Over in the thread [URL=http://www.indie-rpgs.com/viewtopic.php?t=11363&postdays=0&postorder=asc&start=0] Emotion Mechanics and Losing Control of Your Character[/URL], the question, "is it okay for a system to tell you how your character feels and what he/she does as a result" got a lot of "sure, sometimes" and a lot of very intense "hell no, that should be roleplayed out."

Which is great, but now I want to flip this topic around and look not at the negative side of emotion mechanics -- losing control of your character -- but at the positive side: emotion as power.

Pain that makes you angry gives you power -- power to strike back. (Look at the climax of Return of the Jedi, or real-world soldiers going beserk at the death of a comrade). Horrifying sights that make you afraid gives you power -- power to run like hell, or to fight like a cornered rat if you can't run. A kind word that inspires hope gives you power -- power to carry on. (Look at the Sam-Frodo relationship throughout Lord of the Rings).

Most game systems ignore emotions altogether. The Riddle of Steel, with its Spiritual Attributes, and Unknown Armies, with its Obsession/Rage/Fear/Nobility triggers, get at this idea of passion as power-up -- to a degree: They depict permanent features of your character's personality that come into play in specific situations in predictable ways; if one of your buttons is pushed, you always get the bonus for acting accordingly. But even these games don't really get into the highly volatile reactions people actually have to stress. Being hurt, for example, may scare you into running away or anger you into fighting back; having the object of your secret crush actually ask you to dance at the prom may cause you to freeze up or suddenly gain confidence.

And to preempt the inevitable counter-argument, "just roleplay it" doesn't address the issue. If emotional reactions are a source of power at least as important as a character's skills or strength or equipment, they need to have an equal impact in terms of game mechanics. (Even giving bonuses for good roleplaying (a la Sorcerer) doesn't quite achieve this). What I'm talking about, in essence, is making emotional reactions a "power-up": not something which dictates specific behavior, but which makes certain courses of action much easier -- which presents players with the interesting dilemma of surfing the waves of their characters' emotions or swimming against the current.

The question is, obviously, how? What games have people played, read, or designed that address this? And what are the pitfalls of doing this wrong (or, for those who dislike the entire idea, of doing it at all)?

Well, you asked of our own designs...

Yep, I've done that forwards and backwards. The simplest system was probably Planet Linkola, my game of learning Finnish and colonizing space: all character qualities are represented by simple descriptors ("strong", "mechanic", whatever) which determine what a character can or cannot do. When there is doubt as to character success in a task he could conceivably try, emotion statistics are consulted: these are simply numbers from one to ten for different character emotions, a roll under with d10 required to succeed. Each use of an emotion would either rise of lower it's value depending on certain mechanics. If multiple emotions are applicable, roll all and get at least one success. Simple, efficient and more or less simulationism-supporting.

The above was of course the result of trying about a dozen different ways of doing the same, and is by no means the ultimate answer. Another design of mine, the superhero game Power over others applied emotions exactly like other traits: a character could be "Strong 5" as well as "Scared of magic 5" or whatever. The only difference between emotions and actual qualities was that while qualities could hardly be increased (genre expectation), emotions were freely customizable by the player. However, at times a character with strong emotions would be forced to willpower checks, so the player had to be ready to really play according to the emotion if he wanted it to have a really high value.

I've surely made other games with emotion mechanics as well; the IGC contest springs to mind, wherein in my game The Fall of Atlantis and Dawn of Human History I used character emotion as a kind of pacing mechanism: players only gained dice needed in conflicts through demonstrating character motivations (chosen by the player), so they had to pace their conflicts suitably to have a decent chance at success. In another game, the Brotherhood, emotional ties of the characters defined the magic powers they could draw from their connection.

There are of course many published designs with various emotion mechanics. For some reason MLwM hadn't been mentioned in the last thread the last time I checked, although it evokes both emotion-as-limitation and emotion-as-power efficiently. The characters are only defined through their Self-hatred, Weariness and Love, and that's that. Some times the emotions are useful, some times they tie the character to oblications he'd rather avoid.

HeroQuest, on the other hand, again handles emotions exactly the same as any other attributes of characters. Pool, Universalis and any other games with freely customizable skill lists generally do this as well. The basic capability exists in half a dozen of my own games without even being especially mentioned. It's like, two of my three IGC games had the option, and I didn't even think about it. Completely instinctual for a certain kind of design.

IMO emotions are at this stage a fully realized concept in game design, and trying to limit them outside in any aspect is really only defensible for design reasons, not because "player/character freedom is paramount" or other such nonsense. Many games have proved over and over how emotions can be considered.

It's simply not useful to consider questions like "Is it OK for somebody else to tell how your character feels?" or "Can emotions be used as stats?" They're just not even questions without defining some context: for roleplaying games in general the answer is obviously yes, because in some games that's the whole point. In almost any single game the answer is either yes or no, depending only on the game in question. In some it's stupid and wrong, in some others necessary.

Interesting that your mention Return of the Jedi... Over the course of playing Star Wars campaigns for nearly ten years, my group and I have developed a way that we include emotion into game play, without "just roleplaying it." For the most part, roleplaying a character's emotions is like roleplaying him walking into town. A character may become angery or sad or withdrawn and the effect of his emotion on his actions is fairly obvious.

However, when the GM or sometimes the player, realizes that emotion that the character is experiencing is more than that of the "day to day" variety, a mechanic is brought to bear. In the old WEG d6 system, this was simply a matter of determing which attributes the emotion effected, rolling 1d6 and adding/subtracting that number from any rolls in those attributes for the duration of the emotion. Fairly simple, fairly boring, but effective enough... Under the D20 system the same idea applies, except that the penalty/bonus is applied to Vitality. (of course, there are other effects, like if you're a Jedi, anger can gain you Darkside points, etc., but we have house rules that detail these...)
So, IMO, emotion should be a resource that has both positive and negative effects, when tapped into. When to tap into it, what the effect should be, and how power the effect is probably a matter of taste and system.

I've got to disagree with you, Eero. I don't think the question of what it takes to appropriately include emotions in rules has been reduced to obvious tropes. There's a lot of theoretical ground-work still to be done.

Emotions are much more complicated than skills and abilities. They are tightly linked with issues of drama that are still only vaguely understood.

Suppose your character has a Hatred of Aliens: +5. He obviously gets +5 when he's all hateful of aliens, busting carapaces and the like.

And there are also situations where it is clear that he should get a -5... trying to be polite to an alien he dislikes, for instance.

But the fun (and IMHO important) situations are far trickier than that. What about when he's forced to work with an Alien for three days uncovering an assassination plot. And then he has to choose between siding with an important human behind the plot or his erstwhile xeno-partner.

Should he get a -5 to help his partner, a +5 to help the villain? Or should no penalties apply, because it's no longer the generic situation. Or (my opinion) should he get a bonus to almost anything he chooses to do, because he's been wrestling with the issue of his prejudice in the face of a worthy individual, and the emotional resolution of that (whichever way it plays out) is powerful?

I think that if all Sydney is asking is "What does it take to use emotions as stats?" then he's exploring very worthwhile territory. He gives some great examples of things that are hard to represent using traditional game-mechanics: that whole thing about being asked to the dance at the prom is (IMHO) a gold-mine of possible exploration. It is neither easier nor harder for your character to accept the date... I think that, statistically, all you could say is that it should be judged on a gaussian curve that is far less weighted toward the center: extreme reactions seem more likely than middle of the road.

This looks like it should be a very interesting thread. At some point I should probably chatter about the emotion mechanics in my own nascent game, Capes, but I've already written quite enough just now.

ErrathofKosh wrote: So, IMO, emotion should be a resource that has both positive and negative effects, when tapped into. When to tap into it, what the effect should be, and how power the effect is probably a matter of taste and system.

TonyLB wrote:
This looks like it should be a very interesting thread. At some point I should probably chatter about the emotion mechanics in my own nascent game, Capes, but I've already written quite enough just now.

Eero Tuovinen wrote:
HeroQuest, on the other hand, again handles emotions exactly the same as any other attributes of characters.

TonyLB wrote:
Emotions are much more complicated than skills and abilities....extreme reactions seem more likely than middle of the road...

Eero Tuovinen wrote: Planet Linkola, my game of learning Finnish and colonizing space

ErrathofKosh wrote: Interesting that you mention Return of the Jedi... [details of nifty barebones mechanic snipped]

Resource mechanics (hero points, miracle points, whatever) are a way of letting the players decision to take something seriously translate into game mechanics.

It is, in short, a way to have "just roleplay it" that also allows for those type of scenes. Which is not to say that there aren't other, possibly better, ways... just to point out what's already out there.

Sydney Freedberg wrote: And to preempt the inevitable counter-argument, "just roleplay it" doesn't address the issue. If emotional reactions are a source of power at least as important as a character's skills or strength or equipment, they need to have an equal impact in terms of game mechanics. (Even giving bonuses for good roleplaying (a la Sorcerer) doesn't quite achieve this). What I'm talking about, in essence, is making emotional reactions a "power-up": not something which dictates specific behavior, but which makes certain courses of action much easier -- which presents players with the interesting dilemma of surfing the waves of their characters' emotions or swimming against the current.

The question is, obviously, how? What games have people played, read, or designed that address this? And what are the pitfalls of doing this wrong (or, for those who dislike the entire idea, of doing it at all)?

Tony, Sydney: again, I have to disagree. The examples of differences between emotions and other things all assume a fundamentally simulationist approach to game design. Emotions are only different if you want them to be.

I of course agree in that there sure is yet much to explore in emotion mechanics, but then, that's so with wound simulation, too. They are just memes of imagination one and all, and can be twisted and turned in different ways to produce the effect the game designer wants. You all write about emotions like they were some discernable entity that works like this or this, but it simply ain't so. Emotion is a bonus to roll for one game, a sacrosanct area of roleplaying for another, source of bonus dice for a third, a statistic for a fourth... for some games it's nothing at all, for those games do not concern themselves with emotions.

Sydney wrote:
To reproduce this in a game (and I know I'm sounding very Simulationist here, but I think this is GNS-neutral), you can't "just roleplay it," because a character is not just making different choices (which are arguably the province of the player, not of the mechanics), he/she is actually more effective -- which the mechanics have to reflect somehow.

Eero Tuovinen wrote: The examples of differences between emotions and other things all assume a fundamentally simulationist approach to game design. ..... Strictly speaking the mechanic of drawing strength from emotion of course isn't of any specific agenda. Your motivation for including it seems to be, though - if the reason for using emotion mechanics is because they're part of certain fiction, that's sim.

The Ars Magica, Pendragon, and Theatrix examples of applying passions, possibly through the expenditure of some kind of plot points, are good ones -- BUT because they generally do the same thing in a reliable way, often under direct player control, they still don't address the volatility of emotional reactions (e.g. the drive to the extremes that TonyLB talks about). They don't hand the player a ticking time bomb of emotional energy and say, "Now what?"

So let me refocus my question more precisely: How might a game portray the nature of emotions not only as a power-up, but as an inherently unpredictable and even dangerous one?

The goal here would be to make a character's emotional reactions as complex, unpredictable, and exciting as, say, combat (in a well-designed system, obviously, not a hit-point-plinking "I hit you, you hit me, repeat" system), or demon-human interactions in Sorcerer. The hope is for players to have as much fun with the emotion mechanics as they would with combat mechanics in a more traditional game.

Sydney, in the game Hero Quest, characters have "personality traits" that are handled precisely like any other ability. That is, any ability can be used "positiely" when appropriate, or "negatively" when appropriate. So, if my character has "Bouts of Rage" or something, then when he's angry, he can use that to augment, say, attacking somebody. When he's trying to calmly argue with somebody, this becomes a penalty.

As to what's "appropriate" essenetially if anyone feels that it's appropriate, then it's appropriate. Because "feeling" is what drama is all about. That is, the problem solves itself in play as people activate these things when they feel that it's dramatic. Which automatically makes it so. In actual play this works brilliantly every single time.

Seems simple, and solved to me.

Mike

The system of Dying Earth is the obvious answer, as social situations in it are handled largely the same as combat. Character emotions are only partially controlled by the player, and the system is certainly just as interesting as combat.

Another one is Myrskyn aika, which I already mentioned. The idea is that using emotion increases unpredictability. Simple simulation. You'd get the same effect in many systems (say, roll-under, higher is better) by making the size of the die used dependant on character emotion: while the results will be higher, there is some higher possibility of failure due to emotions. In some situations it's even better to be cold and non-emotional.

Or how about this: players can at any time declare an emotion their character is laboring under. When this is the case, the character gains an appropriate emotion bonus in the conflict at hand, but the GM narrates character action. This gives the player a choice: to harness the power of emotion, but risk unexpected results when the emotion is unleashed. For another kind of game the GM could even base his narration on some kind of sincerity roll that would tell him when to screw the character over and when not. Or it could be made a matter of trust, so that the player would have to consider in each case the hidden motivations of the GM.

Mike Holmes wrote: Sydney, in the game Hero Quest, characters have "personality traits" that are handled precisely like any other ability. ... So, if my character has "Bouts of Rage" or something, then when he's angry, he can use that to augment, say, attacking somebody. When he's trying to calmly argue with somebody, this becomes a penalty....Seems simple, and solved to me.

Eero Tuovinen wrote: Or how about this: players can at any time declare an emotion their character is laboring under. When this is the case, the character gains an appropriate emotion bonus in the conflict at hand, but the GM narrates character action.

Eero Tuovinen wrote: The system of Dying Earth is the obvious answer, as social situations in it are handled largely the same as combat.

Sydney Freedberg wrote: The Ars Magica, Pendragon, and Theatrix examples of applying passions, possibly through the expenditure of some kind of plot points, are good ones -- BUT because they generally do the same thing in a reliable way, often under direct player control, they still don't address the volatility of emotional reactions (e.g. the drive to the extremes that TonyLB talks about). They don't hand the player a ticking time bomb of emotional energy and say, "Now what?"

So let me refocus my question more precisely: How might a game portray the nature of emotions not only as a power-up, but as an inherently unpredictable and even dangerous one?

Sydney Freedberg wrote: The goal here would be to make a character's emotional reactions as complex, unpredictable, and exciting as, say, combat (in a well-designed system, obviously, not a hit-point-plinking "I hit you, you hit me, repeat" system), or demon-human interactions in Sorcerer. The hope is for players to have as much fun with the emotion mechanics as they would with combat mechanics in a more traditional game.

John Kim wrote: My experience has been that player-controlled emotions can be highly unpredictable and I still find them interesting.

John Kim wrote: OK, a topic question. I think you've got at least some idea of the sort of mechanics which you are looking for here (i.e. dice-using mechanics with numbers for emotional states and/or personality traits). Are you interested in general for ways to explore and portray emotion in role-playing? This will involve a compare-and-contrast of numerical mechanics against non-quantified approaches. Or are you more interested in detailed design of the mechanical approach?

Sydney Freedberg wrote: How might a game portray the nature of emotions not only as a power-up, but as an inherently unpredictable and even dangerous one?

Sydney Freedberg wrote: How might a game portray the nature of emotions not only as a power-up, but as an inherently unpredictable and even dangerous one? *snip* The hope is for players to have as much fun with the emotion mechanics as they would with combat mechanics in a more traditional game.

I'm not sure that if this response answers anything brought up in this thread or not, but it occurred to me while reading the previous posts, so here goes.

I once played in a game with an ultra-realistic combat system. Big time. Getting shot usually meant a severe wound with months of bed rest or death. The system went into detail about where you'd been hit and what kind of damage you'd taken (so you get stuff like fingers blown off, static shock, collapsed lungs, etc.) The players and characters had to deal with the often devastating effects of combat-related injuries and wounds.

The interesting effect of this ultra-realistic combat system was that it obviated the need for a mechanic for emotions in combat: the players quickly became just as afraid of fights OOC as IC.

I think this is different than "just roleplaying it." The game essentially, through game mechanics other than blunt-force "emotion mechanics" fostered an emotional state in the players that was reflected in how they played their characters.

So, one way to foster appropriate emotional reactions in player characters is to have game mechanics which impose consequences that help create a similar emotional state in the players as that their characters would have.

I think part of the reason that some people feel there is a need for emotion mechanics in some games is that the existing game mechanics don't support the emotions the GM and/or game designer thinks the characters should have.

For instance, in, say, D&D, the player of a high-level character, being familiar with the game mechanics, knows that his or her character has a gazillion hit points and is unlikely to get killed in combat. So, the player, and his or her character, isn't afraid to charge into battle. Is this just a metagaming (I don't know what the Forge term is for metagaming, if there is one) problem? I don't think so. Because, by the time a character gets to be high level, he or she would have figured out what the rules of his or her world are and would know from past experience whether he or she is likely to get seriously injured or killed in a given combat.

So, ultimately, the GM/game designer who feels that a high-level D&D character should be fearful about and during combat isn't taking into account that the combat mechanics defuse the root cause of fear in combat. In general, combat isn't scary just because it's combat. Combat is scary because other motivated agents are actively trying to injure, disable, and or kill you who may succeed. If a game's combat mechanics lessen the threat of death or injury and make it unlikely that a character's opponents can succeed in killing or injuring him or her, then, even though the character is in combat, he or she will not be scared. I think sometimes we confuse "combat is scary" with "someone trying to kill or hurt you who might be able to do so is scary."

Emotions, after all, are basically reactions to things. When, in a game, as GM or designer, we remove the thing to which people are really reacting, naturally, the player and the character fail to have expected emotional responses. Removing the danger and leaving the combat, and then imposing fear upon the character is kind of incoherent, actually.

Anyway, this is just one way to deal with emotions in games. I'm not saying that it works all or most of the time. It won't work in all kinds of situations: for instance, in horror games, there's only so much a designer or GM can do to invoke fear in the players without the them calling the police. And there are definitely ALL KINDS of problems with ultra-realism in games, not just GNS stuff, but also games in which ultra-realism isn't desired or appropriate (say, like in a Star Wars game). I don't even really like ultra-realism most of the time (I play Amber, after all). It was fun that one time, but I wouldn't like it in general. I'm just saying that, from my experience, one way of helping deal with emotional issues in games is to have the game support those issues with its mechanics. I think this is subtly different than the kinds of emotion mechanics being discussed, so I thought I would throw my thoughts in the ring.

M. J. Young wrote: I'm envisioning what might be called a "sweet spot" resolution system. I'd gear it such that I want to roll as close to a target number as possible, but not over it. ....the emotion, if activated, must be called into play before the roll, and it adds its value to the roll..... exceeding the target number creates "overdone", and you messed up because your emotion carried you too far.

Just to throw in a random comment that is vaguely on-track with this topic...

A while back I was tossing around a "Black-Jack" style relationship mechanic for when you meet people, because I was feeling very cynical at the time and thought it was an apt analogy. But basically it goes like this: You play Black-Jack, with real cards, playing against the character you wish to form a relationship with. The closer you get to 21, the higher your feelings towards that person. If you go over, you fall in love. You take the difference between your result and that of the other person, and use this to determine the mis-match of feeling. If one goes over 21, and the other does not, the relationship is doomed. If both go over, both fall in love. If neither go over, the relationship will forever be a friendship or mere association (depending on the value of the result).

When making rolls or whatever, you suffer the inverse difference between your result and their result to any rolls against that person. So that if your result was 13, and theirs was 19, then the difference is +6, so you suffer -6 to your rolls against them, because they are much more motivated towards you, than you to them. "Against them" includes trying to help them.

If your result is below 10, then you begin to not like them at all (because a result of 10 or less in Black-Jack means you are either stupid or not wanting to win). So below 10, you gain the proportional difference between your results as a bonus to rolls against them, because you are motivated by dislike of them.


Anyways, that's just an idea I had which is using some of the ideas expressed here in some way, and since I discarded it a few months ago, I thought I'd throw it out here. The main problem is that it's too bipolar and linear, with a line between love and hate.

I think I'd be intensely interested in seeing mechanics for relationships/emotions which factor the interactions of multiple emotions and their likely effects, as well as the individuals ability to control certain emotions. The best idea I've been able to come up with is a "disposition chart", which is a numbered lsit of emotions. You pick one emotion to be your "default" disposition, and then circumstances would require you to roll to "fluctuate" about that disposition. So that a normally happy person would require something incredibly strong to make them angry, whilst the reverse would be true for someone with an angry disposition. Then you could have individual biases, which would effect how easily you moved towards certain dispositions, acting like a "magnet" of sorts for key dispositions, to try to emulate how some people seem drawn to a few key emotions more than others. However, it all gets very complicated and becomes alot of bookeeping and chart-tracking. This becomes even more compounded if I throw in the ability for relationships, where some people have differing effects on a character's disposition than others, manifest as modifiers to rolls based on results of previous rolls...... very complex. Definately a task for a computer.

But that's enough of my rambling.

-Ben

Ravien wrote: A while back I was tossing around a "Black-Jack" style relationship mechanic for when you meet people, because I was feeling very cynical at the time and thought it was an apt analogy. But basically it goes like this: You play Black-Jack, with real cards, playing against the character you wish to form a relationship with. The closer you get to 21, the higher your feelings towards that person. If you go over, you fall in love. You take the difference between your result and that of the other person, and use this to determine the mis-match of feeling. If one goes over 21, and the other does not, the relationship is doomed. If both go over, both fall in love. If neither go over, the relationship will forever be a friendship or mere association (depending on the value of the result).

The best idea I've been able to come up with is a "disposition chart", which is a numbered lsit of emotions. You pick one emotion to be your "default" disposition, and then circumstances would require you to roll to "fluctuate" about that disposition.

God, I love this forum. There are more good ideas every time I turn around.

Okay, everyone. First, troop over to the latest thread on TonyLB's draft superpower game, Capes. Now bow down before the awesomeness of TonyLB. I know I do.

Now, Tony's thoughts in Capes about powering-up by going into emotional / motivational debt, which you can then gamble your way out of or deeper into, and various ideas on this thread (especially M.J. Young's blackjack-style mechanic), all congealed last night in a sleep-disrupting burst of what I think is insight. To wit:

Gambling.

If you want to do justice to the volatile, unpredictable nature of human emotion, you make invoking emotions in the character require a game-mechnical gamble by the player. (This is related to a key insight of Tony's on the Fear and Confusion thread, that to create the proper feel of a combat situation, the players need to share some of the confusion of their characters). That is, instead of emotion being "just another stat," it's a stat that requires you to put something at risk and which increases not only your chances of success but your chances of disaster as well.

This actually goes beyond emotion mechanics. Let's say we divide all stats into two categories: the things that make you Cool, and the things that make you Heroic. The Cool is what you can always count on -- your skill with a blade, your rapier wit, your dashing good looks, your superhuman strength, etc. -- all the things traditional games include as stats / advantages / skills. Most of the time, your Cool abilities will carry you through: You use Cool to impress routine contacts and plow through mooks. Cool abilities work predictably and reliably, with little or no risk of disaster; in game-mechanics terms, either they work on a pure Karma system (I have a 6, I needed a 5, I succeed) or on a Fortune system with no critical failures, no critical successes, and a narrow range of variation.

But when the dramatic crisis hits, your Cool abilities may not be enough. (Here I'm inspired by Tony's Capes). Now what makes a hero heroic? It's not about Kewl Powerz: It's about laying yourself on the line for something you really care about -- it's about risk. In a combat situation, you gamble life and limb; in a social situation, you can gamble respect (both what others have for you and you for yourself); and in any situation, you can gamble your emotional drives -- your hope, your love, your ideals, your self-esteem. (More Capes). These Heroic Gambles add on top of your Cool abilities, but they use different mechanics: They operate off a Fortune-heavy system with a high range of variance and a serious chance of both critical successes and critical failures (preferably, it should be possible to get both at once).

And if you blow the roll, you lose some or all of what you've gambled. Did you stake your body by running out under fire to grab your wounded buddy? Bam, you got shot. Did you stake your image by trying to impress The One Your Love at the prom? Whomp, you fell down and everybody's laughing. Did you put your heart on the line for something you really cared about? Crack, your heart just broke and that love or idealism or sheer desire is gone. Presumably not all of it forever -- there should be rules for recovering from all of this -- but enough of it and for long enough to hurt.

How to implement this exactly? I'm not sure. One way might be a dice pool where the dice contributed by your Cool abilities act normally but the dice added by your Heroic Gamble can explode into critical successes and failures. (Note: I don't like dice pools, so this is a sacred-cow-shooting moment for me). Or you could have Cool abilities be pure Karma, as above, and then roll for Heroic Gambles to see if they pay off (upping your Karma) or blow up in your face (losing whatever you staked) or both at once (you succeed but at a terrible price).

Note that, especially in a grittier, more realistic game, the player is not always able to choose whether to stake something or not; if someone's putting a gun to my head or dangling my love interest from a bridge, my body (in the first case) or heart (in the second) is being gambled whether I wanted to or not.

So a whole lot of actual design work to do, including (for me) ripping up pretty much all the mechanics I've drafted for My Eventual Game. (Sacred cow massacre). But making players gamble when their characters are putting heart, soul, and body on the line seems profoundly right to me. Or does this make even half as much sense to everyone else as it does to me?

Damn! I wanted to point everyone to Tony's game... It is a very good (if unfinished) example of this topic in action. I too am butchering sacred cows. Here is a thought on how a mechanic could work that involves "raising the stakes":

A character's skills and abilities are defined by karma, and in non-emotion-grabbing situations, these resolve conflicts. However, in situations that become a little more involved, the player (or the Gm) determines how emotionally involved his character becomes. The emotional involvement is defined by levels, each represented by a die type (borrowing from the Window). The player rolls two of the appropriated dice of different colors. The player then is allowed to pick one to add to his appropriate skill karma, and the other determines what happens to him emotionally. Higher is better for success, lower is better emotionally.

Thus, the more separated the values are, the better the character has handled the situation. When the values are close, the character has either not added much to his skill via emotion, or he has, but in doing so, he has caused himself great emotional damage. As the stakes go up, the potential for great success goes up, but so does the possibility of emotional backlash.

Just a few thoughts... Probably needs a lot of polishing and such.

Jonathan

In response to my roll-under suggestion, Sydney Freedberg wrote: Downside: It could make applying any other modifier to the roll tricky (e.g. you can't just add +20 for higher skill, because then higher skill also creates the possibility of overdoing it), so it requires further thought. But I'll play with it a little.

Mike Holmes wrote: ...in the game Hero Quest.... the problem solves itself in play as people activate these things when they feel that it's dramatic. Which automatically makes it so. In actual play this works brilliantly every single time. Seems simple, and solved to me.