Saturday, April 26, 2014

The SUE Files: Something with Numbers, I Don't Know

Okay; the first go-round on this was rather more confusing than I thought it would be, so I'm going to try to lay everything out a bit more logically.

In most games, combat works by blocking off a length of time as a turn, then having some system for letting everyone do (generally) an equivalent amount of stuff within that turn in some stat-dependent order. D20, for example, has its various types of actions. Other systems have a number of action points per turn, perhaps varying by a statistic representing speed. The point, though, is that they're grounded in relatively inflexible assumptions of how long a given action is going take.

I can't help but feel like we'd run into problems trying to limit our characters by time when they can just go faster; after all, they only obey physics as long as they want to. I'd rather limit actions per turn based on cognitive load and track our characters' efforts more directly. It feels more logical to me that way: if you're out of action points, it means you're too busy to do anything else that turn. That's why I called the action points Focus. It represents how hard a character is focusing on something.

So every turn, a character has an allotment of Focus points, which are then spent on particular actions. The cost of an action is ideally reflective of how much one has to concentrate to get it right; relatively simple actions, like jumping, might only take one or two focus points, while more convoluted maneuvers like, say, riverdancing might take considerably more. Maybe six or so. I haven't figured out a scenario in which combat riverdancing is likely to come up, but I'm fairly sure we'll find one somewhere. The point is, any combination of actions a character can fit under their Focus allotment is fair game.

Since Focus is effort, it makes sense that allocating more than the minimum Focus to a task might have better results. For simplicity's sake, let's say that dodging blows in melee combat takes 3 Focus and lets you roll 2 dice to dodge. Allocating 6 Focus to it would increase that to 4 dice, 9 would get you 6 dice, and so on. It doesn't necessarily mean you're moving faster (although it could); it's more a case of more carefully considering where to move. Similarly, allocating more Focus to driving a car isn't going to make the car go faster – but, since you have more dice to roll driving checks with, it does increase the speed at which the vehicle may be “safely” operated. (As an aside, we will have mechanics for bullet time at some point. I just wanted to clarify that running in bullet time is not required to accommodate extraordinary amounts of Focus expenditure.) This is also going to be where skills come into play; they're just dice added to the rolls for groups of actions as characters learn how to more efficiently do things.

The different kinds of Focus are intended as an aid to character diversification, so that people aren't just “good at doing lots of things.”
Physical Focus represents a combination of muscle memory and coordination. It's used for things like combat and acrobatics, where coordinated movement is required.
Mental Focus, on the other hand, is closer to concentration, or perhaps clearheadedness; it is intended to represent how well a character can think critically or abstractly.
Social Focus is a bit fuzzier, but it's loosely a combination of confidence and empathy, how well a character can appear as they wish to others, influence others' emotions, and so on. Strictly speaking this probably shouldn't be internal to a given character, but it simplifies bookkeeping.

In this way, it is possible to build characters who are suave or clever or graceful without necessarily implying they're all three.

Originally, I had intended for the three kinds of Focus to represent the maximum allocation of Focus to each type of action. This is overly complicated, though, and makes for a lot of needless bookkeeping. Instead, characters have three completely separate Focus pools to allocate, although odds are they'll only use one or two at the same time. Just to be simple, let's set a character's Foci equal to certain of their stats as follows:

Physical Focus = Finesse
Mental Focus= Genius
Social Focus= Charm
Immersion = Drive

If we keep “normal” stats generally in a range of five to twenty, we also keep Focus in a very manageable number range. I don't think it's too much bookkeeping to make sure the sum of a small set of integers is less than, say, eighteen – but if it is, we can probably simplify it a bit further later.

Now we have a way for our characters to take turns doing things. Every turn, every character takes actions of total cost less than or equal to their Focus allocations. We need a way to tell in what order they do them, an analogue to Initiative. I like the idea of going down a list of some measure of quickness in descending order, like Initiative, but I would rather it not be one roll and done. Instead, let's start everyone at the sum of their Mental and Physical focus; for familiarity, let's call it Initiative too. Instead of rolling stats, characters can spend either one Mental or one Physical focus to increase their Initiative by one for the duration of that fight, starting in the turn after they spend it.

An example may help. Consider three characters totally not named after variables: Alpha, with PF 5 and MF 5, Beta, with PF 6 and MF 2, and Gamma, with PF 2 and MF 4. On turn one, Alpha has an Init of 10, Beta 8, and Gamma 6, so Alpha goes, followed by Beta and then Gamma last. Both Alpha and Beta allocate all their Mental and Physical Focus to actual actions; Gamma allocates all of xis to increasing xir initiative, which is increased by 6 (2 Physical, 4 Mental) to 12. On the next turn, Gamma will go first with 12, followed by Alpha (10) and Beta (8). This order will persist indefinitely until one of the three spends Focus to change it.

So now we've got a way for people to do things at each other, and a way to decide in what order they do it. We need a way to track damage, and that can go back to Focus too. In a way, this employs one interpretation of hit points: they aren't literal health, but rather a representation of one's capacity to continue fighting. So, if we express damage in terms of Focus, we can run a lot of things through the combat system. Fists and bullets do Physical damage, depressants and arguments do Mental damage, and “drama” does Social damage.

Rather than make every point of damage reduce the relevant Focus by one, I would rather dilute damage down by some factor so we can have a wide range of integer damage values. Let's use the other four stats and say that every [Vigor] damage ties up one Physical focus, every [Acumen] damage one Mental focus, every [Nunchi] damage one Social focus and every [Stubbornness] damage one Immersion. That way we can track damage on a grid, like in Shadowrun, and use that as a simple way to tell how much allocatable Focus of each type we've got left – just mark off the right size grid on graph paper, fill it in left-to-right top-to-bottom, and the number of rows remaining is how much Focus is left.

And now that we have a mechanism for damage, the question of death naturally comes up. Zero Focus should not be death, but rather unconsciousness or equivalent; the character cannot take any actions relying on the relevant Focus, but is capable of eventually returning to some level of activity with minimal intervention. Permanent incapacitation of whatever type ought to happen at, say, [max Focus*-1]; in other words, a character has to take damage equal to twice the product of their Physical, Mental, or Social stats to die, or twice the product of their Immersive stats to break. Death itself is going to be rather varied, albeit scripted in the case of natives according to their home cosmology. Dreamjackers just fall apart; “death” is simply the absorption of so much damage that their mind loses the cohesion necessary to counteract the gradual erosion by the world and they cease. In effect, they doubt themselves to death. This does have interesting implications for recompiling and reloading them once we get to the meta-magic system, though.

Now, Cathexes were in the last post, and looking at them now they feel a lot like level adjustment. So forget them as written. I think they still have a place, though; there needs to be a reason not to just nuke all the chessmasters out there, or at least a way for defeating them on their terms to be viable. Purely speculative: they might end up working something like aspects from Fate crossed with gambling, something long-term but not involved in the XP mechanic at all. I'd like them to be player-defined things that a character considers integral parts of their identity, where they'd give bonuses to rolls but do Focus damage on failure. Say you have a burglar who's really proud of their ability to pick locks; they might get a +4 bonus to all lockpicking attempts, but take 4 Mental damage if they fail. Maybe a brilliant tactician gets a leadership bonus but takes damage when a subordinate dies. That way really forceful personalities might have very high Cathexes and be a more do-or-die. Let's say characters can have up to 10 points of Cathexis, divided however they like between skills. Incidentally, they're probably dreamjacker-only. They make things happen by wishing; this is a way of representing almost unconscious wishing. The idea that you can't stop reality bending to your whim is a bit scary, especially when suddenly it doesn't work.

So what doesn't make sense so far? Again, actions that cost Immersion will be added later, although it's probably somewhat apparent where they'll go.

And before I forget, the stats as follows:

Physical: Finesse, Vigor
Mental: Genius, Acumen
Social: Charm, Nunchi
Immersive: Drive, Stubbornness.

The first set of every pair defines Focus, the next helps define damage grid width. What the stats mean is still vague, beyond the name.

Friday, April 18, 2014

The SUE Files: Morphean Operandi

Now that we've kind of defined what everyone else does, it might be helpful to identify what exactly the Agency does. At their core, the Agency is a glorified cult, albeit not one revering a deity. Their central belief goes something like this:

Worlds are stories that flesh themselves out in the telling. Like stories, they have their own internal logic, operating above the mechanical cut and thrust of physics: narrative causality bridges the gap between what logic would suggest happens and what the story wants to happen. It does this subtly, manipulating the fundamental nondeterminism of the world in a thousand tiny ways, but it does so in certain recognizable patterns. When reality continues to obey these patterns, all observers are satisfied; when it diverges from them, defense mechanisms snap into place to deal with the narrative non sequitur. Like a cellular response to oncogenic mutation, these mechanisms operate in (rough) ascending order of severity, the final one being self-destruction.

In theory, the Agency operates to supplement these responses with more a more intelligently managed antigenic response; in practice, the only thing at all likely to resist until the world snaps is sustained, deliberate meddling to disregard causality in favor of rearranging the world to accommodate a new state, and so they fight SUEs. Their job is to find and isolate the pattern-breaking elements while guiding the integration of the broken patterns into the larger causal flow. In short, neutralize the SUEs, then cover everything up until no one's sure you've done anything at all.

This is a damned hard job. Agents are outnumbered a thousand to one by their potential enemies and a billion to one by the people they're supposed to guard, if not more. They have to listen to a brand-new symphony, identify the off-key player, and silently haul them off the stage before the song's over while said player is beating them over the head with their chosen instrument and the conductor will reflexively shoot them on sight. They have to do it almost totally without support, too.

They do have some help. To start with, they're as immortal as anyone who exists as a projection of their own self-image, aging being fairly subtle from moment to moment, so at least they never have to retire. They've also got forward observers in the form of all the natives, since the Agency briefs them through extremely informative dreams.

Cryptic dreams and symbolism, while classic, don't lend themselves to efficient internal communication. The Agency simply relays huge amounts of sense data to its operatives, having filtered it out of the natives' dreams, and lets them process it however they feel is most informative; the same adaptability that allows them to exist also allows them a certain facility with informatics. Most agents can pull full-dimensional pan-sensory records of the incidents; more adept ones can correlate out ever more remotely connected data to get a broader sense of their destination. These dreams are semi-lucid, as the agent cannot control the data itself but can readily impose their chosen reality filter on it, focus on some parts over others, and so on. Commonly chosen filters include a spy-style briefing, a conversation with a hooded stranger in a tavern, looking into a crystal hypersphere, or listening to a bunch of people tell stories at a party. Either way, enough information is available to avoid a completely blind jump, but only rarely enough to avoid surprises; likewise, the agent can talk across all senses with his teammates during the planning phase, but isn't actually capable of sensing them directly.

They then pick where in the destination world they're going to wake up (and as what) and there they wake. This does have to be a definite destination, not “right behind the SUE.” Thus the mission begins; at its end, whenever they next sleep, they will be informed of such, at which point they're on vacation until they are next needed.

Now, there are times when a team needs to be updated on the fly, or wants to request additional information when unable to sleep. That's why some enterprising agent invented calling cards: little objects designed to act randomly until an agent with a question uses them, at which point they nonrandomly display an answer. These have been magic 8 balls or ouija boards or knucklebones, but for some reason they somewhat tenaciously decide to be tarot decks in a lot of worlds, which accounts for the name. Whatever their form, they share a common weakness. Since they can't reveal their seemingly random results are actually messages, they also can't indicate when they're not: using it when there's no response just gives a random result that looks totally valid. Clever agents use them on definite questions with a small, discrete number of possible answers.

It's probably obvious that the Agency doesn't micromanage. They tend to keep together whatever groups naturally form, flinging them as one unit from world to world. Agents can reject missions by consensus, they quit by not wanting to keep going, and they're given so much running room some factions believe the Agency doesn't even exist except in the minds of its members. Indeed, most of the services the Agency provides are provided member-to-member; about the only thing binding everyone together on the operational level is a shared short-range telepathic “frequency.” As long as two Agency members are near each other, when one speaks, the other will understand and recognize the Agency “metadata” woven into it. With practice this can be expanded into pan-sensory speed-of-thought hivemind communication without actual speech, or even written down into a sort of augmented-reality tag, but everyone can at least manage to understand what their associates mean.

The archetypal Agency unit, the 2-6 member Ops team, tends to stick together and go where the missions are. Members come and go, yes, but it's an event; it's generally assumed that anyone joining a team is doing so indefinitely. Conversely, consultants move around a lot more. They might live in just one world to guide teams through its more abstruse elements, or they might offer occasionally useful services. They make the calling cards, for example, and send messages between worlds when the Agency's normal network will not suffice. Some of this is free to Ops teams; in the Agency service economy, saving the world is a pretty big service, and most consultants who contact the team will be prepared to help gratis. More involved favors generally involve some agreement to help make the same resources available to the next team.

Naturally, there isn't always a consultant available for whatever service a team might need; they're hired on an ad hoc basis whenever a prospective Agent isn't suitable for Ops work but has useful talents, so the resources to set up more than a skeleton network just aren't there. Thankfully there are others capable of picking up the slack. The grey market is big, diverse, and good at hiding; it winds like mycorrhizae through the less tightly controlled portions of the multiverse, in dark tavern corners and well after normal business hours and generally out of the way. Like the Agency, they haven't much use for conventional currency except as a prop, being generally able to short-circuit the economy whenever they really need to. Some take payment in data to sell on to in-world brokers; others take “curiosities”, unique and generally inconsequential bits of world valuable mainly as art. Many want seemingly minor changes in the world, especially from the ever-subtle Agency. While they have no commitment to Agency ethos – and in some cases enjoy pushing their moral boundaries a hair at a time – the grey marketeers tend to reject SUEs out of concern for their own safety. The enemy of their enemy is a relatively reliable client.

From here, I'm guessing people can assume the level of sects and violence they'd like, within and around the Agency, so I thought I'd end on an even more meta note and try to explain what I'm going for with them. The ambiguity surrounding their existence is intentional, of course, and it's intended to avoid giving the PCs a boss. They're their own boss....maybe. Everything else is really just there to give them choices; easily controllable options, in the case of consultants, and moral choices in the case of the grey market, which is also a way of fitting the Agency into a larger undercurrent of interesting folks, giving more places to pull PCs from, and so forth. Beyond that, it's mostly intended to get out of the PCs way and let them do fun stuff without necessarily reassuring them that they're doing the right thing. Remember, the only thing promising that the Agency is the only thing between the worlds and oblivion is...the Agency. Which might just be their own agency giving itself a capital letter. Maybe. Finding out directly is a bit of a job, though.

I don't know. Am I completely nuts here? Once I work out a system, going back in and expanding these powers is going to help elucidate the nuts and bolts; probably at the very least the really complex ones are going to require going to sleep to use so they can't be used in combat.

Thursday, April 10, 2014

The SUE Files: Beyond the Katanas

Why are SUEs bad?

It sounds like a trivial question, but we've split so many hairs that it's helpful to define exactly what they're doing that needs to be stopped. Clearly their methods are not at fault, since agents do the same thing; equally clearly, it's not being mean about it, since plenty of worlds are hellish dystopias 'naturally'. I'd be tempted to say that it's just breaking rules, but that has shades of Marty's M.I.C.

I think, if we wanted a simple distinction, we could say that SUEs break rules to break tropes, where agents break rules to fix tropes. Worlds, after all, are primarily places for stories to take place, and human pareidolia means tropes will always be a thing. Physics merely defines the cognitive space the story can run through; the important thing will always be the story itself, with whatever balance of logic, narrative causality and the Rules of Cool, Funny, and Awesome is necessary to drive it. So why have I so consistently referred to physics in the preceding post on Immersion? Partly because I'm a fool, of course, but partly because I wanted the tropes not to be explicitly the focus of the game. They lose something when the story focuses on itself so closely. That's what that giant 1-10 multiplier for “significance” is for: it's a way of rolling tropes and plot and local significance together into something a DM can throw around quickly. I'm aiming for a very fine distinction here: there are rules, and even when broken the story still follows them backwards, but it's possible to follow all the rules and still end up in a completely different place than you'd think. This is the best, handiest way I can think of to let players be adaptably genre savvy, and to materially reward that genre savvy, without forcing them to enforce “how the story is supposed to go” if they happen not to like it.

At least it affords us a relatively straightforward way of identifying SUEs-as-hijackers-of-stories. It's not just that they aren't local; it's entirely possible for non-natives to do whatever locals can without breaking Immersion, so long as they follow the tone. Non-natives can just cheat very easily, which is where agents come in; if the SUEs were only working in-world, the world could deal with them. Cheating is usually the difference between a weird event and an Immersion-breaking event.

As I see it, we have two basic kinds of SUEs: the ones who want to be something they aren't, and the ones who want the world to be something it isn't. The first resemble the classical Mary Sues, the kaleidoscope-eyed conventionally perfect cardboard cutouts. They might have katanas and trenchcoats, they might be goffik, or they might just really not want to die. In the end, though, they need people only as an audience, whatever form that audience might take. The second type need people as statistics: they want some change to the world, some new social system or physical reality or who knows what else, and they think people will be different for it (usually, but not always, happier.) These can hide better but generally aren't as resilient, since they rely much more on the active engagement of the locals.

This doesn't necessarily imply malice; it's entirely possible for them not to even realize they're doing it, or to knowingly do it in pursuit of a morally laudable goal. The least knowledgeable ones are ironically the most urgent threats, since without knowing about Immersion they can break it accidentally. Thankfully, without conscious shaping of their will they aren't capable of truly world-shattering feats, at least not directly; in any event backlash claims a lot of the ones the Agency can't intervene to save.

The knowing, benevolent ones are in their way even harder: it's hard to ask someone to give that kind of thing up, the locals generally like and protect them, and their effect on Immersion is harder to guess because people so want to believe what they're doing is real. The miracle healers and rebel leaders and so on are always tough. They are, however, promising candidates for recruitment.

If individual SUEs are a headache, their organizations are a leading cause of agent migraines; something about their hierarchies lets the crazily egocentric rise to the top. Some examples to get us started:

The Justice Poets: in essence, they believe poetic justice is a fundamental universal constant; anything exceptionally good or bad happening to anyone is ultimately some sort of karmic response to some vice or virtue they possess. Unhappy with the degree to which the universe validates their beliefs, they pick people they believe to be undeservedly happy and hurt them in some way “ironically” related to their perceived failings – indeed, they consider themselves sculptors of fate. Their sense of proportionality has somewhat degraded as of late, though; they're down to the level of kidnapping people who waste printer paper and feeding them feet-first into woodchippers, then claiming they were “hoist by their own petard” and perhaps leaving a eulogistic limerick on their graves. The old guard are slightly more restrained, still cleaving to the old custom of leaving their targets alive to despair over their fate.

The Chiaroscuro Cowboys apply the idea of justice slightly differently: everything has to have a hero and a villain, if only you dig deep enough. They see it as their job to find the villains and bring them to justice. Naturally, anyone doing this job is heroic, and therefore anyone disagreeing or inhibiting them is the evillest sort of fiend. Somewhere along the way they found a Western to their liking, what with all the lynching, and the name stuck. The newer members are actually almost reasonable, so long as they're operating in a world with sufficiently black and white morality. Just don't ask them about mercy.

Animancers represent the opposite end of the spectrum: they have no real moral views other than a firm belief in the rightness of the customer. Somewhere between soul merchants, chess grandmasters, and sculptors, they simply make minds to order. They might be in the employ of natives, might be filling out some transreal menagerie, or might simply be playing one of their Games. Autonomy, as they call it with some amusement, is played like so: events are engineered such that one preordained person (the Pawn) will, at some future time, have to make a choice of great input, one with a finite number of options. Each team is assigned one of those options and the goal is to manipulate the Pawn into choosing it. Generally speaking, direct effects are not allowed within some considerable radius of the Pawn's person, simply to make the game more intriguing. The death of the Pawn before the appointed time or similar failure to get them to the choice is considered a loss all round; taking a third option, so to speak, is scored as a draw. Variants include Fatalis, in which the Pawn's choice is always fatal; Brotherhood, in which there are multiple Pawns; and World-in-the-Balance, using a dreamer as the Pawn and manipulating only things within their dreamscape. Animancers with high-end Elo ratings get better-paying and more interesting commissions, so the Autonomous Games are played for very high stakes. Rumors of Autonomy variants using dreamjackers, or even the Agency, as the Pawn are of course totally unfounded.

I had been worried that we'd be restricted to cackling “for-the-evuls” villains, but there are apparently alternatives within easy reach, particularly when combined with the corrosive effect of localized omnipotence described several posts ago. I'll come back and add more groups and types and things later, and I'd very much like to see what other people come up with. Next time I'll sketch out Agency procedure.

Tuesday, April 8, 2014

The SUE Files: Needing a Bigger Boat

 Another hilarious guest submission, this time by Joural0401:

 This story isn't about a bad DM or anything of the sort.  It's just… it needs to be shared.  We had an average party at the time, and were in the middle of confronting a pirate king.  Sorcerer, Rogue, Fighter, Barbarian, Cleric.  Pre-gunpowder, so ballistae are the primary weapons for boats.  The entire party is onboard the Pirate King's boat, and he's reaming out the pirate captain who led us here(apparently he's done these kinds of things in the past).  Eventually it's decided that the pirate captain will have a chance of surviving, so they put him on a rowboat, and the DM immediately starts describing the captain frantically rowing away.

Apparently, the thief never read anything on TV tropes, so half way through the description of the captain in the horribly rickety boat, he says "I jump on."

The entire table stands silent for a moment, before the Fighter says "I follow him."

I(the sorcerer) try to stop them, but can't convince them and I fail my grapple check.  So they're on the boat.  The Pirate King turns to look at us, makes a comment about how our friends like to gamble, then starts talking about how the Big Bad has access to all these amazing innovations of weaponry.





Three cannons, all pointing at the little rowboat.  The first one fires.  Miss.  Second one fires.  Miss.  Third one fires… hit.  Reflex save time.  The both make it, so the fighter takes half damage and the rogue takes none.  The DM pauses, and asks for a second reflex save.  Again, the Fighter makes it, but the Rogue fumbles, and is now bleeding 1d3/round.  In the ocean.  The rogue fails a swim check, so the fighter swims down and saves him.

Around this point, I realize what's coming next while the fighter is crowing "We survived!", though given the status of the thief at the time(he got a 1 on a fort save, so now he's bleeding 1d4), might have been premature anyways.  Then the fins appear.

Sharks between them and the boat we came from, sharks between them and land.  Fighter swims for a turn, but decides he'll have a better chance if he drops the rogue, and the rogue barely makes it to swim under his own power.  He's still bleeding out.  Fighter swims off, rogue decides to swim in the other direction because… hell if I know.  A turn later, he fails a swim check and a fortitude check, and ends up unconscious underwater, while the fighter swims a bit further.  The DM is generous and gives the Rogue a fortitude save to have one last chance at surviving.  Says 'you'd basically need a natural twenty to wake up.'

Entire table silent for the fateful roll.

Everyone leaning in.


It was the perfect moment of cinematic you'd-never-believe-it-in-a-movie, but the rogue wakes back up, no longer bleeding, underwater.  "make a swim check."


So the rogue is dead.

Meanwhile the fighter has realized, at long last, you can't outswim a shark, so he draws his weapons to try and fight them off.  He fumbles his first roll, and his sword is out for the round.  Dagger time.  Rolls for it… fumble.  Both attacks miss.  Shark turn.  One of the sharks charges forward to attack.  Critical hit.  The fighter had 3 HP left following the cannon hit- now he's at -11.

So both 'let's get on the boat' people are dead, end of the story, right?

Wrong.  Because I made the mistake of intimidating the Pirate King.

A round later, I was blind.  The cleric tries to guide me to cast a spell but when it fails I just enlarge person the Cleric(phase one of any combat at this point is, of course, to enlarge the barbarian), climb on her back, and tell her to swim like her life depends on it.  We hop in the water(sharks are better than the CR like 10 or whatever guy on the deck) and swim to the boat.  Making it on board, the session ends with the captain of our ship asking where the rogue and fighter went.

None of us knew how to answer until the Barbarian says "Well, last thing I heard them say before the sharks showed up is 'I got this guys,' so I'm sure they're fine."

End of the night, end of the best session ever.

Monday, April 7, 2014

The SUE Files: How the World Doesn't Work

We keep talking about Immersion, but I don't think we've ever even approached an understanding of it. So here's the Theory of Absolutely Everything, so to speak.

Readers accustomed to trawling wikipedia may be familiar with the idea of the brain in a vat; that is, a mind artificially stimulated by a supercomputer into believing it is experiencing reality when it's really just floating in a little vat of nutrient solution. As a philosophical tool, it commonly assumes a perfect simulation. We must assume a thoroughly imperfect simulation – which, indeed, is far more likely. We have people intervening in worlds all the time as a fundamental part of their operation, let alone all the ideological leakage seeping around. To avoid minds fouling themselves up on problems with the sim, then, we need some way to letting them deal with errors without rejecting it entirely. That is Immersion when used according to package directions: a universal coping mechanism to let imperfect operators run imperfect simulations without them eventually breaking.

In keeping with the theme of people being awesome, we've been assuming that people are both aware of a huge amount of information and capable of processing it in real time at a nearly unconscious level. This, in turn, feeds pattern recognition of things they don't even realize they're sensing. That's what Immersion is intended to protect: the sense that those patterns, be they laws of physics or the whims of some divine agency, are persistent. In a perfectly deterministic universe, that sense would never erode even slightly, but naturally universes have people and random elements and so forth. Immersion lets people ride out the bad function calls and the rounding errors and the divine intervention.

Indeed, it serves several useful purposes even without universal cross-contamination. With sufficient pageantry, it's a sense of the miraculous. In small doses it can stimulate curiosity like nothing else: it empowers people to conquer the nonsensical, the fantastic, and the bizarre until nothing is completely beyond comprehension. Unfortunately, the very qualities that make Immersion so useful also make it difficult to manipulate. On a functional level, it is a network of independently controlled redundant processors and exception-handling heuristics all integrated at a fundamental level into the core cognitive patterning engines called upon to instantiate self-aware entities, minds included; in other words, it's everywhere, so it can't be thrown out or shut off, but it also can't be replaced or refilled. (The numerical representations of Immersion are, of course, purely for convenience; one “point” of Immersion could mean any number of things in neural space.)

As I said above, Immersion as a function of time is normally noisy but functionally flat; it is constantly depleted by random errors in computing but constantly restored by everything else. Every second they're experiencing anything, the natives experience most of the world working as they've come to expect, and moreover informing them implicitly that they are existing correctly, if only in a physical sense. This acts to smooth out all the little bumps, and the big bumps, and generally drag the world back to accepting itself as real.

Non-natives, on the other hand, experience just the opposite: not only does nothing work quite right, but it keeps querying them in strange ways, so their Immersion slowly erodes as the nagging feeling they aren't what they think they are slowly colors their awareness. Even the most perfect of them have to stop and translate their sensory input and construct their output, and this leads to a thousand little glitches too small to individually matter but collectively big enough to require them to find another source of Immersion or suffer a break.

For dreamjackers and SUEs alike, then, Immersion works fundamentally differently than it does for natives: it's the assertion that they are what they think they are despite the constant protestations of local reality that what they think they are does not exist. In more poetic terms, they're the only reality they have left, and so they become literally realer than they've ever been to compensate. (Incidentally, this is also why neither of them dream. They're essentially dreaming of themselves constantly just to exist.)

With that in mind:

Immersion is not purely rational; the more something impresses itself on a given observer the more resources they devote to it and the harder it is to isolate before something snaps, which means a bigger bump on the aforesaid graph. To refer to an overused example, this means that a laser pistol hurts the medieval peasant's Immersion less than a laser cannon, even though they work on exactly the same locally nonsensical physical principles. Moreover, people being the emotional creatures they are, particularly emotionally impactful uses hurt more as well; seeing the aforementioned cannon vaporize a random rock hurts less than seeing it vaporize his house. Emotional depth also sucks up cognitive resources, magnifying the response that has to be swept under the cognitive rug. Smart non-natives use weird stuff at one remove from anything huge or important.

Furthermore, that magnitude is significantly affected by the observer's current emotional state. Properly prepared, most minds can swallow a lot more than they can by surprise. This is why gods can enact sweeping changes with relative ease; the pageantry and lights and choirs and so forth act to prep the minds in attendance that something weird is happening, thereby providing an easily rationalized explanation. The same is true of technobabble, making it an extremely valuable skill in sneaking nonsense under the audience's collective noses without anything breaking. Gimmicks take second place behind making sure no one sees anything, but it's a close second.

On the subject of making sure no one sees anything, logical people might assume that simply being out of sight of everyone might work – that is to say, simply popping off to the core of the planet or depths of space is sufficient to allow them unlimited operational freedom. Unfortunately, the simulation itself is always watching, albeit through slightly different eyes. It sees non-determinism, at least on whatever crazy terms it's been told constitute deterministic physics, which means that empty space is positively the worst place to hide from it; with no one doing anything there, logically everything should go where the laws of nature dictate it should, and errors can be swiftly identified and corrected. However, where people are concerned, suddenly errors are constant and logic at best a very strong suggestion. People generate a sort of groundscatter of exceptions, and under cover of that it's harder to detect that non-natives are there, let alone active. This means that, paradoxically, dreamjackers are safest in populated places or when embroiled in major events, particularly those central to the story the dreamer itself is telling – the general confusion of so many active minds can hide nearly anything, so long as it's hidden from the more mundane observers. It's generally easier to hide things from non-omniscient observers anyway.

So what all of this means is that Agency dreamjackers are playing two games at the same time. The first is a simple numbers game: they need to act in their own way often enough to keep themselves from breaking, but present those actions in such a way that no one else snaps from witnessing them. At the same time, they need to weave their actions into the larger story to hide the inevitable errors from the simulation itself, or else the sim will break – which breaks everyone as time itself grinds to a shrieking and asynchronous halt, causality just gives up and physics starts randomly guessing until Lovecraftian horrors arrive. They also need to keep the far less limited SUEs from doing either of these things and avoid a messy death at their hands in the process.

We'll get into the specifics of the mechanics later, once I've ironed out the massive mechanical problems inherent to a system that needs to encompass so much, but for now: every potentially weird event has some kind of base Immersion cost to make it happen based on the magnitude of the effect, like Endurance in HERO6. This is then multiplied, per observer, by some number from 1 to 10 based on the emotional and rational impact. Event-native observers have their Immersion increased by that much; everyone else gets it decreased by that much.

In practice, everyone but the DM has only one multiplier to keep track of, while the DM has at most maybe three; since everyone but the SUEs and the PCs is a native, the multipliers should be identical unless an exceptionally impactful event bumps it up by one or two. It's very wingable, so to speak; the last thing I want is for this to drag down an otherwise eventful scene.

So, in conclusion, the kind of thing we're saying works well is meeting the setting's problems (to wit, the SUEs) with solutions that bear the hallmarks of each agent's background, then carefully picking when to employ the more fantastic elements of their arsenal to maximize effect while minimizing Immersion loss. It's entirely possible to go out of (in-character) character, but that severely drains their ability to keep dealing with the weirdness they're constantly experiencing, so in general you want what your character does to restore Immersion so that what others do can drain it without breaking. If you're a pulp fiction strongman, solve problems by punching; if you're a master physician, go heal some folks. It's very much a case of everything looking like a nail when all one has is a hammer, except the people you're working with only have a saw, a screwdriver, a duck and solipsism, respectively, and the nail hammers back, and there's every chance of a noise complaint delivered via reality breaking.