, I might add) and will look suspiciously on any attempt by you to hasten along their planning process. You need to stop them, though, or they will miss the actual fun of the game, which is the execution of their plans. Let them plan, by all means, but let’s keep it down to 2 hours or so, tops. Be willing to fudge details in their favor, if you must, but keep it moving. As interesting as their plans are in the first fifteen minutes, by hour three they are usually just repeating themselves and getting nowhere. Moderate their discussions. Give them suggestions that would naturally occur to their characters. Ask questions that will direct them towards a concrete plan.

This map should be viewed as a ‘suggestion’ or a flavor piece, not the law.

#7: Thou Shalt Not Obsess Over Minutiae

A lot of games lend themselves to the endless discussion of minor details. Games like Shadowrun, for instance, which has elaborate rules for every piece of gear in the game plus lifestyle costs, exchange rates, and so on and so forth. Old Dungeons and Dragons had it’s famed Encumbrance tables, Riddle of Steel has its half dozen interlocking wound tables, and so on and so forth. Maps like the one to the right here are as common as goblins.

Now, these things certainly have their place in the game, but there comes a point when you need to let it go. Screw up a rule? Whatever – move on. Forget that object X actually costs 30% more than you quoted? Nobody cares – move on. Are the PCs stuck in the middle of a labyrinth with no map and no conceivable way of escaping without said map? Too bad – let them find another way. Keep it moving.

This commandment has a lot to do with commandments 5 and 6, granted, and is also related to the all important commandment 2, but it is unique in the sense that it pervades every single part of the game, from character creation through the doling out of XP rewards. The rules – the book – is always the least important part of any game I run. I follow the rules, sure, but I never let those rules dictate what happens in the game. Why? Well, because just following the rules means the PCs can lose. They can all die. The game can end in misery and disappointment simply because your players were (perhaps accurately) too stupid to save their own hides. That’s not okay; that’s no fun. Don’t let the fine print ruin the bold sweeps of a campaign.

Don’t let PCs’ plans ruin opportunities for this moment.

#8: Thou Shalt Make Things Dangerous and Create Tension

90% of players in the universe are danger averse. They want their dungeon crawl to go perfectly according to plan, they want all of their stuff to work exactly when it’s supposed to, and they don’t want their character to be injured.

Fuck that noise.

Stories where everything goes according to plan and nothing goes wrong are BORING. They aren’t half as much fun as when things go pear-shaped halfway through and everybody has to scramble to pull off a suddenly-improbable victory. Those are the games that players talk about for years afterwards. Those are the campaigns that set the standard for every campaign you ever run again. You want there to be danger, since danger creates tension, and tension is fun. Blow stuff up. Have something go wrong. Make the PCs work for their victory, since then (and only then) will the victory be sweet.

There are, of course, limits to this (consider commandment 4), but as a GM you should always seek to make things just difficult enough that plan A is by no means assured of success and plan B is likely to be shot to hell, too. This may sound mean, but if you give your PCs the opportunities to succeed, then everything will be fine.

Don’t be stingy, now.

#9: Thou Shalt Be Generous

If something is not central to the story of an adventure and the players want it, let them have it. If a player creates a long-lost relative who lives in town and there isn’t some reason to forbid it, let it go. If a player wants his PC to be crown prince to a kingdom, tell him it’s okay. If a PC has her back against the wall and doom settling over her from all sides and begs for one chance to make it out alive, give it to her. You are not the PCs enemy. Let me repeat that:

YOU ARE NOT THE PCs’ ENEMY!

You are the facilitator of their grand adventure and attempt at glory, not their direct adversary. You want to make this fun, not arduous or frustrating. If you ‘win’ and all the PCs are dead and the game is over, you have failed as a GM (unless, via Commandment #2, you set this up for them as a possibility). Don’t be a jerk. This isn’t a power trip, or at least it shouldn’t be.

#10: Have Fun

Basic rule of existence: if you are having fun, people around you are more likely to have fun. Smile. Enjoy yourself. Laugh. Do what you think is cool. Your enjoyment of this game is just as important as anyone else’s, and if your players have beaten you into a place where you no longer enjoy running for them, you are missing the whole point of playing a game. Fun is the whole idea and, as GM, nobody has more power to make things fun than you. Go after it! Enjoy!

My Favorite PCs: Major Russ Carmady

This past year I ran a homemade RPG set in my ‘Frontier: 2280′ universe that was, on some level, a reboot of a Battlelords campaign I had run a few years before, though different in most essential ways. It was a gritty, darkly humorous, hard science fiction game in which the PCs were essentially indentured servants of a large extra-planetary corporation that used them as scouts, guinea pigs,  and black-ops troopers. It was a great campaign full of wonderfully fatal events and lots of explosions and ridiculous happenings. There are a lot of characters worth discussing, but the most interesting in, perhaps, one Major Russ Carmady, played by my friend John.

Unlike the rest of the ne’er do-wells, felons, and criminal miscreants that made up the ranks of the XF CFC corps, Russ Carmady was a company man. He cut his teeth with SPIT-NET, joined the private sector as a junior executive on the frontier, and then screwed up so incredibly badly that the company said they could either hand him over to SPIT-NET for criminal prosecution, or he could descend into the ranks of the CFCs and work off his five years. Carmady, of course, chose the latter option.

Think of this guy, only in space.

Carmady was a character that lacked the capacity for self-reflection. He did not see himself as ‘demoted’ or ‘shamed’ so much as ‘transferred’. He had an eternally sunny disposition, a high opinion of himself despite all outward evidence to the contrary, and was constantly thankful for what he saw as ‘opportunities’ that everyone else saw as ‘deathtraps’. He kept the title ‘major’ even though he was in no way entitled to it. He set himself up an ‘office’, which happened to be in the bathroom of the CFC barracks. He had a desk with a nameplate and everything. He was a font of wisdom, in his eyes, but in reality he was mostly making things up. He was a pathological liar, but a very good one. Everyone else on the team either loathed him or thought he was their best chance for survival.

He was absolutely hilarious.

I could list off the magical, almost superhuman snafus Carmady managed to orchestrate, but I won’t. I will simply relate how he, eventually, died. Carmady, due to his mediocre planning, bad luck, and willful ignorance, found himself in a crashed bounce pod on an alien planet surrounded by deadly radiation in the center of a minefield and discovered he was about to be overrun by unidentified forces and possibly taken captive. There was the distinct likelihood that these forces weren’t even human (a first in that world). I gave Carmady three options:

1. Stay here and play dead and maybe they leave you be.
2. Surrender to unknown hostiles for unknown consequences.
3. Run for it through the radiation soaked minefield.

John, his player, asked me one question: “If I’m captured, do I get a black mark on my record?”

“Yes.” I said.

John cited this fellow as Carmady’s inspiration.

He ran for it. The mine blew his body in half. The table all nodded solemnly - it was the death Carmady deserved. Courageous, ill-considered, and cartoonishly ridiculous, especially since he had ordered the minefield set up in the first place.

I’ve had a lot of silly characters in campaigns before, but Carmady was something special. John, more than a lot of other players, really understood the tone of the campaign. He knew we were, essentially, doing a Catch-22 in Space type thing, and he was totally on board. He was going to showcase the institutional absurdity of Man on a galactic scale. He made a character to fit the moment and embody the feel of what I was trying to do in that game. He, in a very real sense, made the game what it was. He was the compass by which I judged my success or failure in any given session. That, it must be said, is a great compliment. I would encourage players everywhere to follow John’s example: figure out how this game is going to work, and find a spot where you can fit right in. Where not only can you have fun, but you can make the entire game magnificent along with it.

Ah, Major, what will we ever do without you?

That New PC Smell…

Been a while since I waxed philosophical about RPGs, so here we go:

Let’s face it: these people probably wouldn’t hang out. At all.

You know that moment in (almost) every D&D campaign where the PCs all bump into one another in some roadside inn and then, a half hour and a tankard of ale later, they’re running off with these near-total strangers to slay dragons? Did that ever rub you the wrong way?

It’s ridiculous, right? Who does that? I mean, most people don’t run around with total strangers in real life, and we live in a world devoid of roadside trolls and murderous death cults (well, okay, mostly without the death cults). I mean, it would be one thing if they all had compatible personalities, but the dwarves never get along with the elves, the wizards are always mocked by the fighters, and the thieves are always, always dickheads. How do these folks suddenly decide to risk life-and-limb together?

I mean, we all know why: it’s metagaming, pure and simple. PCs have that ‘new PC’ smell about them that draws adventuring parties like bug-zappers draw mosquitoes. You all have to hang out together or you don’t have a party. If you don’t have a party, you don’t have a game. We just tend to close our eyes, suspend our disbelief, and roll with it.

How to Deal With It

There are, of course, a variety of ways around this; ways to justify the all-important meeting and have the PCs hang out together long enough to plausibly build actual friendships. Here is a brief (and doubtlessly incomplete) list:

Option #1: They Need Each Other

This is the easiest and most straightforward method to do things. The PCs have to stick together to survive for a certain period of time. Perhaps they find themselves in a town that is under attack by horrible (whatevers) and find themselves sticking together simply to survive. Maybe they are all prisoners in the same dungeon and have to rely on one another to escape and then, of course, find themselves stuck together as fugitives from whatever force placed them in the prison to begin with.

The options are numerous, but most of them are in medias res type beginnings. This is a bonus or a drawback, depending on the kind of campaign you’re running, since an episodic game with a rotating cast will resolve the issue that is keeping them together rather quickly and then, in the next session, you find yourself back at square one. Furthermore, even in serialized games with long plot arcs, sooner or later the thing that brought them together is going to get resolved. Then we are either left closing our eyes and assuming they stick together or watching them shoot off in various directions.

Option #2: It’s Their Job

This is an easy one and can very quickly build long-term party cohesion: all of the PCs are employed by the same (whatever) and are, essentially, coworkers. They need to put up with each other whether they want to or not. They might be mercenaries, in the military, part of the same secret society, or any number of other options – all of them can work.

I’ve used this one a lot, and I can tell you a couple things. First, this set-up leads to automatic intra-party bickering. Since the characters aren’t required in any sense to like each other, many of them don’t and your players will engage in entertaining-but-time-wasting arguments with each other just for fun. Second, this an ideal set-up for a game with a rotating cast, since you can easily have this or that PC ‘transferred’ for a session or two without straining anyone’s imagination. The primary (and only) drawback of this situation is that you are relying upon external forces to keep the players together. Some players might chafe at this and, furthermore, if the external force gets removed somehow, you are back to square one.

Option #3: They Are Already Friends

This is another easy one that requires just a little background work for each character. All you need to do is have each character start with a positive relationship with at least one, but preferably two, other characters. Your PCs are already buddies, have already been through hell together, and they should join up without squabble or reservation. Give them a collective motive and bingo – you’re on your way.

There is, however, a drawback to this set-up. It is, primarily, that it limits the kinds of characters that can be plausibly connected without straining the feasibility of the relationship. If you are playing in a campaign were Fizziks and Gurkles have been at war for centuries, and one guy wants to play a Gurkle Chieftain and another guy wants to play a Fizzik Enforcer, it’s going to be a tough sell to explain how they’re friends already. You can probably make it happen, but it’s not a natural fit and will require a lot of backflips and contortions. Now, if this doesn’t bother you, then go ahead. It might bother your players, though (after all, that guy making the Fizzik Enforcer made it specifically so he could hate Gurkles and the Gurkle Chieftain had his whole family enslaved by the Fizzik Empire…).

Option #4: Don’t Even Try

There is no law in (good) RPGs that states that parties must stick together all the time to survive. I mean, that’s the case in D&D, but that is more video game than it is RPG, in my opinion, anyway. Use Option #1 just to give them an initial stick-together period and then loosen the reins. Let them go where they will, do what they will, associate with whomever they chose. The characters that most naturally would associate with each other, will. Those who wouldn’t, won’t. No biggie. It’s their game, let them explore it.

The drawback here, though, is a fairly substantial one that has two parts. Firstly, it is pretty daunting managing 3-4 storylines at a time as a GM. It takes a lot of prep, a good head for improvisation, and a sharp memory. Second, and related to the first, you’ll wind up with long periods of playtime where some players have nothing to do. When I used to run long-run campaigns, this kind of thing would happen from time-to-time (sometimes too often), and I’d have six PCs in four locations. If you were Group A, you’d be playing only 25% of the time, and the other 75% was just sitting around and listening. I was fortunate enough in most instances to have my players really engaged in the action of the game, so they often didn’t mind listening. Some, though, got bored, and I don’t blame them. If you try to use this method, make sure to keep it under control and plan on bringing the party back together sooner rather than later.

Anyway, that’s my bit on this. I should note that I mix and match all the methods fairly liberally in my game. No matter what, though, I strive very hard to keep the artificial and the meta-gamey out of my party dynamics.

My Favorite PCs: The Vesten

The 7th Sea campaign I ran from 2001-2004 was an exceptional thing. It produced a greater concentration of fantastic characters than any other campaign I’ve run and, to some extent, forged the gaming group I still play with today. This campaign featured Helmut Dauben Kohb, the statistically improbable, unkillable Eisen, as well as Lord Edward, the shamelessly hilarious rake. But they weren’t alone by any means. I have yet to tell you about the Vesten.

The legend, but not the man…

The first thing that is amazing about the Vesten is that it was, in reality, two separate characters. The first was named Ruin, played by my friend Bobby, who was a towering viking goliath of the Vesten people, a reincarnation of a hero of legend named Valkar, who was fated to slay the Great Wyrm. The second was his friend and countryman, Galdar (played by my buddy DJ), a powerful rune-mage and a man who, at various points in his life, was priest, warrior, pirate, and merchant. The two of them–both powerful, savage northmen who struck down all who opposed them–managed, unintentionally, to find themselves doing things in different places at different times. The common people of Theah could not divorce the two (how many blonde barbarians could be running around destroying Montaigne ships or crushing Vodacce merchant fleets at the same time, really?) and so rolled them into one immortal folk hero/terror called, simply, the Vesten.

Both characters were very different from one another. Ruin was garrulous, reckless, guileless, and without fear. Galdar was vengeful, brooding, careful, and stern. Ruin was a wrestler; Galdar was a boxer and pistollier. Ruin became the master of his own school of hand-to-hand combat, Galdar became a living embodiment of the Rune of Fury and was lightning incarnate. Both of them tread the line between myth and reality: Ruin wrestled with a grendel and answered the riddles of a dragon (one of the most tense sessions ever, incidentally: no rolling, just me asking riddles and Bobby having to answer them correctly or see his character plunged into a fiery abyss). Galdar fought with Krieg, the Vesten God of War, and turned down his proposals and, at the conclusion of the campaign, sacrificed himself to become a five-hundred foot pillar of lightning and storm clouds to destroy the Ultimate Evil. Their battles and their choices built reputations for their characters that topped those of even the most famous heroes and villains of the game setting. They had a little race: the player who had the highest reputation at the end of the campaign would ’win’. It was a close thing–they had reputations of 135 and 130 or so. To put this in perspective, few characters manage to top 50, even in a long standing campaign. When they took action, the world shook.

This is, ultimately, what made them such amazing characters: Bobby and DJ’s actions through their characters actually changed the campaign setting. They diverted the paths of nations, they destroyed long-standing terrors, and uncovered new knowledge and new ideas to change what the world would become. What is exceptional about this is that I didn’t make them do it. They had the choice to take the easier path–to be, like most RPG characters, drifters in the stream of the world–but they refused. They wanted their characters to change the world, and they did.

Managing this as a GM, you can imagine, was challenging. I had to loosen the reins a bit, let them do what they wanted, and allow the plot to form itself in reaction to their decisions. It forced me to let go of my preconceived notions of what the story ‘was supposed to be’ and let it become something else equally awesome. I confess to trying to rein them in from time to time (if for no other reason than to let the other players step up to the plate more often), but overall I think I allowed them a lot of freedom and the campaign was better for it. When Galdar (or one of his aliases–Galdarus or Gillis) wanted the Vendel to build him a great warship, they did. It was the best damned warship in the world, and they used it, dammit. Did it break the campaign? No–it changed it. Change is not bad, or at least not necessarily so. Losing control as a GM doesn’t always mean ruining your campaign. Sometimes, with the right players, it means making the campaign one your players will be talking about for years and years to come.

Killing Their Babies: RPGs and PC Death

Violence, battle, and peril are a constant in RPGs. I’ve explored the why of this elsewhere on this blog in various places, so I won’t go into it here, but suffice it to say that such things are what make the genre tense and exciting in many ways. Few are the games that don’t involve some kind of man-eating monsters, bloodthirsty villains, perilous cliffs, and exploding doomsday devices. It stands to reason, then, that death and, particularly, the deaths of the occasional PC are bound to occur. When this happens, however, it can be a bit of a shock to the players. It can, if mismanaged, create bad feelings between the players and the GM. Of course, if the GM never allows it to happen, bad things also happen. So, how to manage this? Well, here’s my advice on the subject. 

Why it Needs to Happen

At some point as GM, you probably need to step up and kill a PC. The reason you need to do this is the same reason that cliffhangers and adventure stories have a tendency to kill characters from time to time: it makes the danger more real. If every time a player gets his or her character in a fatal predicament they are allowed, somehow, to escape it (through the GM fudging the rules, through random deus ex machina, and so on), the party is going to catch on that they are, in essence, invincible. This is very bad, and for several reasons. 

Firstly, the players will cease to feel threatened by the dangers that the GM places before them. Just like in a bad adventure novel, the GM has given the players ‘plot armor’ that they know to be impenetrable. This makes the game boring, suddenly. Obviously they’ll be able to jump over that chasm as the castle is collapsing around them. Clearly they can live through their death duel with that vampire lord. How do they know? Well, they know the GM hasn’t the guts to do anything about it.

Secondly, and derived from the first problem, the GM can suddenly become ‘bullied’ by their players. The players can have their characters do outlandish things in the utter confidence that, even if they don’t work, there is little risk their characters will suffer for it. This can begin to break the mood of the game (unless the *point* of the game is to be invincible and do outlandish things, like Toon and the like), and things rapidly become more and more absurd. The game begins to morph from a stylized, internally consistent story to a bad improv long-form show. As someone who has been in his share of bad improv long-form shows, they might be funny, but that’s about all they have going for them. The game goes from adventure to joke. I’ve played in campaigns like this in my time, and the novelty wears off quickly. 

Of course, how often and why to allow PCs to die depends greatly on the style of the game. Gritty, violent, and noir settings obviously feature death around every corner, and PCs become much more cautious in their play and less attached to their characters. Heroic or swashbuckling settings feature death much less often, and when it happens it represents a serious dramatic event. Still, even with the most heroic settings, death should be possible and it should be clear that they are possible if things go wrong. Even if the GM doesn’t really want to kill the character if they do something stupid, they should seriously consider permanent disfigurement, maiming, or similar permanent consequences. Consequences are important to create tension; tension is essential for adventuring fun.

How to Manage It

As mentioned above, how to handle killing a PC depends greatly on the mood of the setting of the game. The likelihood and frequency of fatal situations should be made clear to the players prior to the beginning of the campaign. The GM shouldn’t be setting quotas or anything (i.e. I intend to kill one PC every three sessions! Mwa-ha-ha!), but she should say things akin to ‘there will be no holds barred in this game–if you screw up, you’re dead’ or ‘I don’t intend for characters to die for stupid reasons, but they will die if dramatically appropriate or compelling’. This gives everybody a good idea of how dangerous the campaign is, and this is very important for the players to know when constructing and playing their characters. It also should preempt some of the bad feelings that might develop otherwise should a player lose his or her favorite character.

Beyond this, I have a couple rules of thumb:

1. The Good Death: Unless the game you are running is exceptionally dark, grim, or violent, PCs should never be killed due to silly accidents, random events, or simply poor luck. They should be killed by important villains, by exceptionally deadly traps (that they are aware of and attempting to evade), or while knowingly placing themselves at fatal risk due to their character’s traits or behavior. In short, they should die thanks to their decisions (good or bad), not due to their luck. Their death should be dramatic, motivating to the other characters, and serve as a significant plot point for the campaign. It should mean something.
2. Get Them Back in the Game: Unless the death occurs at the very tail end of a campaign (where it would be silly to introduce a new character that would be played for 2-3 sessions tops), always allows the player to make a new character and introduce them into the game as soon as possible. Death should not be a punishment of the player.
3. It Isn’t a Punishment: This bears repeating–PC death is never, never a punishment. If you are a GM forced to use it as a way to regain control of a campaign, you have done something wrong and haven’t correctly set up the expectations of danger in the campaign in the first place (leading to bullying by your players, necessitating death). This is bad news. Ideally, players should think their PCs’ deaths are cool–they get a cool death scene, and they should be allowed to play it up. Then, they get to play a new character (that is every bit as advanced and powerful as their last character, more or less).
4. Make the Death Matter: This is the hardest of the rules to manage, but also very important. A PC should not die and be forgotten. Their death should have a major effect on the campaign and the other players; when they die, something new should be revealed, they should be contributing to the story somehow, and something interesting should happen. Don’t kill for no reason (unless you’re running one of those super-deadly games where life is cheap, and then everybody should be on board with that so it shouldn’t be a big deal).

Beyond this, if you find your players getting into circumstances where they really should die, but it wouldn’t fit with the campaign and wouldn’t make much sense, really consider simply maiming them or otherwise afflicting them with a kind of permanent consequence that makes the character interesting to play, but doesn’t allow them to get off scott-free.

Anyway, whatever the circumstances, one cannot run a campaign without the possibility of fatal consequences. If you are GM-ing such a game, it is your narrative responsibility to allow it to happen. You should do it, however, with caution and care to guard the player’s expectations and to maintain the fun they’re happening. If you’re a player, you should also understand that the death of your favorite character is as important as his life in contributing to the fun of the game. Don’t get upset, just roll with it; after all, it’s just a game.