Back in the old days, when the world was new and before angst had been introduced to our everlasting fight, things were simple. We were evil, probably insane or twisted, and that was all there was to it. They were Good, purity of heart was pretty much a job requirement, and while no amount of blandishments could turn them from their duty, they were so Good that most of them didn’t have the heart to kill us. But the conflicts were too great for that flat, two-color world to hold, and it began to swell, and to grow, and to push itself into three dimensions. And when it did, motivations of each side leaked onto the other. Conversion from Evil to Good had always been possible, but now, with flawed heroes who weren’t Doomed, and honorable motives within our ranks, we had a shot at converting them. And the first thing we learned to use was weaponized psychology.

The principle behind weaponized psychology is simple. It’s an offshoot of ‘know thine enemy’: to be exact, “Know thine enemy better than thine enemy doth know himself.” And, of course, exploit it (but that goes without saying, right?). In most cases, the goal is to cause the target to shut down, or possibly to cause him to join us, though most of us go into a weaponized psychology session fully aware that various constraints both narrative and non-narrative make that one something of a pipe dream. But it’s worth trying, and let’s face it, a good bout of weaponized psychology is fun. So how do we do it?

First step: Know. Thine. Enemy. I don’t care how good your speech on a hero’s suppressed daddy issues are, it’s not going to work if the hero doesn’t have daddy issues. He’ll just laugh at you. So pay attention. Do research. If you don’t have spies, or scrying magic, or whatever, get some. Learn whatever you can about him, then learn to apply it to reach other conclusions, and make sure you stay current. The one thing more embarrassing than picking on a common flaw that he happens not to have is picking on what was a significant weakness when you first met but that he’s already arced through. This also means knowing the kind of conversational style to take. Inexperienced heroes still drunk on their own goodness may respond better to a direct reaming out, but the worldly wise who know your tricks are going to need a subtler, more allusive touch, and if you’re confident enough to challenge a deceptive protagonist, you will need to make sure that you’ve got the same mastery of allusion and metaphor and giving him just enough information to let him come to the rest of it as he does. The ones I’ve met have all been far likelier to play the game for the game’s own sake, so they might stick around for the discussion just because it’ll be a challenge. See below for why this is useful.

Second, make sure you’re dealing with the right kind of hero. He needs to be at least a bit introspective, or else he just plain won’t care about what you have to say to him. On the other hand, he can’t be so introspective that he’s already independently reached your conclusions and found a coping mechanism. Weaponized psychology, particularly when used to paralyze, requires a certain amount of shock/revelation to be truly effective. Perception also bears a relation, but it’s iffy: people who are perceptive about themselves are very hard to surprise, but the ones who understand other people quite well are often vulnerable, particularly if you lead off with a few observations about their teammates that mirror theirs. That establishes you as knowing what you’re talking about, making it likelier that they’ll realize there’s some truth to what you say when you start in on them. Truth is an important element to weaponized psychology; to make it work, we have to counter our reputation for lying.

Of course, weaponized psychology isn’t going to do you a jot of good unless you can get the hero to listen, now, is it? Some people deal with that the obvious way, by giving literal meaning to the captive audience. And yes, that works, but being restrained and stuck listening to what they see as a classic monologue makes some heroes a bit oppositional, which gives extra barriers. I find that it works better if the hero is coming to you in the first place. Maybe you know something he doesn’t and needs to. Or you’ve managed to project the image of someone who despite the whole evil thing could be admirable, interesting, worth working with, and they just can’t get enough of talking to you. Either way, set up a situation in which it’s the hero’s choice to leave or not, and it just happens that it’s more in his interest to stay.

Weaponized psychology is a difficult, individualized science. But when it works, giving heroes logic errors at critical moments is the least of the things it can do for you. Consider it, and if you choose to do it, learn it well. Whether Good is dumb or not, be smarter!

Well, that didn’t go as expected.

The plan was simple. Party’s going after a guy who pretty much controls the world in which they’re dreaming. Split the party, mess with the party’s heads, everyone wins (…eventually, for some value or another of win, whether IC or OOC), and I get to play my favorite antagonist one way or another in four different windows.

In theory. Of course, it would have worked a lot better if the entire group had been present (one’s entire RSVP was an offline IM sent less than an hour before I got home, and as a result I had no warning whatsoever), if one of the players hadn’t shown up most of an hour late, if it hadn’t been so transparent what the plan was, and if I’d actually gotten to play the silly antagonist. Everything’s better when I’m one of my favored characters.

So, the splitting. As far as the antagonist was concerned, it was all about the old divide and conquer strategy. I’d originally conceived it as pairing something I enjoyed with something I didn’t; I’d had a plan involving one fight and a couple of conversations. And with one of the players missing, sure, I had to rewrite my original plan (it wasn’t too feasible anyway), but at least it meant I could take the character whose player stepped out in a medias res situation he himself was responsible for and not have to subject him to a plothole. Anyone who reads this regularly can probably guess the extent to which I hate plotholes. I’d thought I was going to get a teeny bit of the split-session done during session, but it was late, I was behind on my article, and don’t get me started on my Saturdays.

And people wonder why I try to avoid planning specific events ahead.

I spent the last couple of days talking about differentiating between characters’ reactions to other characters, both in making different people’s reactions to the same character different and in finding ways for one character to react differently to different people. But most of those assume that you’re dealing with a situation in which a character can react as she feels, and that isn’t always the case. What might keep a character from reacting the way she ordinarily would?

Cultural mores are a big one. Sure, it might not be something you can always expect to affect player characters (often, they don’t know or they don’t care), but they’re still a good source of standardized reaction. In most fiction, this takes the form of cultures that expect visible emotion to be kept to a minimum, but why limit yourself to that? It might be that the situation calls for a certain definite and obvious emotion, or there’s a ritual way in which one is supposed to deal with another person of a certain sort. Set these sorts of situations loose, and watch the layers fall.

Dealing with people of different status. Going back to Buccaneer’s Guild’s comment day before yesterday, having someone of different status present can lead to wanting to stifle one’s first reaction for any of a number of reasons. If they’re higher status, the object of the game might be impressing them, and that doesn’t work too well if you’re reacting in a way of which they would not approve. Or they might, if they don’t like the way they’re being looked at, make life very difficult for whoever’s looking at them, directly or just by influencing the others present. And there’s always trying to inspire confidence in your ability to handle the situation: World going to pieces? We can handle it. A character might be trying to impress her equals by not looking all awestruck at the Powerful Person, or otherwise seeming to take something Big in stride. And then there are those people who insist on being distant with their inferiors so they will come across as unreachable, set apart, what have you. (Granted, it doesn’t always work…)

And of course, there are tactical uses for changing one’s apparent reaction. Hiding fear, feigning interest, squelching boredom—it’s all about keeping people from being able to use one’s real reaction, or possibly giving them a different reaction to try to exploit. It doesn’t even need to be limited to individual situations; a group that can at least feign a unified front looks a lot more formidable than one that interrupts its pronouncements as a couple of its members argue about their intent or bicker over semantics.

To further complicate scenes, consider how these restrictions on reactions are going to affect people’s internal reactions. One might resent having to project a different emotion than she actually feels. Characters seeing other characters’ socially acceptable or carefully chosen reactions might be favorably impressed… but on the other hand, might be frustrated because now they have to work with that. Just because the reaction isn’t visible doesn’t mean it’s not going to be relevant later, or important for you to know.

So when working with reactions, think about not just whether a character can, and whether a character would, but whether a character should. The result will be interesting, and there’s little that can’t be improved by making it more interesting.

Yesterday, I started talking about non-uniform reactions, or differentiating between different characters’ attitudes to the same person. But that doesn’t take into account internal variation—who says that one person’s going to have the same reaction to a whole bunch of different characters? Not I, not by a long shot.

I tend to take these on a case by case basis, since I’m usually trying to get into the reacting character’s head. It’s easy to start with the same sorts of basics as with differentiating multiple characters’ reactions: overall personality compatibility, shared history, that sort of thing. But once you’re through those, you’re going to want to get more closely into the character’s head, start figuring out who’s seen as different from whom and why.

One thing you might want to look at is qualities and priorities. Your reacting character probably values certain aspects of character and has little use for others, prioritizes some things and sees others as a waste of time. So sure, she’s showing disdain for both these characters here, but in one case it’s because she values loyalty and competence and he’s shown precious little of either, while she can neither fathom why nor respect the fact that the other one would choose love over duty in a heartbeat.

How do these different people interact with her plans? That’s likely to make a difference in and of itself. Often, a character will tend to have more positive feelings towards someone who can help her than one who can’t, and towards one who will help her as opposed to one who won’t. But then you have the times when that conflicts with the rest of her attitude: one character might be useful in the future, sure, but if she doesn’t like him to begin with that might only make it worse. As Buccaneer’s Guild pointed out indirectly yesterday, you’re also going to want to take comparative status into account: not just how the character reacts to the other characters, but whether her internal reaction matches her external reaction, whether it’s allowed to, and how she feels about that.

What about old debts and grudges; how do they play in? Sure, a character’s probably going to feel better towards someone who’s done something big for her…. unless, of course, they never let her forget about it, or won’t let her pay them back, or the like. How do her personality, her compatibility with the other’s personality, and/or good or ill done to each other in the past, affect her dealings with someone who should be a hereditary enemy or who did That One Thing that she feels she shouldn’t be able to forgive?

One important thing to remember is that there isn’t always an obvious link between one character’s feelings about another and the other character’s feelings about the first. Just because she loathes him doesn’t mean he has to hate her; just because she thinks of him as her favored apprentice and possible successor doesn’t mean he has to like her as a teacher. Not only do these imbalances make it more interesting, but they themselves can further affect the reacting character’s opinion of another character by interacting with her personality and the rest of the tangle-up. His unwillingness to mirror her hostility might soften her attitude, but it might also lead to further attempts to provoke him, or to something else entirely.

The end result, if you’re willing to do this for all the important character reactions, is a lot more depth and variety than you’d get from approaching everyone’s feelings towards other characters as being fundamentally the same. Removing the uniformity from the characters’ reactions to each other makes them more real and considerably more interesting. Give it a try!

This is another of those topics that branches from the idea that no two people are the same. Yes, this should be obvious, but it often gets lost in the rush when people are trying to juggle large casts of characters, and what results is oversimplified reactions. Perhaps an entire group worth of people are all reacting to one character the exact same way. Or everybody’s reactions to people seem to divide into “Like!” and “Hate!” This gets boring, and should be avoided if possible.

There are two approaches to non-uniform reactions—looking at the differences between the people reacting to one person, and the differences between the people being reacted to from the point of view of a single person reacting. I’m going to start with the differences between the people reacting: if you’ve got a single character (for this article, the focus character) standing there, and a bunch of people who’ve never shown up for more than cameos in the narrative whose reactions to her you’re going to have to figure out, where do you start?

Many differences go without saying. A person’s friends will generally have a different attitude towards her than her enemies will. People who are too different tend to have trouble getting along; so do people who are almost completely similar but with key differences in particularly important aspects of their personalities. Those who have things they value in common are likelier to get along… usually. Begin with generalities; sort these other characters out depending on how they relate to her, and assign them a general class of impression, something that can be summed up in one emotion word. Try to avoid just a loving/loathing dichotomy; look more for things like respect, envy, pride, disdain, you get the idea. This is a base reaction: how might each character view your focus character if they’ve only heard about her and never actually met?

Next, think about the general effects of whatever impact the focus character might have had on each reacting character’s history. You can still be vague here, but this makes your groups smaller, further differentiating them. Logically, this is where the time the two characters have interacted with or directly affected each other comes into play—wrongs or rights done, shared experiences, the like. But don’t forget indirect effects, where something the focus character did or was in turn did something that affected a situation involving one or more of the reacting characters.

Doing the above two steps is going to get you some pretty small groups. At this point, look at the people in the groups and try to find things that give them different versions of the same sort of reaction. Let’s say there was a grand courtly party, one of those things where everyone who is anyone is there, and your focus character is a relative newcomer trying to shoot to the top. You’ve got four people, all of whom considered her annoying even before the party, and who after that disgraceful mess involving the Earl of Chevar’s son and the punchbowl, now have no respect for her whatsoever. Here, you might decide that one had been dragooned into trying to teach her her way around the court, but the incident was in direct opposition to his helpful suggestions. Another just plain can’t take people who fly off the handle that easily seriously. A third deplores the waste of perfectly good punch, particularly since she hadn’t gotten a chance to have any of it and was about to when the incident took place. And the last didn’t mind any of those factors too much, but considers people who can’t insult with finesse to be below his notice. All the same general class of reaction, but all different flavors and thus likelier to react differently.

Tomorrow, I’ll look at it from trying to give one character multiple reactions.

It gets people killed in horror movies, has led to more misunderstandings (or cases of people knowing things they probably shouldn’t have) than almost any other plot device, often results in the person with the right skill being in the wrong place when needed, and can slow a plot to a crawl. And that isn’t even taking into account the frustration that can come from the GM multitasking, or from figuring out how to keep the rest of the group from either getting bored or knowing exactly what’s going on when their characters should have no clue. But despite that, you often hear someone cheerfully suggest, “Let’s split up, gang!”, the people around them follow through, and the GM agrees to this. Why?

1. Avoiding arguments. If the group wants to split up, who are we to railroad them into staying back together? It’s their own fault if part of the group blunders into a fight that was designed for the whole group.
2. Avoiding boredom or frustration. For the GM who can run two scenes at once (probably due to running by play-by-post or chat), it’s a way to run two varieties of scene at once—and have a scene of the kind we want to write keeping us from being completely bogged down by the scene we’d just as soon get over with. (Note: there are very few GMs who will actually admit this.)
3. Scheduling. Whether it’s a freelancing job or an unavoidable field trip, time zone havoc or unavoidable technical difficulties, one character’s player (or multiples, sometimes) just can’t make it at the normal time. What to do? Let’s send him over here for a while with this NPC and they can meet up at the end. It gets them off screen, but they aren’t sitting around doing nothing, and what they are doing can be figured out at a different time and might be useful later.
4. Connections between characters. You ever seen one of those times when two characters who ordinarily can’t get along to save their lives are isolated from the rest of their companions and have only each other to depend on? (Note that this does not always work; some people will just keep bickering anyway.)
5. Use of characters who just don’t work for the purpose of the scene with the full group. Sometimes, you just can’t get two characters in the same scene without something going wrong. There are two people who aren’t aware of what one of the NPCs really is, and today’s the wrong day for them to find out. The person the group’s trying to get some information from clams up around all of them, but one might be able to talk her into opening up. One player just plain can’t stand a certain NPC, and nobody feels like dealing with the inevitable clash, but there’s something that NPC is needed for.
6. Information. While keeping secrets is the most obvious reason to split the party, a split party is a good excuse to disseminate information as well. Maybe you’re giving each portion of the split group one fact and seeing if they can put it together. Or you’re going for a more Rashomon-style scenario, telling each about the same thing but from different perspectives. Or there’s just plain something you only want a subset of the group to know. Sure, you could do it in sidechat, but this keeps you from accidentally ending up in a paradox, and if you want people to know that something’s up, this is the best way to arrange it.
7. Uncomplicating a scene. Unlike “Avoiding boredom”, which only works for a known multitasker, this works better when you run your split scenes sequentially rather than simultaneously. One scene with ten people is hard. Two with five each can be a lot easier. This happens a lot with fights, but is also likely when dealing with very large parties.
8. Divide and conquer. What list of splitting strategies would be complete without this one? Dividing the group up breaks some of the synergies it might depend on, weakening it by more than would be expected from simple division. It can allow a group of opponents to use tricks targeted to each without worrying as much about one member of the group compensating for another’s weaknesses. If you’ve got a social monster in the wings, or a shapeshifter with duplication capabilities, they might even be able to subvert or replace one of the stray group members. A lot can go long when people aren’t there to look out for each other, can’t it?

Yes, splitting up the group requires more attention, special tactics, and a lot more patience. But that’s a reasonable price to pay for all the advantages. So let them split up, gang—particularly if you have some say in who’s with whom.

Many games and stories feature characters who can turn just about any physical object into a weapon, regardless of any of the many sources of impracticality it demonstrates. And many people realize that this is Awesome and attempt to emulate it, resulting in someone who can explain how to foil an assassination at a high banquet just by using whatever’s on the table and then delivering one-liners about how nobody can say they lack subtleties. Particularly when they’re doing it just as much for Awesome’s sake as for tactics’ sake.

But that’s only a part of what I’m talking about here. Sure, “everything is a weapon” includes being able to smack people with anything in the solid state that shows a decent density, parrying with furniture, garroting people with their own neckties, you name it. But there’s more to everything than objects one can hit things with.

Terrain is a weapon, particularly if it’s terrain that you chose. Things like getting the high ground, giving your opponent unstable footing, and making sure the sun is in someone else’s eyes are logical. And of course, rigging traps is a popular and reliable means of further changing the battlefield. One can kick up dust, bring down scree, use one element of the backdrop to affect another element, put an element of the backdrop to uses for which it wasn’t intended… there are a lot of possibilities.

Allies are a weapon. And I don’t just mean fastball specials or the like, where an ally is literally serving as a projectile or a reach enhancer or a source of extra momentum. They can force an opponent to split his attention. One’s rapid-fire attacks can serve as a way to herd the target into another’s more focused strike. Synergy with their skills can lead to things one person alone wouldn’t be able to do, whether it’s taking advantage of one’s ability to provide lots of entangling vines or trusting another to be able to pop out of nowhere in just the right spot. Coordination is a powerful thing.

The enemy is a weapon. Granted, that’s usually obvious when he’s trying to defeat you. But how many swords have only one edge? A socially inclined character can bait the opponent into making a mistake, get him to lose his cool, or even just have him laughing so hard he can’t fight. One who understands the enemy’s psychology can use it to predict his movements and take advantage of the predictions. Even with natural physics, helmets and capes can be turned into opportunities by exploiting a lack of peripheral vision or sticking them to the ground. Weapons are a weakness in that they can be removed. And the closer you get to Rule of Cool physics, the more improbable you can make your exploitations of your enemy.

At first glance, this probably looks more useful for a gamer trying to impress the GM with his resourcefulness. But that doesn’t mean other people can’t use these tips. Maneuvers good for the player are good for the GM as well, after all. And a writer utilizing these tricks can both demonstrate that whichever of his combatants using them is quick-witted and resourceful (show, don’t tell!) and make for a far more interesting fight than people just running around generically bashing each other with big sharp things or colorful magic blasts.

Everything is a weapon. Watch it change your fights for the better.

Normally the management and I don’t get too involved in each other’s work, but here we have a perfectly good riff on dreams, and nothing about how we deal with them. See, the dream is an excellent way to mess with a protagonist’s head. What better way of influencing them than in a situation where they can’t be sure if it’s us interfering or the Forces of Good sending a message or their own neurons firing in funny patterns or whatever metaphysical garble they explain the pretty pictures with? And better yet, they usually aren’t too good at shielding their minds when they’re asleep. Not paying attention, after all. But if you’re going to mess with their dreams, you’ll want to figure out what you’re trying to accomplish first, so you can adjust your battle strategy.

Usually, when messing with the protagonist’s dreams, we’ve got one of three objectives. In some cases, we’re just trying to deliver a message, one way or another. If your target is quite aware that this is how you’re going to contact her, go ahead and be obvious about it—adjust the dream to your personal advantage, speak your piece, and leave. While you can converse, I don’t recommend it, particularly if you’re up against a deceptive protagonist; it leaves you open to the same kind of influencing they’re doing on you. On the other hand, it’s quite possible that you want to deliver a message but don’t want it to seem to be from you. You might take on the semblance of someone else, or sculpt a dream that can be interpreted as suggesting whatever you want to get across.

Another reason why you might be in a protagonist’s dreams is to influence her emotions, either on a certain subject or in general. Maybe you’re trying to put the fear of you in her, or to make her more friendly towards one of your plants, or just to induce a sort of universal ennui or melancholy in her. For this, subtle is better. If it’s obvious that you’re sending the dreams, then they’re going to be resisted, but if your target thinks it’s just part of her, she might even make a point of justifying them when someone else tells her she shouldn’t be so wrapped up in them.

As a combination of the above two, you might be in a position to try to make the target want to follow a certain path—maybe come to a realization and then follow it wherever it leads. This, of course, takes time: you’re going to want to figure out what you want them looking for, and then you’re going to need to deliver it in pieces, in such a way that they’re having to work for every little piece. After all, nobody trusts a free lunch—and small wonder, given how often the things are poisoned—but the stuff they have to work for, they value, because, well, they worked for it. There’s a fallacy for that out there somewhere.

Then there’s the third reason: giving them the wrong impressions about themselves. This one is for masters; it requires utmost subtlety and a greater understanding of the target than any other method. If you want to make your target think she’s slowly losing her mind, you can’t put your face on every creature she walks past in her dreams unless what you’re going for relates to paranoia about you; if you want her to believe her friends are rejecting her, they have to speak in their voices and use their phrasings, not borrow yours. If it can be traced to you, if all of it is clearly your work, then it all shatters, because there’s a clear cause. But if there isn’t a clear cause, then it Comes From Inside, and if it Comes From Inside it must be true.

Knowing your objective is the first step to figuring out your strategy. “Evil will always win because good is dumb” only works when evil is smart like us. Think it through!

Dreamscapes. Such lovely things, when you want to try something new. This week, I got to apply them to my own game.

It had been coming for a while, after all. The group’s old worst enemy had cropped up in the head of a friend of theirs, something a lot stronger than a memory. He was useful, sure, but dangerous. So one of the players got the bright idea of extracting him from his host’s mind and making him into a talking paperweight, so they could still ask for advice but wouldn’t have to deal with him engaging in underhanded politics through his host when they weren’t looking.

Having discovered the joys of battle in dreamscapes in a game not long before (and driven my GM crazy in the process), I had but one solution to this: Get them into the host’s dreams and let them battle it out. So my group found themselves trekking through the mind of their young friend Nandin on the way to beat his passenger into a coherent and hopefully singular state.

The first thing I decided was that Nandin, being originally from a rather Sahara-esque desert, would have a desert for a dreamscape. Under the houserules we’d used in the other game, temperament determined dreamscape; Nandin’s would be comfortable, for a desert, reflecting his kind personality. So it was an expanse of white sand, glittering in the sun which shone down from a brilliant blue sky. There were landmarks, though not too many of them: the mountain which housed his dreamscape’s guardian, the oasis in which Nandin himself was hanging out, and the wall of haze that separated the Nandin!dreamscape from the Jalil!Dreamscape. It didn’t matter how far away they were, though; if the group intended to go somewhere, the group would get there: instantly in the case of the place that were Nandin’s, in five minutes to reach Jalil’s side. The guardian herself, whom the group visited first, ended up being a roc/phoenix/something blend with the voice of someone the group had never met and the mannerisms and text color of Nandin’s closest friend (and, for that matter, Luath’s older sister), Irayo.

Jalil’s, on the other hand, was different. It wasn’t just the terrain, though going from glittering white sand to a brownish-reddish sandstone can’t have hurt the contrast. It wasn’t just the occupants, though the sand-fleas and “cute innocent vicious rats” (basically what happened when I gave the little critters the outward appearance of kangaroo rats and the diet and tactics of piranhas) certainly made it seem more hostile. It was also the metaphysics. Nandin was pretty normal, and his dreamscape behaved about as one expected dreams to. Jalil, on the other hand…. messed with the laws of metaphysics a little. After all, when I’d designed him, he’d broken one of the Cardinal Laws of Endgame Bosses by having just about everything better than his physical stats. In life, this caused him issues. In existence as a dream figment in a place where people aimed with their manipulative ability, powered blows with their charisma, and resisted damage by being pretty, particularly where most of the group had been hamstrung by this little aspect of the dreamworld, he could be deadly.

Though I’d had some fun with some of the ideas I’d had for Jalil’s architecture, they didn’t get there. Last I saw them, they were being swarmed by the same rat-piranhas and another creature, and between one of them turning into a rolling ball of DOOM, another creating small-scale novas with his fist and not much else, my getting to have some fun with my own crazy dream action concepts…. I think I am remembering how to run a fight that I love to pieces. We had to end early, but I’m looking forward to seeing how it ends.

(As a side note: Happy Halloween to everyone! It finishes off a pretty exciting week; this blog hit the 500-post mark two days ago, and my birthday was four days past. Party!)

Some plot twists just can’t get themselves taken seriously; just mentioning that a story ends in one is enough to make a hefty portion of its audience look for something else. “All a dream” is one such, and quite possibly one of the most reviled of the offenders. But that doesn’t mean that the basic premise, that the scene that just happened was just a dream, isn’t still usable. It just means that you have to be clever.

The first rule of “All a dream” is that you shouldn’t use it at the very end unless you have a twist to that twist. What makes “All a dream” such a reviled ending, after all, is the combination of two factors: one, that it’s so common as to be cliché, and two, that it’s a waste of potential. Sacrifices? Gains? Losses? All those motions of emotional intensity that kept the readers hanging so much? Not real even by the story’s standards. That subplot that could have been interesting if it’d been taken to its logical conclusion? Never going to happen. The plots won’t go anywhere, the emotions didn’t matter in the end; what’s the point of it? If you’re clever enough to find a new way to play with it, they might go along with it, but it’s going to have to be some pretty good artistry.

The second rule is that the dream should in some way matter to the story, even if it was just a dream. Does it push the story forward? Is it hiding something else that’s going to matter to the plot later on? What purpose does it serve? A dream without a purpose will likely be seen as filler, or [scene that could only have been possible through the dream] for scene’s own sake. You don’t want that.

Under normal circumstances, the third rule is that it shouldn’t be a way to try to dodge the implications of something that happened within the part of the narrative that was designated as just a dream. For one thing, it’s seen as cheating, and for another, sometimes people want to see where the new situation would lead. But this isn’t always the case. In a game, for instance, everyone agreeing that a certain situation is in nobody’s best interest can be an excellent reason to declare it a dream and carry on from there; likewise, people in semi-collaborative works (expanded universes and comic series usually) often use this as a patch to deal with differences in author style. (It’s usually more acceptable when a character’s creator is trying to salvage a serious case of derailment.)

And of course, if it’s not trying to be a twist, you can get away with a situation that’s clearly a dream easily; the characters might even seek it out.

I’ve seen a game in which “All a dream” was done successfully (if inadvertently): the GM had run a rather major plot twist that just about all of the players had some sort of problem with (power level, playstyle, general improbability, and a few other things), and midway through it was adapted into a dream sequence and finished as such. It led to some interesting results; the GM had to actively think about why we got the dream in the first place, one character was influenced to what we called “the shortest betrayal ever”, and we were using bits and pieces of the sequence as clues to later people’s behavior pretty much until the end of the replacement story arc.

Was it all a dream? Think about it carefully; answering yes may well be plausible.