First of all:
I'm thinking Respect and Trust ought to be unified if both are going to be arrays of dispositions toward various elements of a character. Since Respect was intended to be a measure of how much a character trusts another to do their "job", it's already covered by Trust. Seem reasonable?
Authority on the other hand has me scrabbling for synonyms, because I think the word itself implies leadership, rank and power structures, which aren't really part of the concept. Put simply it's just the "power" one character wields over another, and the main reason for its existence is ongoing intimidation. Rather than making one-off checks to intimidate someone into doing something, you're constantly asserting yourself in a position of power above them, and using that domination to achieve your intimidation goals.
To that end, I think I'd keep it as a single disposition toward a character rather than breaking it down any further, and I'm thinking it probably needn't be a scale between submission and domination, since submission is merely the reciprocal action. For instance, if Angela has a dominating disposition toward Nancy, there's no reason to have a variable on Nancy's side to reciprocate what is already known. And in some cases, I'd like to be able to have mutual blackmail and so forth.
Now, I really like the function of Authority, though it needs a new name:
Nancy spots Angela trying to pick the lock on Dave's door. She calls her out on it, and decides rather than publicise the fact, she'll keep it to herself and use it down the track as blackmail. Fairly simple, right? But the beauty of it, rather than integrating blackmail into a quest script, the game is actually tracking the power Nancy can exert over Angela, and that can be applied to any unwilling action from Angela.
Just another side note about Respect: this might not bode well with the limited amount of NPCs at your disposal. I just recently read the story of Boatmurdered. A game of Dwarven Fortress that went horribly wrong as elephants started to attack and kill several dwarves. The other dwarves had a higher priority set for looting corpses instead of continuing their tasks which would ultimately lead to more and more dwarves being killed by the still remaining elephants. While such AI hick ups are tolerable (and quiet entertaining at times) since there is, technically, an endless supply of new dwarves it won't work the same way with a small not expendable cast.
Boatmurdered is a great read, and it's a good example of the things I love about Dwarf Fortress. But you're exactly right, because of the nature of the game - where failure is generally at least as entertaining, if not more entertaining than success and actions like forcing the front doors to lock dwarves
in could loosely be considered part of the strategy.
I share some of your trepidation as to what AI hiccups might mean with such a limited cast, but having no schedule works in our favour. We can playtest the hell out of this thing until it works. However, a couple of things to keep in mind - I plan some redundancy within the society. With thirteen characters and about five major archetypes, at the very least there will be a certain overlap, so the loss (or casting out) of one character will make things more difficult, but not impossible. I've also intentionally tied two major, non-combat archetypes to the Brains stat-pair, so if worst comes to worst, one of your inventors will have decent stat line to switch to a doctor. And as Galsiah has mentioned in one of his later posts, as long as the game stays entertaining in some way, and the player always has things to do, that downward spiral toward complete destruction could be enjoyable.
Heh, sorry but this in combination with horror reminded me of a little gem from the past. I believe that it could turn out to be a little out of place but as long as it makes more sense than in Arcanum I'm fine with the idea.
Funny clip, looks like my sort of movie. But of course, nothing I'd want to strive for with Synaesthesia. The main idea is to establish various "faction lines", when it comes to attraction and so forth. I want a few overlapping webs where characters are linked according to their various dispositions, and so there's a handful of sides to be taken in any kind of dispute. I think it would be a bit dull, and entirely unlike a plausible human society if it wasn't part of the system.
However, I think there's an important distinction here between internal and external emotions. I.e. between the true emotions of a character, and the effects he shows to the world. I don't think it's a good idea to give a PC (/NPC) any direct way to influence his internal emotional state - no matter what his skills. I do think it's a good idea to give characters the ability to control the expressions of those internal emotional states (i.e. "external" emotions).
A character skilled in debate might have great control over his tone and the words he uses, but this shouldn't mean he's able to reduce his anger emotion - only that he's able to hold its effects in check (in some circumstances).
That's exactly right. The player character will be able to address their emotional state to a certain extent, but the vast majority of control comes through controlled expression, both in terms of keeping your emotions in check, and going out of your way to find outlets. An angry character may find that defending against a horde of zombies is actually fairly therapeutic.
I think they might be. After all, why not? - it gives hugely more replay value, so long as things can be kept entertaining for any group such recalculation produces (naturally the recalculation needn't be entirely random, and could have some balancing aspects enforced).
This would also take significant pressure off otherwise undesirable game situations - like self-destruction of the group under some-strange-circumstance. So long as a playthrough which leaves everyone dead in a couple of hours hardly gives anything away about the next playthrough, it needn't be a problem - so long as the downward spiral is entertaining.
I definitely plan to have randomly generated characters, though the extent will depend on how effectively we can implement them. At best, I want to be able to "roll" personality stats completely at random within reasonable parameters, and at worst, I'd have a pool of pre-generated personality stat lines that have been tested enough to work. And as for self-destruction, I'd hope that it would be entertaining in itself. I'm thinking Dwarf Fortress, Arkham Horror, Dead Rising or even Team Fortress in this respect, where it can actually be more enjoyable when the game becomes an overwhelming struggle to just hang on for dear life and lose with dignity.
And drama. Drama is important. I don't want everyone to die because of some AI quirk where everyone refuses to get their fatal wounds healed by a doctor that's perceived as incompetent, or gets trampled by elephants because they wanted to loot a corpse and take possessions back to a stockpile. However, a complete community implosion where a gunman loses it and shoots the doctor in the head because he couldn't save the gunman's attractive counterpart does have a certain drama to it, especially if there's a slim chance of salvaging something from the aftermath. Same goes for a character who dies trying to pull a fallen teammate back to cover after an elephant trampling. It's going to take an inhuman degree of playtesting, but I'm hoping some kind of automated, time-accelerated prototyping will help in this regard.
I think this is pretty reasonable, since it allows a distinction between highly conflicted/ambivalent/bi-polar emotional states, and the absence of any strong emotion. Someone with both high Joy and high Melancholy might be on a knife-edge between elation and depression, while someone low in both might be in a fairly calm/stable emotional state.
While this kind of thing could be designed in as more permanent personality quirks, I think it'd be good to be able to make such a distinction for anyone. There wouldn't only be inherent bi-polar characters, but also the potential for those driven into that state by circumstances.
This sounds interesting, yet has a higher chance of doing strange things. For simplicities sake, I'd rather avoid such conflicted states. Though, something I've been thinking, is that perhaps on a linear scale from Joy to Melancholy, there exists a band and the characters state fluctuates between the high and low point:
Code:
MELANCHOLY [----||||-] JOY
In this example, the player is more predisposed toward Joy, but can still swing a bit. But I haven't really given a whole lot of thought to the idea aside from the premise of defining a band an allowing mood swings within that. It would certainly be more interesting than a single point on the same scale, but it's yet more complexity to consider.
I think that's basically what you were saying anyway, but maintains the notion of opposing emotions, and (for me) it's easier to visualise than two independent variables. To further rationalise the idea - I think it would be reasonable to afford the player a certain degree of control over where they sit within the band, but for the band to be completely automated.
The important thing is to be clear about what is being represented with each stat: external expressions<---->feelings<---->subconscious mindstates. Each can reasonably be treated differently. E.g. seemingly conflicting states make more sense towards the subconscious end of the scale, whereas clear feedback, understanding and control make more sense on the external expressions end. Squashing these sorts of concept together runs the risk of trivializing the potential conflicts and complexity of mindstates just to avoid nonsense external states (e.g. Sad+Happy).
Yeah, I think we're on the same page. Band represents subconscious, feelings represents a point within that band, and expression is feelings +/- the conscious control the character has.
I think what is difficult about player/designer control over inner emotions is that it is unknown to the designer what type of personality the PC may possesses. You could easily tip off players by dictating 'their' characters how they should feel about a given situation. You could either avoid this by giving the player the option to modify his emotional scales during character creation or by adding thoughts/actions which influence emotions during play (I would favor the later for the sake of added interactivity. Do I tend to be an arse and therefore kick more of those or do I rather prefer to be the calm silent person with dangerous tendencies to laziness). Either way there will be at least one moment where the player will have direct control over inner emotions if you want to keep it fair. The only emotion that should definitely be left alone by the player is fear as the designer's secret boogie man to screw up players intentionally.
In general (as in, I haven't yet considered the personality variables) I want to give the player a degree of control either at character creation or very early in the game, though I'd like to black box it a bit. So rather than the player explicitly specifying "She's a chain-smoking goth bounty hunter trying to make a difference in a man's world", I'm considering the idea of Jagged Alliance 2/Ultimas/Elder Scrolls personality quizzes. I've always liked the concept, but never really liked the often transparent nature of the questions. I think that's mostly related to the way such tests try to tie themselves into the more utilitarian side of character development, rather than the more nebulous personality black box.
I also want to give the player a reasonable amount of ongoing choice to develop as they see fit. Part of this will be fairly direct, through the end of day reflection phase (I've touched on this in another thread, yeah?) and part will be simply the culmination of dialogue options with common trends. For instance, if the player likes to pick dialogue with swearing in it, it'll eventually weight more options in favour of course language and so on.
The more I think about what bugs me about Respect the more I start to believe that it has to do with the Master of Orion 3 vibe I'm getting. Would it be really feasible to make Respect a separated entity? It can be easily absorbed/overruled by most of the other dispositions listed while not having a great effect by itself (Positive effects at least. I would argue that in a small band of people the qualities and flaws of each individual would show up eventually by default). For me, it's like another one of those countless statistics in MoO 3. They didn't really add anything to the experience while making it timid to keep track of them all. To make things worse in this case you wouldn't even know those statistics. If you got many different factors playing together in one situation without the slightest of indication which they were (or with indicators far too vague to be really recognizable) you leave the player to wild presumptions and wonderment why everything turned sour at the end.
MOO3 has to be the best argument against needless complexity the gaming world has ever seen, and I'm thinking respect is pointless if trust is expanded to cover character archetypes (see above somewhere).
Let's revisit S8 initial example and try to elaborate it from my biased position. [...]
Obviously, it's all goign to come down to playtesting and toying with variables, but I'd hope Example A would never occur. I don't think a complete novice should ever be favoured over anyone but the most negligible medic, but of course it's always going to be a possibility if we don't weight the dynamics of it all effectively.
Also, something to keep in mind, is that a "failure to cure <injury>" would never be a single step failure, and I'm thinking it's probably best to bias away from the end result in most cases, so NPCs actually notice the small successes that ultimately didn't quite amount to an overall success. Probably slightly unrealistic given the way most people tend to evaluate according to end results, but practical in this case.
Example B on the other hand, is exactly the sort of emergent drama I want to get out of these dynamic interactions. And for most of those steps, I'd like to provide the ability for meaningful interaction if the player is at hand. For instance, the player argues with Nancy that she doesn't have the necessary skill (low trust in her medical skills) and suggests they wait for Nigel (another NPC he has a high trust of) I'd like the possibility of Dave doing more than just verbally assaulting Nancy, and the player having to step in and grapple and restrain Dave to stop him killing Nancy in a fit of rage. Etc, etc.
Sounds neat. I guess it depends if S8 wants more of a do-it-yourself narrative like in Dwarven Fortress (which is more of a wild collection of events that may or may not take place) or a game with a clearly defined coherent narrative (not saying that they necessarily exclude each other. A NPC personality and dialogue can be fairly static albeit maybe to the price of coherence).
I'm trying to tackle the "DIY narrative" but also attempt to integrate a degree of "recognition" from the game itself so it sees things happening and adds narrative content accordingly. I'll cover that more when I get into Plot Arcs, but basically, the game either has a conditional trigger that kicks off a narrative script, or in some circumstances, forces a narrative script unobtrusively. And that's what I see as being the bulk of the game content. Scripts that look for dynamic interactions and trigger so the player can actually explore a narrative in an interactive fashion rather than simply having fisherman's tales like Boatmurdered. I guess you could liken it to an automated Let's Play, but we'll see.
Another objection I got is if the player creates a medic type of character for example and finds himself in a group with not many good fighters he might have trouble later on. Kinda reminds me of older RPGs where you actually had to roll up your stats randomly. Might be a wee bit too hard to balance out (if desired).
By default, there will be a "balanced" community, where a fairly optimum combination is generated to fit the player's choice, but I also want to allow completely random generation and custom generation for a completely different dynamic.
Whew. Keep the thoughts coming. Anything I missed?
