Brief Critique of Relationship Mechanics

[Walt Freitag]

Hi John,

Your analogy of paying to cement relationships being equivalent to the GM reducing the player's XP unless the player acts in-game to negate the potential relationship doesn't hold up. The type of relationship mechanism we're talking about never requires the player to alienate or drive away or otherswise negate the relationship. The player can cement the relationship or not. Not doing so does not require or dictate or imply any action on the player-character's part. It doesn't mean the relationship disappears, it means it remains under the GM's control. (Or under whatever other form of shared control the game normally uses, but let's assume it's the GM's control for the sake of argument.) The GM can allow the relationship to fade into the background, or alter its nature (e.g. by betrayal), or use it as fodder for deus ex machina plot twists, or have the subject(s) wiped out off-camera, or whatever. (The same is also, often, true of "freebie" relationships in systems where relationships are not backed up by currency mechanics, as Gareth [contracycle] pointed out.)

Describing the currency spent on a relationship as "penalizing" the player is also overstating. I could, for instance, say that the guy who sold me a new suit has "penalized" my bank account, but others might wonder why I've described my choice to buy the suit in such a peculiar way.

Look at the choice of paying XPs for the relationship this way:

GM: Your character has become involved in an interesting new relationship. Do you want to make that relationship an important part of who your character is, and something that you want to make sure continues as an important plot element of the evolving situation in the future?

PLAYER: Um, no.

GM: Okay, I'll give you some points that you can spend later on some other interesting new thing for your character instead.

- Walt

[Jason Lee]

So, if relationships are adjustable on the fly, that says to me the currency isn't a limited resource.  Man, I'm repeating someone now, I'm sure of it.  

If the currency isn't a limited resource, then it doesn't serve game balance in the traditional sense. It could serve a game balance-esque purpose in providing a general measure of effectiveness of a character that you can use to compare to other characters (Bob is better than Sally because he has more points); just some way for the players to quantify who's better than who, so they have a scaling guideline.  Which I'm perfectly peachy with as a purpose for currency - it's how I use it.  But, that isn't typically how currency is viewed.

If the currency is a limited resource that circles back around to my original problem with it implying stability.

[John Kim]

Now, is this a tad abstract? Yes, neccessarily so. The system handles everything with this level of abstraction. That said, it provides the most complete and plausible cataloguing of character abilities that I've yet seen. If it ain't on the sheet, the character ain't got it (DIP aside).  

Well, it is playable to handle relationships abstractly -- but it doesn't need to be abstract.  For example, my Vinland RuneQuest game is extremely relationship-heavy without assigning abstract numbers to relationships.  For this game, at least, I don't want to handle relationships at the same level of abstraction as everything else.  

In contrast, say, I have gone with abstracted Wealth rules in my RQ campaign.  I had a specific intention:  I wanted to convey to players how affluent each character was, but not get too caught up in the details of trading cows, land, and so forth.  

Well, John mentioned that it already makes sense for equipment. But, in fact, I can't think of anything more ephemeral in life than relationships. That guy you used to work with who you swore you'd keep in touch with? Why don't you talk to him now? Lot's of possible in-game reasons, but, mechanically, it's because you didn't pay the HP.  And like equipment, you have to explain how you lose touch.  
...
So, again, I really don't see the problem. If you can accept the metagame idea that you have to pay for equipment (and I can understand if that's your gripe right there), then I think you ought to be able to see how it applies overall.

Well, I said that the superheroic equipment rule is playable and works for some people.  However, I would say it does not make sense on some level.  It is intrinsically irrational, and exists to emulate a particular quality of comic books -- not because it makes sense.  

And I'd agree that seeing equipment or stats as different from relationships is mostly a result of tradition. Let's put strength aside as an example: can you say that Wisdom is something that makes more sense to rate by number than a relative rating of the strength of a relationship? Swordsmandship skill? Painting ability? Give me a system, and I'll come up with something in it that it rates that's as least as abstract as relationships.  

Sure, games are full of abstractions.  Different games will abstract different things.  But I would say that what elements a game chooses to make abstract makes a big difference in the game play.  I think everyone here agrees that it is valid and playable to abstractly rate relationships.  However, some people (like me) are saying that they don't prefer that approach.  I feel the same way about Wisdom, for what its worth.

[Mike Holmes]

Your assumption is that to abstract something in game terms is to neccessarily handle it in the abstract in-game. But the systems in question don't do that. They, in fact, require detailed descriptions of outcomes of Conflicts of the resolution systems. I've heard some of the most realistic and detailed descriptions of combat using the same abstractions. And the use in terms of relationships has only served to introduce them in more detail, not less.

In any case, to not have any system is to handle the relationship entirely in the abstract. That's how you're doing it in your game. You can't have it both ways. Or are you saying that by having system for your combat in RQ that you're abstracting it?

Mechanics, no mechanics, it's all just system, and a way to come up with what happens in-game (Lumpley Principle). How you do it is only a matter of preference.

And it's a valid preference. I've only argued that the other side is valid as well, which is the only point in contention here. So we're not really disagreeing apparently.

Mike

[John Kim]

In any case, to not have any system is to handle the relationship entirely in the abstract. That's how you're doing it in your game. You can't have it both ways. Or are you saying that by having system for your combat in RQ that you're abstracting it?  

You're making the fallacy of systemless play.  The difference here is simply that I do not have a mechanic which assigns a semi-permanent number to each relationship to be written down on the character sheet.  There is still method to what happens in relationships in Vinland.  Non-numeric play  is not necessarily abstract.  Consider combat in largely non-numeric systems, like Everway or Amber.  Combats can still be played out either abstractly (i.e. "After five minutes of fierce struggle, you win.") or in detail (i.e. "OK, you have knocked the fifth thug off his feet.  Now what do you do?").   Sometimes numerical mechanics are abstract.  For example, D&D combat is rather abstract.  On the other hand, Millenium's End combat is also numerical but it is highly detailed.  

By using the RQ system for combat, yes I am picking a particular level of abstraction for my resolution.  For example, there is no distinction between most maneuvers, i.e. all sword swings are a standard attack.  I chose this for a reason.  In the genre I am trying for, combat is not supposed to be cinematic or fanciful.  It is violent and gory and fateful.  When you go into combat you go to meet your fate.  A big part of it is finding out who gets what wounds, to match the family sagas I am basing the game on.  If I wanted more detail, say, for fancy combat maneuvers -- then I might use a different system or significantly change it.  

Similar would apply to relationships.  For example, as I consider it, I can definitely see myself using abstracted relationship mechanics for a political game where I wanted to encourage the PCs to attempt Machiavellian manipulations.  Like in the combat example, I see some logic behind that level of abstraction for what I want.  But for other games, I don't think I would want them or at least would want different ones.  

Mechanics, no mechanics, it's all just system, and a way to come up with what happens in-game (Lumpley Principle). How you do it is only a matter of preference.  

And it's a valid preference. I've only argued that the other side is valid as well, which is the only point in contention here. So we're not really disagreeing apparently.  

Well, yes, but saying that it is all System is a non-statement.  We're agreed that they are valid preferences, but the question is what is driving those preferences.  What differences do we see.  

Your assumption is that to abstract something in game terms is to neccessarily handle it in the abstract in-game. But the systems in question don't do that. They, in fact, require detailed descriptions of outcomes of Conflicts of the resolution systems. I've heard some of the most realistic and detailed descriptions of combat using the same abstractions. And the use in terms of relationships has only served to introduce them in more detail, not less.  

I didn't intend to express that.  I was expressing my preference for how things are handled in game terms.  Yes, you can describe in detail events which are just an abstract result in mechanics terms.  However, the mechanic itself still has an affect.   i.e. Take two combats.  One is just narrated based on a result of "Side B won" -- while the other is played out in detail.  Even if the details of the verbal description are identical, these two combats will feel quite different.  

I would be curious to hear more about your experience with the relationship mechanic.  It seems strange to me that adding a single number to rate a relationship would result in more detail for it.  I certainly didn't experience that in my admittedly brief HeroQuest play.

[RaconteurX]

I do not have a mechanic which assigns a semi-permanent number to each relationship to be written down on the character sheet. There is still method to what happens in relationships in Vinland.

The method, however, can be said to be completely arbitrary. If you as narrator need a relationship to sour, there is very little the player can do to gainsay your decision. In HeroQuest, it is the player who decides, through investment of game currency, whether (and to what extent) a relationship is important to the hero. I find that this makes narrating far easier, as I can rely on my players to provide vital cues to the direction they would like to see the story take by examining their expenditure of Hero Points.

It seems strange to me that adding a single number to rate a relationship would result in more detail for it.

It does not result in more detail, but it does result in less ambiguity since relationships (indeed, any ability in HeroQuest) have tangible mechanical effects... much as passions do in Pendragon. It is most decidedly a design conceit which Greg Stafford considers important as it has appeared, in one form or another, in every prior attempt to render an epic Gloranthan roleplaying game to which I have been privy (and I have seen fragments of at least four completely different attempts). The intent is to provide a concrete distillation of factors into a single value which can be used to simplify a complex process... much as the various weapons skills in RuneQuest do, as a matter of fact.

I certainly didn't experience that in my admittedly brief HeroQuest play.

Your limited experience leads you to make poorly-informed assumptions about the game, I think. Whereas RuneQuest made us think about the importance of culture, HeroQuest goes one step further and asks us to consider the importance of community. The relationship mechanics derive from what I imagine we can all agree is the most central theme of Glorantha, drawn from the archetypal hero's journey of which Greg is so fond: individual as embodiment of his or her community. Surely you would not suggest that so vital a setting element be ignored in the rules?

[John Kim]

I do not have a mechanic which assigns a semi-permanent number to each relationship to be written down on the character sheet. There is still method to what happens in relationships in Vinland.

The method, however, can be said to be completely arbitrary. If you as narrator need a relationship to sour, there is very little the player can do to gainsay your decision. In HeroQuest, it is the player who decides, through investment of game currency, whether (and to what extent) a relationship is important to the hero. I find that this makes narrating far easier, as I can rely on my players to provide vital cues to the direction they would like to see the story take by examining their expenditure of Hero Points.

I guess this depends on how you handle things.  You seem to be arguing in favor of relationships being controlled strictly by Hero Points and player choice.  i.e. If the PC has a fight with someone he has a relationship with, the player still controls the relationship number.  You as GM have no say.  Relationship numbers are thus unlikely to change based on varying events.  Also, it is impossible for relationships to be feigned, except with the player's knowledge.  (i.e. You would have to tell the player that what appears in-game to be a follower doesn't really have a relationship number.)  

That approach does seem more empowering to the players.  However, personally I enjoy seeing the dynamic of relationships change as caused by in-game events.  For example, two sessions ago Melnir (an NPC) became estranged from Poul (a PC) after Poul went berserk.  Melnir is married to Silksif (another PC), and their discussion of what happened was an important point of the game for me.  

If you keep the mechanic but allow the GM to have some control -- for example, to say that killing your own follower causes another to run away -- then this ultimately comes down to GM judgement just like in my case.  The veneer of the number doesn't mean much since the GM is capable of changing the number.  

Whereas RuneQuest made us think about the importance of culture, HeroQuest goes one step further and asks us to consider the importance of community. The relationship mechanics derive from what I imagine we can all agree is the most central theme of Glorantha, drawn from the archetypal hero's journey of which Greg is so fond: individual as embodiment of his or her community. Surely you would not suggest that so vital a setting element be ignored in the rules?  

Well, yes, I am saying that.  Having played both RQ and HQ, RQ is closer to what I want for my game (though I have made a lot of house rules).   I should note that my Vinland game is not set in Glorantha.  It is set in the Hudson river valley in 1392 of an alternate history where the Icelandic Vinland colony flourished rather than fading out.  Nevertheless, I am basing my game on the genre of the Icelandic family sagas, where community is extremely important.  Feuding and peace-making in particular are central themes.  

So community is extremely important.  Does this require numerical mechanics?  I don't think so.  As a counter-question: is theme and moral choice important to your games?  Surely, then, there should be a numerical stat for theme and the players should roll dice for their moral choice and thus the meaning of the story -- right?

[Jason Lee]

If you keep the mechanic but allow the GM to have some control -- for example, to say that killing your own follower causes another to run away -- then this ultimately comes down to GM judgement just like in my case.  The veneer of the number doesn't mean much since the GM is capable of changing the number.

One very tiny quibble, which is oddly a very important distinction to me.  I'm personally less concerned with the GM having control than I am with the player that owns the character in question having control.  Yes, for NPCs, this means the GM is the player who owns the character.  Like I said, it's a quibble, but the difference in perspective has important ramifications when dealing with inter-PC relationships (are ratings allowed for PC's in HQ?), elements of mystery in a setting, betrayal, and some forms of player empowerment.

If the player owns both the target of the relationship and the rating, no problem.  If one player owns the value of the relationship and another player owns the target character (or organization, whatever) I see a strong limitation in power for both parties.

An interesting way to make this work might be having to use the target's relationship rating with you.  Of course, then the resolution would have to be obfuscated somehow.  Ok, I'm straying.

Oh, and just so I'm clear.  This isn't an arguement against you John, hopefully it's a clarification.  Hard to say if you feel the same way though.  Also, whether or not this is best for HeroQuest, I dunno...just talking pros and cons, not right and wrong.

[M. J. Young]

You seem to be arguing in favor of relationships being controlled strictly by Hero Points and player choice.  i.e. If the PC has a fight with someone he has a relationship with, the player still controls the relationship number.  You as GM have no say.  Relationship numbers are thus unlikely to change based on varying events....

If you keep the mechanic but allow the GM to have some control -- for example, to say that killing your own follower causes another to run away -- then this ultimately comes down to GM judgement just like in my case.  The veneer of the number doesn't mean much since the GM is capable of changing the number.

I don't know enough about HeroQuest or RuneQuest to address that aspect of it, but it has jarred my memory that we do have a sort of relationship mechanic in Multiverser, specifically related to what we call associates, but which could be called followers or henchmen or a dozen other things.

The basic rule is that there is a check for the principle (who presumably is a player character) and one for the associate (who presumably is not). However, it is specifically stated that this isn't some sort of binding mechanic--it's an effort to mechanically determine the real relationship between them. The roll for the principle is a charisma check, to see whether his personality continues to cement the relationship; that for the associate is a will power check, to see if he determines to split off in his own direction. Yet what is most significant is that the rolls are specifically not made in most cases.

First, the rolls are never made if both characters are player characters; players presumably can decide whether that fight they had last week has driven a wedge between them or merely created some tension.

Second, the rolls are not made of both characters are non-player characters. Such relationships between non-player characters are plot devices controled by the referee, and what happens between them is entirely up to his discretion.

Third, the will power check is never made unless the charisma check was made. In turn, the charisma check is only made if there's some reason to doubt whether the relationship is still solid--and this is very much a subjective judgment. It almost never happens that a character's spouse would be checked for, because it is presumed that there is a solid relationship there unless there is clear evidence of mistreatment or deterioration in it. Associates who are well treated and have no reason to break the relationship never require charisma checks. The function of the charisma check is to determine whether the primary character was able by virtue of his generally likable nature to overcome any reason he may have given the associate to depart.

If the charisma check is failed, the will power check is still made to see whether the associate asserts his independence; if it is also failed, the relationship remains intact. If the charisma check is successful, the will power check is still made, because a superior success would dissolve the relationship.

The check is only made at the most critical moment: when the player character changes universes, to determine whether the associate goes with him.

So it is an objective character relationship mechanic, but it is only put in play when a subjective determination has been made by the referee that the player character has not maintained his end of the relationship. In a sense, it provides a way to settle the question of whether the player character has been "nice enough" to the NPC or not, without giving that decision entirely to either the player or the referee.

Does that meet your criteria?

--M. J. Young

[John Kim]

We do have a sort of relationship mechanic in Multiverser, specifically related to what we call associates, but which could be called followers or henchmen or a dozen other things.

The basic rule is that there is a check for the principle (who presumably is a player character) and one for the associate (who presumably is not).
...
So it is an objective character relationship mechanic, but it is only put in play when a subjective determination has been made by the referee that the player character has not maintained his end of the relationship. In a sense, it provides a way to settle the question of whether the player character has been "nice enough" to the NPC or not, without giving that decision entirely to either the player or the referee.  

Does that meet your criteria?  

Well, I don't see any logical problems with it -- but from the sound of it, I don't think I would use it.  You are keeping subjective GM judgement, but you are adding in a random roll.  So as I see it, the point of this is injecting randomness.  i.e. If relationships in my game were too predictable, then I might try this mechanic to add more variety/spontaneity.  However, I find that in general relationships are quite unpredictable and don't really feel the need to add more randomness.  

As a side note, I wonder if social status is part of the question here.  Your mechanic is focussed on the PCs having followers.  There was a similar focus in Hero Wars, although relationships are also used more broadly.  

In my Vinland game, social relations are in many different directions.  Some PCs are of greater social status, and some are of lesser social status.  Matunaaga (a PC) is the huscarl of Skallagrim, like an employee or henchman.  Vagnhild (a PC) is the step-daughter of Skallagrim (a PC).   On the other hand, Skallagrim started out as the huscarl of Arnkel (an NPC).  He later acquired his own land and a socially advantageous marriage.  Like most people, the PCs are somewhere in the middle of the social ladder.  The desire for relationships mechanics seems to be more for PCs towards their social inferiors, rather than for the superiors of PCs.

[Aelios]

Lots of opinions here, but I feel we are missing something important.

Given: Relationships are, by nature, unpredictable. A sword skill doesn't get worse if you don't buy it flowers. Even a love of ten years could, and has, forsaken people within a very short time. But this probadly isn't the result of random chance, but it may be uncontrollable (by the PC).
Presumption: Relationships can be bough with Hero Points, experience, or whatever else.
Point: What happens when that relationship dies? Do you "refund" the experience? Is it lost forever? Is that fair to the player?

I see nothing wrong with giving a relationship a numerical value, but if you are going to ask a player to pay for it with experience then they need some guarantee that the relationship is stable, which is not realistic (the guarantee). The only thing a PC can be sure of is how (s)he feels about the NPC. And even that can change quite often.

Perhaps what Daredevel needs is a seperate relationship pool or bank that can be used to allocate points to different relatioships, but that can have points put back into the bank if a relationship goes sour. Or is that just replacing the problem with a new one?

[Gordon C. Landis]

OK, so we want relationships to be an important part of play.  We can accomplish that (I'll claim) in many ways.  Let's have some categories here:

1)  The relationships are given some numerical representation in one or more parts of the game system.  Hero Quest works this way.
2)  The relationships are represented in the system in some particular, methodical way, but are not literally quantified.  As far as I can tell, John's Vinland game works this way (the fact that his method is of his own devising and not explictly part of a published system is unimportant to this analysis).
3)  There is no specific method to deal with/manage/handle relationships.  This does not mean relationships are system-less, under the Lumpley Principle use of System, just that the actual "system" used is non-numeric and non-methodical.
4)  Relationships and/or the impact thereof do not occur in play.  This one is mostly theoretical - I think you could almost always find something that fits in #3 - but for completeness, here it is.

I think we can all accept as given that a particular play group can succede in realtionship-heavy play in any of the first three descriptions (and maybe have fun in #44, if that's what turns 'em on).  It seemed like Joachim was worried that #1 might in fact work AGAINST that, and sure, each description can have particular problems.  They share some issues -  like who has access to influencing the impact relationships have on play, and via what means?  But they can all be overcome.  The question, I guess, is HOW?

I and others have already covered why we think Joachim's concerns don't HAVE to be problematic under description #1.  Equally, John points out why his system works for him.  Neither approach invalidates the other, and at some level, it becomes a question of taste.  The best we can hope for (I guess) is to help those who want to find a way past the concerns they have with a particular method to do so.  Along which lines . . .

Aelios brings up a new concern - that (my paraphrase) if you allow the "effectiveness" granted by a numerically-quantified relationship to go away when the relationship does (as would happen in "reality"), what happens to the points that were "spent" on that relationship?  Regaining points lost in such a way sounds like a fine solution to me. Note, though, that the whole "number on the relationship must reflect the current reality of the relationship" assumption is NOT one that HQ (and others) necessarily subscribe to.  The relationship goes to hell - you could still USE that.  A 10 year love affair ends - there's still a LOT of emotional energy there, for a long time, which the player/GM/group can use to inform why the "Leila-12" ranking still applies to that character.

Again - the HQ (and other) relationship method does not neccessarily mirror the reality of what's happening - it need only tell you what "source" a player is able to draw on in fueling their effectiveness.  Why and how that fueling occurs can be entirely determined by the play group.

Gordon

[John Kim]

OK, so we want relationships to be an important part of play.  
...
I think we can all accept as given that a particular play group can succeed in relationship-heavy play in any of the first three descriptions (and maybe have fun in #4, if that's what turns 'em on).  

What interests me more is the differences between the methods.  To take the Champions parallel again...  Both the superheroic equipment rule and the heroic equipment rule work, and both can be used in games which emphasize equipment.  But they work in different ways, and they are not seamless replacements for each other.  The rules do make a difference.  

I would think that the HQ relationship rules also make a difference.  So I could use them and still have a relationship-heavy game, but it would be different from my earlier games.  The approach in your post is that these are just different approaches towards the same goal.  I suspect, however, that the goal is subtlely different.  Relationships are important in HeroQuest, but I suspect they have a different focus and/or purpose than relationships in my Vinland game.  

I guess I should give some examples of what to me was important in Vinland relationships.  The most important relationship recently has been the one between Silksif (a PC) and her husband Melnir (an NPC).  Melnir had been a continuing character from early in the game, but it came as a complete surprise 6 or 7 sessions ago when he arrived in force at Tjaraholt and asked for the hand of Silksif in marriage.  She accepted mostly because of his generous offer of providing for her family.  Since then, she has been slowly figuring out his attraction to her and what he is about.  

I guess I see this in part as a mystery.  For a while, Silksif as well as Heather (the player) weren't sure about what Melnir's true feelings were.  If we were playing in a numerically-rated system, I think that I would have told Heather about the relationship rating.  She could then role-play Silksif as not knowing, of course, but it would be a significantly different play experience.

[Aelios]

John, what kind of system do you use for your relationships?
I think I can safely assume you don't assign numerical values for each PC/NPC relaitonship. Do you jot down notes like "Melnir is falling in love with Silksif" and just play it by ear or do you have a more concrete method?

[John Kim]

John, what kind of system do you use for your relationships?  I think I can safely assume you don't assign numerical values for each PC/NPC relaitonship. Do you jot down notes like "Melnir is falling in love with Silksif" and just play it by ear or do you have a more concrete method?

For something like this that is fairly major, I wouldn't say that I play it by ear, but it isn't objectively systematic.  I developed Melnir as a detailed character, of which his attraction to Silksif eventually became a part.  

In this case, system came into the question through Whimsy Cards.  Our rule is that the player of a given character has veto power over any card which controls their play, but should seriously consider it.  In one session, there was a feast at Brygjafael.  Katrina was trying to attract Melnir.  One player (Liz) played the "Romantic Interest" card to suggest he was interested.  However, Silksif later intruded on this, and another player (Laura) played the "Misguided Love" card and declared that actually Melnir only had eyes for Silksif and was only talking to Katrina to get her attention.  

That said, it could simply have remained casual flirting at a party according to the system.  However, as I thought about Melnir's character after that session, I decided that it was something much more serious.  I did certainly give weight that card play should have lasting importance, but more importantly it resonated with his character.  Melnir is a skeptic of things supernatural, while Silksif is a gydja (i.e. prophetess or shaman) of the Norse tradition.  Melnir is practical rather than prejudiced, though, so to him her status was exotic and intriguing.  

On the other hand, other relationships have happened without card play.  For example, Kjartan (a PC) is engaged to be married to Thjohild of Groenholt (an NPC).  Kjartan had been told by his grandmother what house he should marry into (Groenholt, the stead of Vigfus the Proud), but he could choose whom he wanted among the grand-daughters of Vigfus.  I had described all of the women of Groenholt (a paragraph each), and Liz the player debated which would make the best match for Kjartan.