Infallible Precognition versus Player Agency (and super heroes)

[James_Nostack]

A comics friend I got to talking about Alan Moore versus Steve Ditko, and (like a fool) I said, "I would totally run "Watchmen '66" as sort of a mash-up between Moore & Gibbons and the Silver Age antics of the Charlton originals."

The adaptation, using Marvel Heroic Roleplaying, has been extremely smooth: the game's rules allow you to, for example, inflict direct harm on someone (stress), inflict indirect harm on them (complications), or create things that lend support to yourself or others (assets).  And there are ton of weird little tricks you can play with the game's economy, so that even some of the odder quirks of the characters are pretty easy to implement.

Except I'm having all kinds of trouble with Doctor Manhattan's ability to perceive his personal future.  In the comics, this is Alan Moore completely blowing away mainstream comic book storytelling conventions.  I feel that there must be an RPG mechanical equivalent, but I lack the analogous genius to see it.

Here's how this thing supposedly works in the fictional world of the comics:
* Doctor Manhattan is experiencing his whole future and past simultaneously
* He therefore has perfect foreknowledge of his future
* But apparently neither he nor others have the ability to change it

As a literary device, it's mainly used to demonstrate the character's alienation from the human condition, his abdication of personal responsibility, and of course foreshadowing of things to come.

Folks on StoryGames and RPGNet have been pretty strongly against the idea of using Dr. Manhattan as a player-character at all, for several reasons.  One: he's unbalancing.  (My response: anyone volunteering to play in a Watchmen game knows that there's going to be a balance issue, and I'm not convinced it's insuperable anyway.)  Two: eh, just use his time-knowledge as a piece of incidental color since it never amounts to anything meaningful.  (Bah!  This is perhaps the most crucial part of the guy's characterization, and it would be cool to give it serious mechanical meaning.  I want this guy, mechanically, to make the player feel Manhattan-like.) 

So the problem I'm running into is that Dr. Manhattan's always correct about his own future, but even armed with that foreknowledge apparently nobody can resist it.  It's essentially introducing a railroad about a specific future event.

The closest thing to a solution I've come up with so far is, assuming the player succeeds at the relevant die roll, allow Dr. Manhattan to frame that scene, burning through resources to ask the other players leading questions about their involvement.  In Marvel Heroic, normally only the GM can frame scenes so this is a pretty serious grant of authority, and in some sense should feel like having a "deputy-GM" at the table when the power is active.  Which, I think, is kind of what playing alongside this guy ought to feel like. 

But if you have other thoughts--preferably ones easily translated into this game system--I'd be happy to hear them.  (I don't need to hear: "This guy shouldn't be a PC" or "This is really just incidental color."  Others have made those arguments already.)

[Ron Edwards]

Hi James,

I'm wrestling with this one because I would probably have (irritatingly) provided a long discourse on incidental Color, if you hadn't headed that off.

So, I'm thinking back on the comic - not the movie, not Watchmen fandom, just the comic, full stop. Here are some of my notions.

1. My impression is that Doctor Manhattan can't scan the future and past at will, for any topic of his choosing. Instead, as he experiences life, he perceives the entire continuum for the experiences of that moment. There's still a sense of wave-front at work there; he genuinely feels the feelings of that moment. Arguably, he never anticipates in the sense we think of the word.

2. His perspective is wholly self-centered; he can't shift it out to see what happened or will happen in places he wasn't standing in, e.g., to look for clues at a crime scene. He can't know what others will do in the future when he won't be there to see it. Putting #1 and #2 together, for instance, he doesn't simply look ahead and know that Ozymandias was/is going to disintegrate him (I don't buy the idea that he knew but let it happen because he also knew it wouldn't hurt him) - instead, it happens, that's all.

3. There are lots of things he doesn't know and therefore he can still use observations and deductive logic in an ordinary human way. For instance, unless I'm totally mistaken, he didn't know who Laurie's father was until she herself figured it out.

So that makes things a little easier. And you have very crucially specified that your design goal isn't to simulate Dr. Manhattan's power/condition, but is rather to generate the aesthetic/experiential content for the person playing him. Which is totally cool.

Hurm, as another character in the story would say. Might there be a way to permit back-story editing and retroactive "so this is how X was made to happen," or even, "Y will happen because of what is happening right now" type input from this player? I would even posit that we can use the GM-task concept productively here: so the real/big GM still does all the scene framing and weaving and other similar structural stuff, and has total authority over anything Dr. Manhattan doesn't experience, but that player has a useful content-adding function, including material for future scene-framing, which is definitely GM-grade as far as traditional role-playing standards go.

It amuses me that the other players might become as irked at the Dr. Manhattan player as the characters in the story do at Dr. Manhattan.

Best, Ron

[Eero Tuovinen]

I wrote a lengthy treatise on parallel worlds and observation bias, but the argument got convoluted, and ultimately it's all speculative anyway. I'd rather go straight for the point:

I think that you can take care of Dr. Manhattan's precognition in your average roleplaying game by allowing the player the following powers:

• Dr. Manhattan may arrive in any scene whenever the player feels like it. He may also make retroactive preparations for any scene he participates in, including bringing/sending others to participate as well.
• Dr. Manhattan may make pronouncements about the future, but these pronouncements are veiled in play until they actually happen. That is, you can say that Dr. Manhattan describes the future to others, but you can't tell what he says until the scenes in question are played and/or otherwise narrated as per the ordinary rules of the game. If the player fails to follow this rule of veiling and Dr. Manhattan's prediction does not come true, then he lied when he said those things that he did before.
• Dr. Manhattan is outside of time both physically and mentally (indeed, he does not have a real physical existence as vividly demonstrated in the comics). This means that every time he comes into a scene the player may choose any version of him to arrive. What this means depends on the system details, but mostly I'm thinking that there is no reason whatsoever for injuries or attitudes to carry from scene to scene with him. He is eternal and immutable, essentially, for the presence that he excerts on the world around him originates in a single point of choice, the moment of his birth.

The reason for why I believe the above powers sufficient is a complex argument about authorial observation bias. The key point is that we're arbitrarily choosing to observe a world where Dr. Manhattan acts like a human and does not cause a time paradox. Nothing in his own nature as a being who observes his entire time-line simultaneously certainly forces stuff like slowly reconstituting a body for himself after his birth (to pick a dramatic example from the comics), but the rules of good story-telling do: the comics would not have been published if Dr. Manhattan had not acted within certain parameters. For rpg purposes I suggest that a proper emulation of Watchmen should enforce a similar limitation on Dr. Manhattan: he cannot ruin the game because we're specifically choosing to observe a time-line where he doesn't ruin the game.

The reader may be interested in why I allow Dr. Manhattan to appear in any scene the player wants. Doesn't this mean that he is omniscient? My interpretation is that the causal reason for why Dr. Manhattan chooses to be anywhere at any given moment is in all of his experiences over his personal time-line; this means that Dr. Manhattan knew to teleport or walk or whatever into this particular scene just now because he knew when starting on his journey that he would arrive at this moment in time. There is no omniscience involved, it's just that Dr. Manhattan often acts for causal reasons where the cause of his action is in the future, and disconcertingly often is nothing more than his own knowledge of what he is going to do in that future. I'm pretty convinced that in-fiction his "choice" to do anything at all is arbitrary and ultimately caused by observation bias: there is no ultimate cause for why he does the things he does, except that we are choosing to observe a world where his actions as suitably dramatic and human-like. In this case that choice is being made by the player of Dr. Manhattan: as a co-author he thinks that it would be interesting to have Dr. Manhattan appear into this scene.

Another power worthy of comment is Manhattan's precognition. The reason for why I don't allow the player to actually make prophecy is epistemological: Dr. Manhattan cannot warn others of things that might come to pass, he may only speak absolute truth (unless he lies or speculates like anybody else, of course). If we're talking about the absolutely last word on what is going to happen in a given scene of a roleplaying game, then the proper procedure is not to allow a single player to declare it - it has to be played according to the proper systematic process. This is easiest and most practical to achieve by leaving Dr. Manhattan's predictive ability color only, with no factual content for the players to worry about. The same point is obvious when we consider situations where a player character summarizes the events of a past scene, it's just with this special case where I even need to explain why a scene needs to be played before it can be summarized.

Note that I could say that Dr. Manhattan may force a scene from the future to be played so he can utilize knowledge about that scene's resolution here and now. However, doing this would cause time paradox if anybody utilized the resulting knowledge to avert the scene that was played out of order. Because we've admitted to observation bias ("Dr. Manhattan shall not cause time paradox, for it would break genre!"), we know that this cannot happen. As storytellers you therefore have to play in a way that preserves enjoyable story. I choose to achieve this by allowing the player of Dr. Manhattan to retroactively declare preparations for any scene when it comes along in the normal chronology: sure, Dr. Manhattan knew in scene #3 that some event would happen in scene #5, which is why he did, indeed, do this particular thing then that I'm only just now revealing. It's just the most feasible way to handle this sort of stuff.

Of course, the cherry on top is that Dr. Manhattan's ability to "make preparations" could strictly speaking also cause paradox or break the integrity of prior scenes altogether. However, in practice we've seen that similar abilities work just fine in roleplaying games as long as players don't go out of their way to break them, and the rules system perhaps puts some limits on what constitutes dramatically acceptable "preparation" and what's just avoiding scenes altogether. The technical narratological reason for why such abilities work is that we never map and narrate everything about what characters do, so it doesn't break the logic of the story to make a surprise reveal later that yes, indeed, Dr. Manhattan came here in between his last two scenes (or even simultaneously, in his case) and made some fanciful preparations that we just didn't talk about before now.

Also, of course, Dr. Manhattan's nature seems to be that of a person who mostly doesn't concern himself with the fine details of how to abuse his powers. I, of course, assume that this is because the author is choosing to observe a world where Dr. Manhattan is a detached and mysterious figure. It just makes for a better story than one where Dr. Manhattan is omnipresent everywhere.

[James_Nostack]

So, I'm thinking back on the comic - not the movie, not Watchmen fandom, just the comic, full stop.

All I'm interested in. 

I'm wrestling with this one because I would probably have (irritatingly) provided a long discourse on incidental Color, if you hadn't headed that off.

Please indulge yourself: it would give me the opportunity to argue why I don't think the character's precognition is incidental color at all, but instead the li'l thematic motor of that character, a huge plot McGuffin, and an incontrovertible trump card in the ethical debate played out in the novel.  (But that's tangential to my main purpose here.)

1. My impression is that Doctor Manhattan can't scan the future and past at will, for any topic of his choosing. Instead, as he experiences life, he perceives the entire continuum for the experiences of that moment. There's still a sense of wave-front at work there; he genuinely feels the feelings of that moment. Arguably, he never anticipates in the sense we think of the word.

That's correct, and supported by the text.  "To me, it's already happening."  He's experiencing his entire lifetime (or relativistic world-line) all at once.

2. His perspective is wholly self-centered; he can't shift it out to see what happened or will happen in places he wasn't standing in, e.g., to look for clues at a crime scene. He can't know what others will do in the future when he won't be there to see it.

That's also correct, though I don't know an exact cite.  We do know that he knew about JFK's assassination in 1961 because he overheard TV coverage in 1963, so he knows what others will tell him even when he's not a direct witness: hearsay, I suppose.

Putting #1 and #2 together, for instance, he doesn't simply look ahead and know that Ozymandias was/is going to disintegrate him (I don't buy the idea that he knew but let it happen because he also knew it wouldn't hurt him) - instead, it happens, that's all.

He actually states that he doesn't know what's going to happen next, and thanks Ozymandias for reminding him of the experience.  Chapter XII is probably the only one in the book where Jon experiences something like what we consider free will.

3. There are lots of things he doesn't know and therefore he can still use observations and deductive logic in an ordinary human way. For instance, unless I'm totally mistaken, he didn't know who Laurie's father was until she herself figured it out.

The "until" in that last sentence gets twisted around a bit, due to his living everything at once, but yeah. 

Note in that scene, though, how Moore isn't really playing honest with us, for story-telling reasons.  At one point in that Chapter, Laurie gets frustrated and asks to know how the argument ends.  Jon answers something like, "With you in tears."  This leads both the reader and Laurie to expect that her pleas on humanity's behalf will fail.  Tension!

But in fact, Dr. Manhattan could have answered, "It ends with you in tears, because you find out who your dad is.  But holy shit, your mom is soooooo fucked up, it's a level of the human condition I couldn't even imagine!  Christ on a crutch, I guess that generalizes to everybody.  I should stop feeling sorry for my blue omnipotent ass, and stop them nukes."

There's a few other scenes where Jon tells similar half-truths, which I attribute to Moore's needs as a writer.  But it could be that he isn't especially curious about his future since it cannot be changed, and when he does "look ahead" he just goes with the first impressions.  I dunno.  He's pretty selective about when he uses it.

And you have very crucially specified that your design goal isn't to simulate Dr. Manhattan's power/condition, but is rather to generate the aesthetic/experiential content for the person playing him. Which is totally cool.

Marvel Heroic Role-Playing is sufficiently traditional that those two goals aren't too far out of alignment, but creating a "Doctor Manhattan experience" is more important than getting it exactly right.

Might there be a way to permit back-story editing and retroactive "so this is how X was made to happen," or even, "Y will happen because of what is happening right now" type input from this player? I would even posit that we can use the GM-task concept productively here: so the real/big GM still does all the scene framing and weaving and other similar structural stuff, and has total authority over anything Dr. Manhattan doesn't experience, but that player has a useful content-adding function, including material for future scene-framing, which is definitely GM-grade as far as traditional role-playing standards go.

Yeah, that's kinda where I'm headed with this, to the extent I can cram it into the core mechanic of the Marvel Heroic game.  The question is how far you can go with a co-GM who's also a player, without bumping into the Czegge Principle or other weird problems.

[James_Nostack]

Incidentally, over at my blog some fan-boy blogging occasioned by a close reading of the text, and part 1 of a somewhat deeper (though still kinda shallow) look at the work's themes.  Hope it's not violating forum policy to list.

[glandis]

As long as what you're looking to capture here can include some of what the reader experiences with Dr. Manhattan, here's an idea: at the start of each scene/sequence/whatever, tell the Dr. Manhattan player (publically, I'd think) what the mechanical outcome on his character is going to be. Detail might be flexible - I'm not familiar enough with the MHRP system to know exactly how to frame this (e.g., could you say "you gain a new asset"? Sometimes  - flexible detail - even you gain a PARTICULAR new asset?), but it seems likely to be workable.

I've used something like this in a fantasy/prophecy context, and it worked well for that group.

[Ron Edwards]

Hi James,

My thought on the "It ends with you in tears" issue comes in two parts: my take on the source fiction and my take on the game-design question you've posed.

1. As I see it, Jon was able to answer Laurie's question at point A in the continuum because she asked it. As her question to him at that point is genuine in-A input, then he's able to see it "ahead" - but in his terms, in his interpretation of the question. He's such a needy fuck (or arguably, such a romantic) that all he cares about is how she feels, not about the content of her question as she sees it, i.e., the debate over his participation. So that's what he sees - and I suggest, not because he looks, but because he always sees the whole continuum as it pertains to that very moment alone. And his response is so neutral because he can't react, he doesn't know it yet as an experiential thing ... until the time comes for it to happen. Until point B happens, even seeing it from point A doesn't make it happen for him, even in anticipation. As a side note, Laurie's interpretation of her reply is characteristic of her, hearing X but perceiving XYZ.

I make no claim actually to interpreting The Watchmen in a fashion that anyone should pay attention to. I only offer this because it's the foundation for my suggestion - more generally, because I do not interpret Jon as Omniscient Temporo-Man, Master of Time, that I think the fanboy discussions may have graded into.

2. So, in play ... if the player can drop bombs like that one into play, then it becomes a constraint for upcoming play - I don't see how that would be any harder than the what-are-they-called, the Interlude or whatever moments in InSpectres, when we shift to a post-event interview with a character about the scene which is about to be played. Players quickly learn to provide information in such scenes which is not narratively descriptive or which pre-determines outcomes, but rather deliver opinions and judgmental content: "If only Bob weren't so reckless!" or "Damn, this hurt!" (pointing to a bandage on their head). This energizes everyone regarding content for the next scene to be played, but doesn't actually establish that content in a boring, play-it-out way - instead, working it into the material becomes fun as hell and even sets up for precisely those reversals of expectation that happens in this exact example.

(All that said, I agree that Moore plays fast-and-loose with the metaphysics of his story in order to generate maximum thematic obviousness, sometimes to the point of dunce-cap haranguing. The very existence of the comic-in-the-comic device is nothing but a big red underline for the basic story, for instance, until we find Ozymandias is having visions of the pirate comic, of all things. But now I am geeking out - must cease!)

Best, Ron

[clukemula]

Hi James,

Have you thought about something similar to Archipelago's Destiny Points?

In Archipelago, each player writes one Destiny Point for each other player character, an event that will happen in that character's future. Once everyone has written these events, each player gets to pick one as their character's Destiny Point for play, which all other players can see throughout the entire game. These events constantly drive play in a certain direction, but none of the players know exactly how they'll be played out.

So my character, a big-shot Wall Street banker who is having intense problems at home with his family, could have a Destiny Point that says "He kills his father." Me and the other players know that this is going to happen (though my character probably doesn't), so all of the story ends up being focused to see how and why my character kills his father.

I think this could work well with Dr. Manhattan (having other players write events for him), since he would perceive these events in his future, knowing that they will inevitably happen, but not knowing them experientially until they do happen.

The only thing I'm not sure about with this is how to incorporate his sense of knowing when the events will take place, if that's even important to emulate.

- Luke

[bankuei]

Hi James,

An easy thing to do is to have Dr. Manhattan's memory jumps be rolls to make Assets which can be used by anyone who is involved in the future events.  If you want to make sure you're not stepping on toes along the way, you can set these up as directed, but open questions to the players involved - "At some point in the future, you're in trouble - what kind of trouble?"

As a fun sort of balancing factor, probably a good complication that can come of them is when Jon misinterprets the situation or gets lost in something that's really not relevant or simply drifts too far from humanity ("Oh Mar is awesome! Oh shit, you need oxygen, sorry!")

Chris

[James_Nostack]

My thought on the "It ends with you in tears" issue comes in two parts: my take on the source fiction and my take on the game-design question you've posed.

1. As I see it, Jon was able to answer Laurie's question at point A in the continuum because she asked it. As her question to him at that point is genuine in-A input, then he's able to see it "ahead" - but in his terms, in his interpretation of the question. He's such a needy fuck

Yeah, I think there's definitely something to this.  He in theory knows all this stuff (the way I remember the girl I had a crush on in second grade) but doesn't need to consciously call it to mind unless somehow prompted (Judy Barstow!).  And there may be an element of, "He'll answer exactly what you ask, so be careful how you ask it" type of thing, which I worry in play would lead to D&D players lawyering their wish spell.

As a side note, Laurie's interpretation of her reply is characteristic of her, hearing X but perceiving XYZ.

To be fair to Laurie, total nuclear war is the premise of their conversation, and sometime during the discussion Jon mentions that at some point soon his vision dims, possibly from mass nuclear detonations.

2. So, in play ... if the player can drop bombs like that one into play, then it becomes a constraint for upcoming play - I don't see how that would be any harder than the what-are-they-called, the Interlude or whatever moments in InSpectres, when we shift to a post-event interview with a character about the scene which is about to be played. Players quickly learn to provide information in such scenes which is not narratively descriptive or which pre-determines outcomes, but rather deliver opinions and judgmental content: "If only Bob weren't so reckless!" or "Damn, this hurt!" (pointing to a bandage on their head). This energizes everyone regarding content for the next scene to be played, but doesn't actually establish that content in a boring, play-it-out way - instead, working it into the material becomes fun as hell and even sets up for precisely those reversals of expectation that happens in this exact example.

Yes, I think that's approximately how I'm going to try to run it in playtest, and if it flubs I'll go with Bankuei's suggestion of Assets, which is the more conventional, by-the-book approach but I fret won't be as play-twisting as I'd like to be.

(All that said, I agree that Moore plays fast-and-loose with the metaphysics of his story in order to generate maximum thematic obviousness, sometimes to the point of dunce-cap haranguing. The very existence of the comic-in-the-comic device is nothing but a big red underline for the basic story, for instance, until we find Ozymandias is having visions of the pirate comic, of all things. But now I am geeking out - must cease!)

The thing that bugs me is that Adrian goes to all this trouble to muck up Jon's vision of the future--yet also has access to his psychological evaluations, which would surely reveal that even if Jon knew of what Adrian was up to, he wouldn't do shit about it.  In fact, Jon's ethical sense, to the extent he has one, aligns with Adrian's pretty closely.  Driving Jon mad with the cancer scheme, and hiding things with the tachyons, is basically a pointless detour.

(Assuming, of course, that Adrian had access to perfect information.  I guess two ways to rationalize this would be, Jon really "pin down" the future in an immutable form until it occurs to him to look, so driving him to Mars is a way of keeping Adrian's plans open to chance.  Or: Adrian doesn't trust that others would be as impassive as Jon if Jon shared his foreknowledge of Adrian's plans.  Even if their interference fails, it's still an unexpected element to fret over.)

[James_Nostack]

An easy thing to do is to have Dr. Manhattan's memory jumps be rolls to make Assets which can be used by anyone who is involved in the future events.  If you want to make sure you're not stepping on toes along the way, you can set these up as directed, but open questions to the players involved - "At some point in the future, you're in trouble - what kind of trouble?"

Yeah, I think that will be my back-up plan if the more ambitious scene-framing thing falls apart in testing. 

As a fun sort of balancing factor, probably a good complication that can come of them is when Jon misinterprets the situation or gets lost in something that's really not relevant or simply drifts too far from humanity ("Oh Mar is awesome! Oh shit, you need oxygen, sorry!")

Yes, I'm including that as a limit: when the Precognition fails, opponents can turn it into a Complication: Lost in Thought for free, or step it up for a plot point or a doom die.  Which would put it at a d12, one step away from completely immobilizing him for the rest of the scene, and forcing any allies to try to snap him out of it. 

Chris
[/quote]