I was thinking over the weekend about various ways that games preserve antagonism (in the form of significant antagonist NPCs) from untimely ruin. Of course not all games are concerned with the kind of story arc that husbands its antagonists for a thematically satisfying conclusion. I've played trad games that let the chips fall where they may; if someone manages to soak the evil wizard in burning oil, well, he's dead, regardless of what the GM had planned for him. And I've played nar games that simply whip up a thematically satisfying climax and conclusion from whatever's still standing at that point in the game. But when a game concerns itself with the preservation of antagonism it seems to me it does it in one of four ways:
1. It makes the antagonists really tough, in hopes they endure until the end. (And maybe there's some tacit support for the GM fudging results so they do.)
2. The mechanics have a threshold that must be crossed before the game can have a climax involving the main antagonist. My Life with Master and Extreme Vengeance do this. And I think With Great Power as well. Though I could maybe be convinced that With Great Power is actually type #1: the antagonist just starts with an overwhelming advantage.
3. The game mechanics enlist the players in appreciation of the value of the antagonist as a force that protagonizes their characters, and so there's player investment in the preservation and usage of the antagonist. Often, but not always, this is done by giving players some control over NPCs, including the antagonist.
4. The game has some sort of genre convention, or in fiction constraint. Most superhero games fall into this category.
Am I missing some other major way that games do this? Also, I'm particularly intersted in type #2 and type #4 games. What are our other type #2 and type #4 games?
Paul
Interesting topic.
Other #2 games include:
-Geiger Counter: The Menace cannot lose dice until has reached its maximum threshold
-Dust Devils: You must enter the End before you can be removed
I think it could possibly be worthwhile to discuss whether the preservation of antagonism is in fact necessary, but that's not really suitable to this thread.
Hi Paul!
The way The Shadows of Yesterday do it (to kill a major antagonist, you have to enter BDTP and risk the death of your character) could be seen as a different category, or as a particular case of #2...
In the 4th category, as "fiction constraint", you would consider even setting constraint? (as for example, in Ars Magica, the Order of Hermes law that prohibit enemy covenants to destroy each other). Often they were rather crude and force-based ("don't destroy the GM's toys or he will have the entire setting come after you, HARD!) and difficult to use with players who didn't care to learn anything about the setting.
In Jonas Ferry's "One Can have Her" the GM can veto any conflict that has as stakes the death of a character's named enemy, but it could be said that in that game the true opponents are the other player's characters (and any player can veto the death of his own character as stakes in a conflict), and no Player character or Enemy can be killed without using a conflict with that as stake. It's the most explicit (as "in-the-rules") matter-of-fact protection of any guy someone designed as his "enemy" that I remember at the moment without having the enemy protected by some endgame conditions as in MLWM.
Hi Paul
Care to share some experiences that could inform us on why you appreciate a type more or less? Also, how do you analyse antagonism in GM-less games and in Polaris in particular?
With Great Power... definitely falls into #2. The villain's Aspects explicitly CANNOT be devastated (that is, affected in any lasting way) until the final space on the Story Arc is filled.
The old TSR Marvel Super Heroes game had an easily-missed rule that any supervillain could spend 50 karma points to escape certain death. I'd put this in your category #4, primarily.
I'd say that Extreme Vengeance bridged #2 and #3, because players got XP for rolling more dice against the bad guys AND for having the bad guys roll more dice against them. Even though players had no control over NPC villians, they got more points the longer the bad guys could put up a fight. A great idea that was, unfortunately, extraordinarily tedious in practice. Someone had to write down every die roll in the game!
Spirit of the Century uses something like #3--antagonists can offer concessions through which, although defeated in a current plan, they get away to scheme again. The players have to agree to the concession though, otherwise the conflict can continue to the complete defeat of the villain.
Shock uses method #2--the protagonist cannot usually resolve their story goal until the antagonist credits are reduced below a certain point.
Mel
Hey, Cristoph:
I imagine that they're intended to fit into #3 or #4, depending.
To hit on Polaris, it uses something which, if I squint a little, is kinda-sorta like what Paul categorizes as #3. Mainly, if I cut off the first clause halfway through "The game mechanics enlist the players in appreciation of the value of the antagonist." Full stop. None of this lit-theory stuff. Polaris doesn't particularly tell you anything about how the antagonist protagonizes the antagonist blah blah blah but because the kill-rate is so high for non-Heart characters in Polaris, characters (which, naturally, include antagonists) only survive if they are of interest to the players. So antagonists survive based on player understand of their own "coolness."
The best example of this I can come up with was a game I played with a giant demon named Altair, who was so tall that his head reached the stars. Now, just like any other character in Polaris, he was supremely mortal. Any player could have said "I cut off Altair's head and he dies" and that would have most likely been that. However, we were all so invested in his size and invulnerability that killing him required traveling in to the sky, making pacts with the moon, and so on.
Note that we're not doing this because we understand in any conscious way that Altair protagonizes our characters. We're doing it because Altair is damned cool, and we're all really into how cool he is, and we don't want him to casually die*.
yrs--
--Ben
* This is all about the preservation of antagonist in Polaris at an individual level. The preservation of antagonism without respect to the individual antagonists, is more of a matter of Paul's #4: the fiction of the game contains a source of infinite and unending antagonism. Bliss Stage works the same way.
P.S. Hey, Paul, this may be kinda an edge case, but what about games that simply don't contain antagonists as such and get protagonism from other means?
I was reading through Heavy Gear recently and that game seems like DP9 made a concious effort to cram every bog standard 90s design paradigm it could into a single game.
It deals with antagonists largely with #1 but there's no tacitly about it. They use this "Chessmen" system to identify important NPCs from less important NPCs so that GM's can avoid messing up the meta plot by killing off key characters (DP9 was apparently REALLY big into the idea of doleing out the meta plot one splat book dollup at a time)...they spend a ton of time talking about these Chessmen...clearly a device they were very proud of.
Then there's a section on how to roll dice where the advice (VERY explicitly) is to allow the players to roll their own dice but never tell them the difficulty ("Threshold" in HG terms) that way the GM can set it to whatever they want after they see the rolls to make sure the players don't ruin the story...this is not extrapolation...its direct instruction in the text (too bad I didn't have this text handy for any of the Impossible Thing Before Breakfast threads). And that includes saving NPCs who are "restricted".
This isn't concidered "fudging" its considered proper play. In fact, the advice for players who are "advanced" enough (direct from the text) is to forgo dice altogether and just let the GM adjudicate everything that happens. That way you make absolutely certain that those nasty players don't ruin the GM's ability to regurgitate DP9s meta plot.
Yowza!
But in terms of protecting Antagonists...it doesn't get any more solid than that.
Hey Christoph,
I can't say I appreciate one type more or less as an enduring and fundamental personal preference, but I can tell you why I'm currently interested in types #2 and #4.
In conversation with Thor Hansen on Saturday I made the observation that I'm not a very good roleplayer. And I supported the observation with references to great roleplaying I'd seen at Gen Con, by Eero Tuovinen in particular, and John Harper as well, and from Thor himself. He challenged the observation with a reference to my excellent roleplaying of a specific NPC when running Lacuna a few months ago.
And y'know, I'd enjoyed the hell out of that particular NPC. And it got me thinking about how so many of my favorite gaming experiences, where I had great fun and feel like I was roleplaying well, were with NPC antagonists. So now we're going to be playing Grey Ranks next, but after that I'm wanting to run something that'll support me in running NPC antagonists. I don't want a type #3 game, because I want unshared ongoing ownership of the NPCs. And I don't want type #1 because I've seen it fail to work too often.
So I'm interested in types #2 and #4. Know of any non-superhero games that fall into type #4?
Paul
Paul, I think #4 is not usually something that's part of the game itself, in terms of general mechanics or setting. Rather, it's part of scenario design. For some Forge games that's largely congruent as scenario design is part of the game, but for others and certainly for traditional games it's not.
Say you have a contemporary setting and the PCs are cops or some other sort of law enforcement guys. And your antagonist doesn't give himself away easily. So, they could just put a bullet in his brain but they wouldn't, because that would be illegal. There's your #4 at work.
Or say your PCs are Hogwarts students and your antagonists are a teacher and an annoying Slytherin student. The social context within the fiction and also the genre conventions prevent the PCs from just hurling an Avada Kedavra at the antagonists from behind. That's your #4, too.
In both cases, it doesn't matter whether you play The Pool or HERO. So, I can't point you to a specific game that does this, but my point is: You don't (necessarily) need one. Thoughtful scenario creation will do that trick, too.
- Frank
How about games that have an explicit meta-game resource for the antagonists? This could either be general "Twists of Fate" the GM can employ in any situation he wants, or specific "Villian Points" that in-color antagonists have. This way the antagonists can be defeated by sufficiently skillful actions of the players, but won't die from a single exceptional roll. Combined with an interesting economy like in Pace, this could allow a great protagonist-antagonist dynamic.
Depending on what you're seeking from the game, it's possible that the constraints introduced by scenarios #2 and #4 would be unsatisfying -- it might be the case that being undefeatable (at least permanently) would make the parts before the endgame less exciting.
James
Hey, Paul. Do I read you rightly that you're not interested in the preservation of antagonism but in the preservation of antagonists?
If the character ceased being an antagonist, but you got to continue to portray them, would that be satisfying on not satisfying? Why?
yrs--
--Ben
Hey Ben,
Quote from: Ben Lehman on September 16, 2008, 12:13:10 PM
P.S. Hey, Paul, this may be kinda an edge case, but what about games that simply don't contain antagonists as such and get protagonism from other means?
Name one?
Quote from: Ben Lehman on September 17, 2008, 02:09:24 PM
Hey, Paul. Do I read you rightly that you're not interested in the preservation of antagonism but in the preservation of antagonists?
If the character ceased being an antagonist, but you got to continue to portray them, would that be satisfying on not satisfying? Why?
I dunno. I'm not interested in an antagonist to protagonist reversal that eclipses the protagonism of the player characters. I'd hate that from the other side of the table. I'm also not interested in owning a bunch of NPCs who're largely window dressing to protagonists and antagonists being run by the other players. If some NPCs fall by the wayside from my own failings, I can handle that. It's on me. Failures like that are part of the creative fun. I'll have other NPCs that might fare better. What I don't want is my best efforts getting authored out from under me or killed early on the upslope of my enjoyment of them.
Paul
Hey Frank,
Quote from: Frank Tarcikowski on September 17, 2008, 06:22:37 AM
Paul, I think #4 is not usually something that's part of the game itself, in terms of general mechanics or setting. Rather, it's part of scenario design. For some Forge games that's largely congruent as scenario design is part of the game, but for others and certainly for traditional games it's not....it doesn't matter whether you play The Pool or HERO. So, I can't point you to a specific game that does this, but my point is: You don't (necessarily) need one. Thoughtful scenario creation will do that trick, too.
The challenge for scenarios with things like codes of bushido, or the requirement that cops be law abiding, is that players recognize the protagonizing power of transgressing and pursue it with their characters, because most genres that have such behavioral codes are actually about the transgressions. Perhaps I'm failing to see the forest, but I can't put my finger on a single other genre besides classic superheroes that actually breaks if the the code is violated. And that's what you need for #4. Something fundamental to the genre, but in a way that doesn't invest a transgressor with protagonism. What am I missing?
Paul
Well, more likely it's me who is missing something because I just don't get superheroes at all. It's not my cup of tea, as our British friends would put it. You are of course right about transgression, but speaking of a dramatic structure, wouldn't one usually build the action towards said transgression so that it will only happen at the climax of the story? I mean, you don't want to preserve the antagonist forever, do you?
- Frank
Quote from: Paul Czege on September 17, 2008, 04:12:15 PM
Hey Ben,
Quote from: Ben Lehman on September 16, 2008, 12:13:10 PM
P.S. Hey, Paul, this may be kinda an edge case, but what about games that simply don't contain antagonists as such and get protagonism from other means?
Name one?
For three, the whole Trollbabe -> Dogs in the Vineyard -> Drifter's Escape lineage can do this, if the GM wants to (or, in the case of the Drifter's Escape, if everyone wants their characters to be sympathetic.) Basically, it creates a situation where a lot of people have done bad things to each other, and the protagonist is in a lynchpin position to figure out a resolution to their grievances, but no one is providing opposition to the protagonist in any way. The protagonism comes from the difficulty in resolving the situation, period, rather than from any effective resistance.
I'm thinking in particular about a Dogs in the Vineyard game Tony ran involving some pre-marital sex and some teenagers falling in and out of love and making stupid mistakes. All of them were basically good people, but the situation was completely unresolvable in any sort of satisfactory way: we ended up ordering a boy to fall back in love with the girl he had previously been in love with and so on. No one in the game opposed us in any significant way: they were aware that they had screwed up, and that their situation was difficult, and they were relying on us to help them sort it out.
Not an antagonist in sight.
I'm not actually sure that antagonism (as in: resistance to the protagonists that allows them to shine) has anything to do with antagonists (individuals hostile to the protagonists) at all. Clearly, a functional antagonist is one possible source of antagonism, but it's not like the antagonism itself goes away when a specific character does.
Not to say that keeping your favorite NPC alive isn't also a noble endeavor. It's just unrelated to protagonism or antagonism or any other -ism.
yrs--
--Ben
Hi Paul,
Here's an interesting technique for #1, found in the game Zero. In that game, a given character's roll is 2d6. What you do is, when rolling for a bunch of foes and when one is supposed to be pretty tough, you roll all the dice for your NPCs in one big go. Then you sort them into 2d6 pairs of your choosing, assigning them to characters as you see fit. That way, you make sure that your favorite baddies get the good rolls.
In Hong Kong Action Hero, NPCs are rated on a strict scale, which is to say certain villains are practically unbeatable and the mooks are really one-hit mooks. I'll have to review the specific technique, but it was very formal. This kind of NPC categorization was really big in early-mid 1990s design, like the Chessmen thing Ralph is talking about and lots of others.
The thread topic does confuse me a little, because I'm not sure I understand the difference between antagonism (-ist?) and adversity in general. Nor am I sure whether it's linked with certain expectations for scenario prep and play, i.e., a climactic confrontation. It'd help if I could work with a specific example.
Best, Ron
Quote from: Ron Edwards on September 18, 2008, 01:45:26 PM
...I'm not sure I understand the difference between antagonism (-ist?) and adversity in general...
Sorry if this quote seems a little out of context, but this highlights some of the ideas that have been going through my head with regard to this thread.
Especially after Ben's comments about a DitV game without personified antagonism.
I'd always considered antagonism to be an obstacle that prevents a "happily ever after" conclusion. This could be in the form of a person who is instigating storyline events that need to be resolved, or the dramatic tension of events that need to be addressed. Without antagonism of some form, is there a story?
Sometimes minor events get in the way, and need to be addressed before the true villain is revealed. Sometimes minor villains need to be addressed before a mencaing event can be seen for its truth.
I perceive most of these systems as different ways to categorise the immediacy of threats. Methods to isolate protagonist and antagonist (personified or otherwise), until suitably dramatic moments. That probably a narrativistic way to look at things, but it's the first that comes to mind.
Isn't a relationship map a similar concept when viewed in this light...heaven knows we've seen a lot of them lately. The characters have an ally in the form of Person X...Person X connects with Y and Z in these ways...the players can't access Person Z until they've resolved the fight between X and Z...oterwise they could go through Y to get to Z but this has it's own issues to be resolved.
The antagonism between person X and Z isn't directly motivated against the players, but they'll need to bring it on themselves to get further.
I guess that's the whole point.
Stories need antagonism of some form as a drive. Complex stories develop this further by making certain antagonisms hidden, while their ripple effects may be felt further afield. For a character to develop, they need something to work against...
...that's why they put snakes on the snakes-and-ladders board.
V
Hello,
Ouch. Michael, that's a very distorted description of relationship maps; it describes older techniques found in hundreds of modules, not how relationship maps are used (you may not know that the term is highly specific to a single game text, not a general/casual term). Rather than derail the thread, I invite you to hop down to the Adept Press forum and talk about it there.
The thing is, though, I agree with you fully about the presence of adversity, or even better, relevant adversity in making any kind of fictional experience worth anyone's time. But I don't know whether that point applies. I don't even know if Paul is talking about adversity in general, antagonism (which is to say adversity in the form of other people's priorities), or specific antagonists (characters chosen to deliver antagonism).
Paul, this thread is really struggling. Can you provide an account of play that nails what you're talking about?
Best, Ron
Hey Ron,
Quote from: Ron Edwards on September 18, 2008, 01:45:26 PMThe thread topic does confuse me a little, because I'm not sure I understand the difference between antagonism (-ist?) and adversity in general. Nor am I sure whether it's linked with certain expectations for scenario prep and play, i.e., a climactic confrontation. It'd help if I could work with a specific example.
Personally, I reject Ben's assertion that protagonism can be achieved without antagonism. NPCs with needs they can't resolve themselves who attach themselves to player characters are antagonists. NPCs with flaws that create problems for player characters are antagonists. They are the source of the conflicts upon which the story turns.
The first time I ran The Pool I had an NPC named India Vaunt, an accomplished mercenary leader who was a mentor to Scott Knipe's lower ranked mercenary character. In the second session of play he used a Monologue of Victory to kill her melodramatically during a trial by combat in which they were both acting as champions. As the GM, I wasn't done with her yet. Scott scarfed her for a moment of melodrama.
I'm finding myself needing to run a game that lets me own, and invest time and creativity and depth in NPCs. I'm weary of having NPCs slain by vagaries of the dice (a risk of type #1) or authored out from under me (my problem with type #3) before I've even scratched their surface.
(Moreno, would you believe I had just made a PayPal payment for a copy of Ars Magica 2e before seeing your post on the 15th? It arrived two days ago. I've barely had a chance to crack the cover, but I'm really looking forward to seeing the pre-licensed game.)
Paul
Hi Paul,
H'm - without investing into the details of the Ben/you post, I think relevant adversity is the source of protagonism, and antagonism is the most reliable route to adversity. And antagonism is most obviously delivered through avatars or representatives, i.e. antagonists. But I do think you're talking about the latter.
So OK, that means we just focus on the latter and not get into the perhaps-metaphysical higher categories.
One thing about The Pool, PTA, Trollbabe, and a number of other games is that antagonistic NPCs have no special value. There are some practical nuances. As Pool GM, I can withhold all Gift dice when a particular NPC is involved, forcing the player to rely only on one Trait and whatever he or she chooses to gamble. I've found that Trollbabe play produces antagonists mainly by seeing which conflicts end up hurting or incapacitating the protagonist, and thus "casting" that particular NPC into the role of "tough." And in PTA, a Nemesis character has to be available for play/use throughout a season, by definition.
But all that said, yes, I see your point. The GM rides with what's happened, period. Dust Devils can go the same way; it did for me, when a fairly tough NPC did in fact get a really nasty draw against him, and went down for good in an early fight scene.
Here's the thing: this is a feature in given systems which may not meet your needs (in this regard) as systems. I know why it's such a common feature - people have been writing games in which protagonism is reliably consequential, and in which GMing is more about working with what's been shaped in play. Too many of us have spent too much time chipping away hit points, or STUN, or watching our characters' actions get re-shaped into whatever the GM wants to happen no matter what we say.
I'm not sure protagonism must suffer in these circumstances. I modestly suggest that Trollbabe play delivers nicely in those terms, and it is very, very close to The Pool in this design feature. (Of course, taking down a foe often risks grave harm in Trollbabe; maybe that's involved.)
But again, to stick with my agreement with your point rather than the quibbles, I want to investigate what systems do really allow meaty antagonist play. (And I imagine, without simply giving them uber-high never-chip-it-away scores of some kind, right?) What's out there?
Unfortunately, I'm not coming up with much. In some cases, the antagonist is untouchable, even a little intangible, as My Life with Master and The Mountain Witch (although secondary characters can be quite tough in the latter). In others, they're effectively player-characters and therefore subject to sudden reversals on occasion (The Shadow of Yesterday, Sorcerer, Dust Devils, Legends of Alyria, The Riddle of Steel, Nine Worlds).
A good question. I actually struck toward it a little bit with my older notes for Doctor Chaos.
Best, Ron
May I suggest taking a look at Darkpages. Darkpages involves creating a specific Antagonist NPC. For all intents and purposes they're the same as PCs but go down a little faster because they process Pain directly where PCs have a few channels to dump Pain into.
That might sound like exactly the OPPOSITE of what you want except that antagonists work off the same Re-incarnation rules as PCs (at least I think they do). Which means that they can come back... just slightly changed.
Jesse
7th Sea might be worth a mention here. Most of the approach is fairly conventional, with three mechanically distinct tiers of antagonists who are progressively more capable. But more interesting is the Wiles that villains can have, and which correspond to various special powers held by PC's. These Wiles are often specifically aimed at preserving the antagonist, and are framed in terms of the antagonists dramatic role; one of them is simply "Recurring". Other Wiles allow villains to, for example, re-roll dice, ignore Surprise, cause a PC to fail a roll, have stronger minions, etc. This is a lot more flexible than simply having a lot of hitpoints or even being physically tough at all, and are rather more like mechanical interpretations of the kinds of fudges GM's use to keep their villains alive.
An interesting example of either type #1 or type #4 or a kind of mixture, drawn from my own actual play, is provided by Steve Jackson's "In Nomine". In the game's setting, PC's (and frequently opposing NPCs) are either angels or demons, who (on the corporeal plane) walk around in physical forms known as Vessels. If your Vessel is killed (or perhaps the more accurate term is destroyed), the angel/demon does not die, but instead returns to Heaven/Hell, passes through a sort of coma-like state called Trauma, and eventually (with sufficient rolling of dice and passage of in-game time) awakes from Trauma and could purchase a new Vessel and return to the corporeal plane.
In practice for the GM providing the antagonism, this provides an inbuilt reason why a "killed" antagonist can return whenever the GM would like them to do so, although eventually the idea that the angel/demon involved can keep acquiring new Vessels without much apparent personal cost will strain credulity (especially for demons). I myself have done this in my ongoing game, although I have been at pains to portray the demons involved as having suffered for losing their old Vessel, and have yet to have a demon who has been twice Vessel-killed recur a third time.
Should it be wished that a particular angel or demon be more permanently erased, it is possible to "soul-kill" them, although this takes a lot of damage to achieve completely (and retreat to heaven/hell is comparatively straightforward). It's also extremely difficult to engage in the type of combat in which soul-killing is an option without all combatants opening themselves up to it.
Given the inefficacy of Vessel-destruction as a means of permanently disposing of antagonists, and the inherent difficulties of soul-killing, it often seems more practical from a player's perspective to commit "character assassination", or death by political manoeuvre - trying to get a given demon to screw up so badly that their patron Demon Prince would not let them return to Earth, or at least would reassign/depower them.
I haven't seen this said yet, but why can't antagonists preserve themselves?
Why does the gloating villain gloat? Because he is relishing a temporary moment of invincibility, which he has set up. I think sometimes the problem with many game villains is that they are all front; they don't have the backup work to be as invincible or dangerous as they say they are. Now this requires engineering; the villain must have some region of advantage, some comfort zone in which they are hard to take out, based on the rules of the game (explicit crunch-sim or otherwise). This gives them comfort to be a villain in the first place, because they feel untouchable. It's like anonymity on the internet!
Going back to old school, why do we have stakes at all? Or why do something like bringing down the pain? Because people in the game need to have warning that something they are invested in may get lost, or irreparably damaged. People in the game includes the GM!
So just like edge in Shadowrun has a use for player characters, it has a use for NPCs too. Because these mechanics are never really about PCs being favoured by the cosmos, but because their players don't want to loose them. This is different to bringing down the pain because it cannot fail! One of the problems with the "Are you sure?" method is that sometimes people will say yes, they will risk the character you thought they were totally stuck to and take out your apparently more beloved NPC. This is one of the many reasons I think Universalis is brilliant, because it tries to make explicit how invested people are in stuff, but the other opportunity to taking a wise gamble is not to gamble at all.
Also, putting story elements in front of a player is asking him to change at least one of them, so you are always risking something. One alternative to this is internalising a character's conflict, so they carry their antagonism around with them. I'm talking more heavily psychological systems here, so people can have issues, and be forced to deal with them and adapt by situations, perhaps even gaining new ones. This flips the problem, as the player is now having to accept the world encroaching on their pristine character concept. Another alternative, like the mechanically constructed safety zone above, is to have explicit moods for a scene, that people can shift by their actions, which provide the context for acceptable actions. This forces you to think about what situation elements would give the that mood sufficient force, and so design a situation that signals to the players what to expect.
Another tack is that when people solve your originally immensely tricky scenario, resolve your antagonism in a non-sociopathic way, they may have actually taught you something for life! In other words, if your antagonism has been weakened by a flaw in your model in the world, does this apply to situations you have found insolvable? If not, and you can find a parallel, how about ramping it up by including those elements that make it harder for you. Now this hardly solves the emotional investment problem, but it could mean something pretty wonderful for the broader effect of your role-playing.
Another view that is rarely used in games is that sometimes finding the source of the problem is the problem itself. Working out what you are supposed to be opposing. Now this is what is often called a "soft system" problem, the semi-existential one about "What are we supposed to be doing here?". This is generally considered a fundimental deficiency of a role-playing method rather than an in game issue to fight, perhaps because people sometimes play games to escape from such unstructured problems, and dislike it as horrible blankness. That is I think why solving it during games design is a big thing, but that's not the only way to go.
It's essentially a constructive approach, encouraging players to build rather than defeat, with mismatches between built structure and the setting environment appearing every now and again, so that players cause their own conflict via over-corrections, or are simply satisfied with what they have created, and we move on! But I suppose this last form is a bit off topic.
I started reading this thread because the topic heading relates to something I am currently discussing with a friend.
Upon reading it, it seems to be about something only tangentially related to the issue I'm discussing (despite the topic heading), so I'll start a different AP thread about that (look out for [Land of Nodd] in the title). It's about "antagonism" (as a counterpoint to your conception of protagonism) as opposed to the "antagonists" and their survival.
However, I do have two comments. Both have been hinted at already:
1. I don't know if this lines up with your #4, Paul, but I think it does: there are many settings which are clearly written in such a way that total, complete victory for the protagonists is not a possibility. A good example might be the world of "The Matrix" (from a certain point of view, anyway), as well as some versions of D&D ("point of light" surrounded by infinite lands of danger), Cthulhu-type stories, many zombie horror movies or apocalyptic settings, etc.
We know, just from seeing the setting and the premise (little "p"), that we do not expect the heroes to overcome the adversity of the wide world around them. Unlike the Lord of the Rings, the little guys don't have a chance to change the state of the world. It's already built-in: there will always be more monsters, or the Evil Emperor is so far out of reach that we can't confront him meaningfully, or the zombie infestation has spread over the whole globe, so we can only hope to salvage our own survivial, not change the evil that threatens the world. Another example might be Grey Ranks: we're pretty sure that the stories of our protagonists are not going to be about them overthrowing the Nazi threat as a whole. So, we know from the beginning that the larger-scale antagonism will never disappear.
2. It would be interesting to see a system where players in some way demonstrate their interest in particular antagonists, and that interest translates into in-game bonuses for those antagonists--maybe they become more effective, or merely harder to kill. Are there any systems out there that do this?
A short one: I made a #2 - kinda game. Each player choses a mission for his character. The mission we played yesterday was: I have to avenge my brothers death by slaying the evil Lord Viggo. Now there are two resolution systems. In play, characters get into conflicts (system 1) where they may earn mission dice. When a player has earned enough mission dice, he may call for initiating the end of his mission (system 2). Here, he rolls his earned mission dice one after another, with each roll narrating how the character comes closer to his goal or how has to retreat a step. When two dice match, he narrates how the character completes the mission (in our case, he finally struck the evil lord dead). If he doesn´t get a match, well, tough luck, the villian escaped (or whatever). Also, there is no other way of completing the mission but with system 2; normale conflict resolution may not result in the mission´s end, and if the mission is killing the evil lord, he is invincible until the endgame is initiated.