Character Death Mechanics?

[Mike Holmes]

Disagree.

When Indiana Jones is captured, is that a bad thing from the mind of the moviegoer? When Bond's wife is shot, is that a bad thing to the person in the audience sympathizing with that character? No.

The protagonist does not always succeed. But in all cases we feel for him. That is the only requirement. That the player have some reaction to the things that affect the character. Then it's all good, no matter what happens. Even if the character dies for some good reason, the player will be satisfied. When Bond's wie dies, it's a great thing from the player's POV. Sure it sucks from the character's POV. But for the player, now he has an excuse to get medieval on the badguys.

It is possible to have "negative" repercussions for the character that are all "positive" from the player's POV. The only way this would not be true is if the player were actually the character. To quote Mel Brooks, "Comedy is when you fall down a manhole into a sewer and die. Tragedy is when I cut my finger." The player is not the character, and can (and always do) regard them as a separate entity who's story they are watching unfold.

So the "connection" is one of observer. Not one of being. This is critical to understand. Tension always comes from the potential that the character can fail. Thus it does not matter that the conflict causing the tension has a failure condition that is death, or any other loss. Just that there is one, and that it's dramatically important. In that case, no matter what the outcome, the player will be pleased.

The only way this is not true is if the failure does reflect back on the player. That can only happen if he is invested personlly in the conflict. That only happens when there is some player metric involved. Hence Gamism. In Gamist play, failure means that the player has failed, and that sucks (but one accepts it as a neccessary part of competition). In Narrativism the player has no such stake. The only thing that is important to the player is that something dramatic happens. And since failure is as dramatic as success, failure is never less pleasing than success (and often quite a bit more pleasing).

This is why it's very dangerous to try and mix Gamism and Narrativism. Either the player will invest in the competition, and therefore be dissatisfied when he fails or does not find a decent challenge, or he will promote failure inhis character, and be punished for it.

So, which is it, Sparky? Do you want Gamism or Narrativism?

Mike

[Sparky]

Hi Mike,

Most of the tension will not come from the characters being killed off. Really. And I like the NPC guideline you mentioned.


Hey Damion,

I haven't nailed down all the particulars, but the conflict resoloution mechanic should provide for some fun consequences. For example, I'm going to avoid wound penalties to actions and any sort of death spiral. Those kill the fun for me, so to speak.

One kind of result I'm imagining: "Success and Minor Advantage" which someone (player or GM) will use to narrate a result description. Maybe the PC uses the small advantage to have slightly better Initiiative next time or leap to a better position or a small landslide of rubble crashes between the foes, giving the PC a headstart on running away from the foe.

One of the aims of this is to give the players who don't GM the chance to get used to creating interesting consequences both positive and negative. Kind of like a meaured narrative mechanic.


Back to Mike,

Do I want them to be detached observers or personally invested? Pick one or the other?

My answer is that the individuals constantly demonstrate both and so I expect both. I know it sounds like heresy, but bear with me a sec.

A given player will sometimes limit their actions simply for the sake of character. Sometimes they'll want to succeed in the worst way, gaming for all they're worth. Sometimes they'll act in the best interest of the storyline. Sometimes they want to see how the dice fall and take it from there. It goes a long way towards satisfying them to allow them their choice in approach, and it gets more narrative as time goes on.

I'm basically trying to let the resoloution system do the work and let everyone ad-lib off it within the bounds of setting. Sometimes it will be positive, sometimes negative but almost always interesting.

I set-up a few things ensuring the sim needs of a multiple GMs in the setting and a few things that are gamist for just the simple pleasure of dice rolling fun. Being true to the characters'abilities and personalities results in a reward. That leaves the GM free from most system judgement calls to use their energies in-game to be the most creative with the storyline, or so my theory goes.

I think that I can have both Gamism and Narrativism if both are a bit limited. I don't see them as mutually exclusive. I won't really know how successful it is until it's playtested for a while. Hopefully I won't have to eat any crow. ;)

Thanks for your thoughts, guys.

Sparky

[M. J. Young]

I think that I can have both Gamism and Narrativism if both are a bit limited. I don't see them as mutually exclusive.

You don't know how much I hate to be the one to say this, but:

I don't think you can accomplish this this way.

Gamism is very much about playing to win; narrativism is very much about exploring the premise of a moral, ethical, or otherwise personal issue.

Perhaps it might help if I suggest (at the risk of upsetting someone) that there's a difference between telling tales and telling stories. All role playing games create tales. You can go back and retell what happened, and sometimes enjoy very much describing how your team faced the dragon and brought him down, or fought the space pirates until you managed to escape in the life boat just before the ship exploded. This is telling tales. You can similarly suggest that playing just about any game or sport tells similar tales. The announcer at the baseball game relaying the events on the radio has his audience locked on to the tale he tells. The play-by-play recounting of a chess match may be a very tense and interesting tale to the right reader. But none of these things are stories.

Gamism may tell great tales, tales you'll tell again and again, about how you won or how you lost. But they don't, in the main, deal with the same kinds of personal issues as telling stories. It is never about whether it was the right thing to kill the dragon, or blow up the space ship. It's about whether you succeeded in the objective.

Look at the movie Dragonslayer. There's a tale there of a sorcerer who kills a dragon. But the story is about many other things--about a girl who has been hidden in plain site as a boy for her entire life, so she won't be sacrificed to the dragon; about a princess who discovers that she has been shielded from the risk of the lottery while all the other girls were at risk, and so sacrifices her own life; about the apprentice who thinks he has power and discovers he has little; about the faithful servant who doesn't know a bit of magic but understands his master's intention of coming back from the flames to kill the dragon, and dies trying to make that happen. The story is about all these things, not about killing the dragon. The quest to kill the dragon is in this case merely a catalyst to create the situation in which the issues are explored.

Now, when you're writing a book or a movie, you have control of the outcomes, and so can make the tale come out the way you want even while exploring the themes of the story. But when you're playing a game, you have to make choices between whether to do what is best for the tale or what is best for the story; that is, do you run this in such a way that it's about facing the challenge and winning or losing, or do you run it so that it is about those other issues?

I am convinced that you can play in a game that is gamist, narrativist, and simulationist; but I don't think you get there by limiting any of these. You get there by cutting away the things that limit them and allowing the game to find its center through the nature of the setting and the desires of the players. A game can't really be partly narrativist, partly gamist, and partly simulationist, I suspect. It must be fully dedicated to each of those it wishes to support, and prepared to let them fly. If that means that the last session was a gamist challenge to rescue the princess from the demons but this one is a narrativist conflict about whether to marry her or continue doing what the character does best and the next is a simulationist exploration of the lands around the kingdom, maybe that's a way to get there. But you can't really have a limited gamist effort to rescue the princess and a limited narrativist conflict about whether to marry her; that's nonsense. Maybe you can design a game that has controlled drift between them, or which manages them in some other way; but you can't really have a game work which doesn't at some point know what the player goals are, even if that changes.

I hope this makes sense; I'll probably get trounced for something here, but really, "limited narrativism" and "limited gamism" don't make sense to me, even apart from trying to combine them in one game.

--M. J. Young

[Sparky]

Hi MJ,

I see no one trounced you. ;)

Perhaps I've used some terminology incorrectly, misunderstood responses and/or not expressed myself clearly. I certainly didn't mean to. If that's the case, you all have been very patient with me and I'll apologize right now.

I also tend to speak in generalities and apply general principles instead of focusing on just the particulars so I when I say something like 'story,' I mean 'a connected series of events.' On this forum certain words like story might have connotations that I'm not aware of or think to specify. Being aware of this, I usually try to offset this with examples.

I've acted under the impression that Narrativism was about story/theme and Gamism about accomplishing objectives and Simulation about being true to a particular source or setting.

IME, Gamist goals can be woven into a storyline and be dramatically brought to a Narratively satisfying climax, with or without an emphasis on Sim concerns. I'm using a Gamist mechanic which is limited (or modified or perhaps tempered) by the other approaches. Many Gamist games have a paragraph or two in the GM's section about fudging the results to make for a more satisfying story. Isn't that Gamism 'limited' by a dose of Narrativism? Maybe the use of the word Narratism in this sense is incorrect.

The goal of play, for me, is an interesting experience within the game world. The characters will act and react, exploring and discover things about themselves and the setting. This includes elements in the setting both pre-planned and implied, but are a product of everyone's decisions. So long as the play and resoloution is interesting, I am not concerned about a tempered bit of drift towards a certain approach from one scene to another. The resoloution is structurally Gamist, but that doesn't preclude the other concerns from affecting play.

I hope this makes more sense to everyone. Please advise.

Thanks
Sparky

[damion]

Mike:Agree.

However it can be hard to alwasy come up with a method that is sadisfying to players as well as charachters. If a charachter dies fulfilling their goals,they probably don't mind all that much, but they player may still hate it because they have to make a new characther.

Well, I think your termenology is wrong. But hey,

The goal of play, for me, is an interesting experience within the game world. The characters will act and react, exploring and discover things about themselves and the setting.

Just because their are dice, doesn't mean the mechanics are gamist. Anyway, I made the same mistakes when first read GNS, still do quite often.
This appeares simulationist, (Actually, Sim Bond would be interesting).

Sparky, any more details about this? Mechanics, ect? It sounds kinda interesting.

[Mike Holmes]

Mike:Agree.

However it can be hard to alwasy come up with a method that is sadisfying to players as well as charachters. If a charachter dies fulfilling their goals, they probably don't mind all that much, but they player may still hate it because they have to make a new characther.

I want to think I'm getting through, but there still seems to be some sort of communication barrier. Maybe not, but I'm going to try one more time.

James, in Narrativist play, nobody gives a damn about how the character feels. That is, if he fails it's just as good as if he succeeds, as long as it fits the developing story. No Narrativist player or GM would ever have a character die unless the player wanted it to happen, or was at the very least not opposed.

Only in Narrativist games designed poorly in this regard (many of them at this point), or in games designed to support other modes, can a character die when it makes no sense to the storyline. InSpectres is a great example of a well designed game. Death has absolutely no place in an InSpectres story, so it's just not possible mechanically. In Hero Wars, it can happen if the GM is not attentive, but is still grossly unlikely. Hear that, everyone? A Fantasy Game where it's impossible to die "accidentally".

As such, characters can only die when someone playing in Hero Wars thinks it's appropriate. Thus, even if the character wanted to die, but the player does not in HW, he will not (assuming that the player is playing Narrativist; a Sim player can make it happen, and would be satisfied that it had). Were you aware that this was the case when playing HW with Ron, James? Does this reduce the tension while playing?

In Narrativist play, the player does what he wants at all times, and the characters motives are only important insofar as the player will want to make the action sensible (why would a player want to do otherwise).

So, what does this mean to Sparky? Well, players who prefer Narrativism will be disapointed by mechanics that allow accidental death. Players who prefer Gamism, will be disapointed by mechanics that don't allow this.

I fail to see how you can have both.

This is the sort of insoluble situation that has the prevailing attitude here at The Forge going with the general assumption that trying to mix your modes of play in design is not a good idea. At the very least it's terribly hard, and I personally have yet to see a set of rules that comes even close to fixing such problems.

So, if you think you are the one to do it, more power to you Sparky. But you'll be in uncharted waters that have swallowed a lot of good designers before you. Consider that it may be impossible (or just way more hassle than it's worth).

Check out these threads:
Can a game designer work for all three (G/N/S)?
is it the slice of the pie, or the size of the piece

Actually, the way that you stated it, it almost sounds like you are suggesting Transition. Do a search for that term if you are interested in trying to create a game that alternates it's support. However, consider that my assumption is that that this is just creating three similar games, and allowing people to pick the one they like.

Mike

P.S. It just occurs to me that MJ's game Multiverser does not allow death. It alows a failure of sorts, but that failure is just to stay in the current universe. As such it's not even close to a counter-example in the traditional sense. That is, it's not a question of Death, or whatever, but of whether or not such punishes the player or not. As such, I'm going to guess that in Multiverser, that you can die "accidentally" and as such that it's either a Sim or Gamist part of the game.

[M. J. Young]

P.S. It just occurs to me that MJ's game Multiverser does not allow death. It alows a failure of sorts, but that failure is just to stay in the current universe. As such it's not even close to a counter-example in the traditional sense. That is, it's not a question of Death, or whatever, but of whether or not such punishes the player or not. As such, I'm going to guess that in Multiverser, that you can die "accidentally" and as such that it's either a Sim or Gamist part of the game.

I would say that it's a bit more complicated than that, but in the main that would be  correct. The mechanic can seem simulationist or gamist, depending on how it's run; but the game has complexities beyond that. The rules provide interfacing with other games; we're working right now on an interface with a highly narrativist game in development elsewhere. Under Multiverser rules, if you verse into a "game world", you're subject to the rules of that game to the degree that that game covers your actions and equipment. (Multiverser rules still fill in gaps, where the player has an ability or piece of equipment which logically should fit in that world but doesn't adapt easily to the other game system.) Thus when we verse someone into a narrativist game system in which death mechanics are strictly limited in that way, the character can only be killed under the terms of that game system. Whenever he is killed, he is converted back to Multiverser rules for the next world.

This interfacing aspect is often overlooked. We're working on a book of interfacing rules for various games, but have to make arrangements with each author/publisher to be included. (In exchange, we let them include Multiverser interface rules in their publications, however they wish to distribute them.)

--M. J. Young

[damion]

Mike:

Sorry. What I was trying to say is that  even in narrativism charachters may get into situations where the charachters dieing is the best for the story, and in fact them NOT dieing may be bad for the story, but this may not fit the goals of the player. In Gamism char death happens from 'failure'.  Gamists don't really like 'random' death, they want death to be the punishment for failure.
In Simulationism char death happens because that is what the system dictates(I.e. death fits with what is being 'Simulated').  
In Narrativism it is either player choice or a conflict between the player and the story(still player choice, but the player decided that they would rather have the story than a live characther).

In most systems charachter death comes down to GM control. (Even in HW). Basicly there is some point where the GM has to take explicit action to kill a charachter. Quite alot of GM'ing advice is how to turn a system where the system controls death to one where the GM controls it. (By say avoiding situations where death is a uncontrollable mechanical outcome).
It seems to me this is what Sparky is suggesting. I think it's an accepatable solution, as it allows the GM to tailor things to the group.

[Mike Holmes]

In most systems charachter death comes down to GM control. (Even in HW). Basicly there is some point where the GM has to take explicit action to kill a charachter. Quite alot of GM'ing advice is how to turn a system where the system controls death to one where the GM controls it. (By say avoiding situations where death is a uncontrollable mechanical outcome).
It seems to me this is what Sparky is suggesting. I think it's an accepatable solution, as it allows the GM to tailor things to the group.

Hmmm. That's a correct point. Yes, a GM can do that. The question is how powerfully the system enables the GM to do this. If it's obvious machination from the GM not supported by the system, that will alienate players who expect a Sim-looking system to act like one.

IOW, I suggest at least a HW level of GM and player empowerment over the outcome of the situation. And that means, basically, conflict resolution, FitM, and the "Failure means conflict" principle.

Mike

[Sparky]

Hi all,

Mike: I'll read those threads and do the search for transition before I use any fancy terms like Gamist, Narrativist. Give me a day or two more and I should be able to eliminate that barrier and respond more accurately to your posts.

Damion:
I really should create a pdf explaing the whole thing, but here's a quickstart-type thing. The dice rolls are meant to provide a structure for play. Most of my players would be confused by the freedom of (say) Donjon. I'm just trying to make it easy for all of us.

=========
Non-dramatic tasks can usually be resolved by just stating the desired action. Certain tasks (like equipment repairs or singing) where the quality of result is important can be resolved by rolling two d10s and comparing them both to a difficulty number set by the GM. If desired, Trait points can be temporarily spent to succeed despite poor rolls.

To resolve a Scene, the player states their general intent and rolls two d10 against a Success target number and an Advantage target number. One of the dice is then assigned as the Success die and the other as the Advantage die. As above, points can be temporarily spent to improve one of the dice rolls, but only one of them.

These rolls are written as Success-dot-Advantage. When spoken, the numbers are simply stated as "number number." Ex: (9.0) is spoken as "nine zero."

The goal is to beat the target numbers. The die rolls minus their target numbers become the Result. Successive Results are added to the Results from last round or 0.0 at start. At this point the action is narrated according to the final Result. When the Result matches the target numbers, the goal of the scene is successful.

What happens is that each of the result numbers rise and fall with the ebb and flow of battle. The numbers drive the action and the resourceful use of Trait points can save the day.
==========

Scene Example:
Dirk the Swordsman is ambushed by the first of two monsters (difficulty of 5.5 when apart and 10.10 when together.) Pretty tough odds. The goal of the scene is obviously to survive.

First Round:
Since Dirk has been ambushed, he is on the defensive. He rolls a 5 and a 4, which he places as 4 Success and 3 Scene. He decides to defend the attack by spending one point from his Swordsman Trait to turn the 4 into a 5.

The Result is now a 5.3 and we take away the diff of 5.5, leaving the final Result of (0.-2) The player narrates that Dirk held off the monsters' attacks by some fancy footwork and some parrying with a quickly drawn sword.(This represents the 0 Success Result.) Unfortunately, he also leapt over a small chasm onto a narrow rock outcropping, which represents the -2 (dis)advantage with limited room to maneuver. It also has the effect of limiting his attackers to fighting only one at a time-much better odds than two on one.

Second Round:
The GM describes one of the beasts clumbsily leaping over the chasm, landing with a snarl and a spray of stones and pebbles. Dirk goes on the attack against a diff of 5.5  and rolls a 7.5. The new result is (7-5).(5-5) or 2.0, which make the final Result (+2,-2) for this round.

The player states that Dirk's sword strikes quickly and often, tearing into the monster's midsection but he does not drive the creature off. The fight would now continue into round three.

There are a few things I left out for sake of simplicity in the description, but that's the essence of how it works. It's all bits and pieces of many other games, of course, but it's meant primarily to act as a tool for creating interesting gameplay. What do you think?

Thanks,
Sparky

[Sparky]

Hi again,

I wanted to detail just a couple more things, but seperately from the last post.

Mike, I'm concerned that that you might feel blown off because of my lack of immediate attention to your posts. Don't feel that way, I just want to really understand where you are coming from.

However, one of your posts made a comment about trying to succeed at linking two approaches where so many others have failed. I'm only trying to design a fun game system for my group...I had/have no pretensions of this effort topping anyone else's. Matter of fact, it is everyone else's good ideas that have allowed me to get something together I really like.

Anyway, back to the details:

Scene Conflicts
These are meant to provide tension by slowing down resoloution. They don't just get used for combat. They can be used for chase scenes or  escaping a trap or whatever.

Character Death:
A PC can only die during a Scene Conflict. Their highest trait determines the Result required. If a PC's highest trait is 9, then the final Result would would need to be (-9.-9) or worse.

A character is made with 20 Trait points and can gain more by making strong weaknesses. Reward points (Cores) can be spent to help them succeed in rolls or to escape the danger.

In the end, you'd almost have to want to have your PC die. If you entered a dangerous combat while low on points and didn't run away the PC might get overwhelmed and die. But that's really a choice. You can see it coming a long way away.

Spending
Points spent from a trait (like Dirk's Swordsman trait) actually double when used appropriately. So really, Dirk should have been (+1.+2) by the end of round two. It was too confusing to include that option in the example.

Limits on the GM
The GM is limited to a budget of twice the highest trait among PCs. (I'll call them tokens for right now.) Difficulty is purchased with Tokens at the beginning of the scene resoloution, with 1 Token equalling 1 Success and 1 Ad target number.

Tokens can also be used (before the roll) to increase the Success or the Ad target numbers. Tokens can only be recovered in the same amounts that the players recover their tokens, so they must be spent wisely.

(Scene difficulties include the factors that affect all participants all the time while Advantage diffs affect only one party at a time. Points spent after the Scene starts only apply for that round.

In the second round of the example, I could have made the Success target higher by saying that the second monster began throwing rocks at Dirk, for example. I could have just as easily said that Dirk was now looking west into the setting sun for a harder Ad target.

So the GM has currency to balance out the threats that can be brought to bear on the PCs while the PCs currency outnumbers the GMs.


Thanks,
Sparky

[Sparky]

Hi all,

As you probably guessed, I'll be absorbing the GNS theory and its applications for quite awhile. I'll post any ideas and questions I have about GNS in the Theory forum.

For this game, I wanted a more Gamist approach because that's our current preference. I realize that this is not to the prevailing taste here and so I really appreciate the feedback and encouragement that I received.

Thanks!
Sparky

[Mike Holmes]

I don't feel blown off at all. Take whatever time you need, Sparky.

For this game, I wanted a more Gamist approach because that's our current preference. I realize that this is not to the prevailing taste here and so I really appreciate the feedback and encouragement that I received.

It's not true that we don't prefer the Gamist approach, really. I think that what most people really object to more, is the fact that most gamist designs are broken. Mostly because they are mixed with other modes of play. Games often effectively tell players, "Play to win; but only if it's "realistic" or "dramatic" to do so." Then the player is left to choose one or the other, and play gets all messed up.

Take a look at the Iron Game Chef competition for some great examples of coherent Gamist designs.

That all said, what you seem to be doing sounds pretty Narrativist to me. The idea that a player can see and avoid death situations, thus putting the power to control when he might die into his hands is pretty Narrativist. There are, strangely enough, some strong similarities between Gamism and Narrativism (the willingness to accept a lot of non-continuity driving metagame). You might be surprised.


James, I re-read what you posted. And I just have to comment again. If you are reading a book, and the main character dies, and you think, "damn, that sucks, I wanted to see more stuff happen to him", is that a good story? A good story is one in which the sort of things that the audience wants to see happen, happens. So, at no time is the character dying in an RPG when the players don't want it to happen, good for the story. As the players are the audience. What makes a good story is whatever sorts of things that they'd like to see happen. Resolution systems, then, are only springboards for creativity. To the extent that they cause stuff to occur that the players don't like, they are anti-Narrativist.

This is key to understanding Narrativism. There is no story for story's sake. There is only story for the player's sake.

Mike

[M. J. Young]

Mike, I'm going to take issue with this idea:

A good story is one in which the sort of things that the audience wants to see happen, happens. So, at no time is the character dying in an RPG when the players don't want it to happen, good for the story. As the players are the audience. What makes a good story is whatever sorts of things that they'd like to see happen. Resolution systems, then, are only springboards for creativity. To the extent that they cause stuff to occur that the players don't like, they are anti-Narrativist.

My thinking is that there is a balance here. A good story is one in which the reader is surprised by what happens, but not unhappy about it. Hey, I wanted the Fellowship of the Ring to stay together all the way to Mordor; Tolkien had a better idea. I wanted Gollum to reform and Frodo to bravely throw the ring into the crack in Mount Doom; again, Tolkien's idea was better.

The point is that a story needs to do things we neither expect nor want, in order to be better than the story we would have created ourselves.

That doesn't mean that whether a character dies or lives is of no consequence; it does mean that it might be a better story when the character we wanted to have live dies.

Looked at another way, I write stories. Often when I'm writing, I will write myself into a corner. It will become apparent that what I thought was going to happen, or what I wanted to have happen, can't happen because of what I've already written. At that point, I could decide that I've messed up the whole thing, toss out fifty pages, and try to rewrite the whole thing. I'm not like that. What I do is try to figure out where the story will go, how I can make it better than what I had intended. Usually I succeed. Often I find wonderful plot elements from trying to resolve story events which I would not have found otherwise.

As an example, I recently put a book character into the situation of being born over again, with all his past knowledge intact. This character had telepathic abilities, which was not normative. Before he was more than a few months old, he decided to telepathically contact his new mother to try to figure out what was happening to him. Once he managed to convince her that she was not crazy, she did the logical thing that he--and I--had not seen coming: she told his father. Now he had to choose between letting dad believe mom was crazy, or showing dad his abilities. Dad did not take it terribly well; this being a rather primitive setting, there were immediately questions about whether this was some sort of demon child. Mom defended her son. Perhaps, she said, their child was some sort of god-sent deliverer, to help the people in their time of need. Oops--I'd had no idea about a deliverer, or a time of need. I actually had not yet figured out quite what the character was going to do when he grew up, as I was primarily exploring bringing the character through growing up. But this gave me a major story, about racial oppression, in a fantasy setting.

It's certainly not necessary; but it isn't entirely detrimental to narrativist play to have entirely unanticipated and seemingly negative events including possibly character deaths. Hey, I thought the movie Executive Decision was terrible despite a good cast; and I really did not like it at all when Segal's character was blown away in the boarding effort. But that's one of the best scenes in the (otherwise weak) film, and it sets up the real conflict of the movie, the aspect of the commando team taking orders from the inexperienced guy. It might have been a good film (it's probably a good book) given that story. (What I don't like is the constant repeat of "go--no, wait" which starts to become ridiculous.)

--M. J. Young

[Mike Holmes]

We really don't disagree in principle, MJ. We disagree in how to get the effect sought by the principle.

I've said that randomness is OK. That it is a springboard from which creativity comes. Yes, it should provide unforseen twists and turns. I've never said, or meant to imply that the player should simply get what he wants. Just that he should find the output satisfacttory in al lcases. We agree that, in the end, the result should not leave the players unhappy. That's all I've said.

Further, as in your writing experience, I think that there is no way that in a game with multiple players that one could ever have an outcome that was entirely predicted by one of the authors.  Certainly not in Narrativism as defined. So that's completely a straw man. It's going to be random in any case.

My point has been that unexpected death is very rarely, if ever, a pleasing result to the NArrativist player. Whoops, rolled really terribly, and Bob the Magnificent died by the hand of Mook #4 leaving a huge number of really excellent story threads hanging.

How is that at all something the player wants? This is not a dramatic setback, it's the end of the character's story. As I've said, it's not success that we're concerned with maintaining, but simply the players' power to continue to play until such time as they think that a termination of the story is appropriate.

Again, we're assuming Narrativist here. You're comments certainly make sense if you are talking about Simulationism. Where the player is more concerned with internal consistency or something than he is with story. It's your refusal to understand that this is precisely where the line between Narrativism and Simulationism is drawn that leads you to believe that there is no line, and that a game can support both equally well, it seems to me.

Now, I'm not saying that you can't have a coherent hybrid that plays of a bit of both. That's an option. But, as with all hybrids, the probable result is drift to a manageable mode. IME, the usual response to such a hybrid is Illusionism. Wherin the GM takes personal responsibility for ensuring that the story is maintained by, amongst other things, making sure that the characters somehow survive as long as neccessary. This was my mode of play for the longest time. And it works just fine for Sim players. But the players who prefer Narrativism hate it. As such, if that's what you want to support, then one needs to put the ability to survive (read "continue the plot)" square and openly in the hands of the participants.

Mike