Is Alignment an Anti-pattern?

[Mike Holmes]

Well I think identifying alignment as a pattern that doesn't incentivize certain kinds of play is problematic. Because for the most part it does (or even forces adherence). Rather, most of what are alignment rules are actually what you'd define as some sort of reward, right? Or penalty? Or guage, or characteristic? In Hero Quest a character could have "Antisocial" as an ability (characteristic), and it would be used normally. In D&D a player might be penalized EXP for playing outside of alignment, and this is a reward system. Etc, etc.

I've can't think of a system that has "alignment" that as a mechanic isn't some other pattern. What you describe above as alignment is a characteristic (trait?). That is, how is "Tall" as a trait that merely has certain effects listed for it not the same as "Lawful Good" if it's mechanically handled in the same fashion? I'd just chuck alignment altogether as a pattern. I'm not seeing it anywhere as distinct from other patterns.

Then I'd point out in the other appropriate patterns how alignment in game X is an example of that pattern.

Mike

[John Kirk]

Well I think identifying alignment as a pattern that doesn't incentivize certain kinds of play is problematic. Because for the most part it does (or even forces adherence). Rather, most of what are alignment rules are actually what you'd define as some sort of reward, right? Or penalty? Or guage, or characteristic? In Hero Quest a character could have "Antisocial" as an ability (characteristic), and it would be used normally. In D&D a player might be penalized EXP for playing outside of alignment, and this is a reward system. Etc, etc.

Well, it's possible that I need to rename Alignment to something else because the name itself might interfere with some people's understanding of it, given their pre-conceived notions of what it must mean.  (Any ideas there?)   The pattern, as currently written, does not "incentivize certain kinds of play", as you say.  It contains no reward system and does not enforce "adherence" because there is no "adherence" to enforce.  The pattern refers to a common characteristic that is set to one of a pre-defined set of values.  It is not a gauge, because their is no graduated scale associated with it, only categories. It is also not a trait, because the game designer must provide the categories.

In fact, the pattern has far less to do with the actual common characteristic and its possible values than the impact it has on the rest of the game structure.  I believe it is this feature that distinguishes it from other patterns.  It essentially must be used in conjuction with pre-defined lists of abilities and/or classes because it is there that the pattern truly appears.  Any skill, handicap, gift, or flaw that has differing effects based on what category into which a character falls exhibits the pattern.

In software development, this pattern is characterized by a lot of "switch" statements, or numerous "if, else if, else if" blocks scattered throughout the code that cause a program to behave differently based on a characteristic of some object.  It is considered an anti-pattern in software development because of the problems it introduces when a new category is added.  The same is a problem in game design.  Suppose, for example, a new alignment category was added to a simple "Good" and "Evil" alignment system.  Say, "Wild" was added.  This would require every ability whose effects differed based on alignment categories to be visited and re-written to accomodate the new category.  That is potentially a lot of work and is something to be avoided if possible.

But, if a game has a set of alignment categories that is not going to change because the categories themselves are core to a game's focus, the arguments against the pattern's use are weakened.  After all, D&D's alignment categories have not changed over the years.  So, no "re-visiting" and "re-writing" has been necessary because of this.

[Mike Holmes]

You might go to programming here for a label. This would be a "Property" of the charater, no? Basically a flag for booleans. "Flag" might work, too.

Because, again, the problem is that rarely does alignment only work this way. Usually in addition to being a flag for these sorts of tests, it's also linked to a reward system, and/or other stuff.

Mike

[M. J. Young]

That which in Multiverser is most like alignment is called "affiliation".

It is mandatory for "gods" and most supernatural entities, but optional for characters. It has essentially a positive, negative, and neutral position.

Multiverser has no classes. Affiliation will impact the use of some kinds of magic, particularly "holy magic", obtained by petitioning a supernatural being for assistance, but anyone can do that.

As with D&D alignment, the three slots are all-inclusive: anyone has to be either for, against, or uncommitted, as there are no other possibilities.

I mention all this because a lot of what you say about "alignment" doesn't fit my understanding of it, probably for these reasons, and because "affiliation" might work as a term for what you're saying.

--M. J. Young

[Mike Holmes]

MJ,

Actually this sounds pretty much precisely like what Kirk is talking about for Alignment. Though you might want to look at his "faction" pattern to see if that's part of it, too.

What I'm thinking is that alignment and such are often "Alignment" by Kirks definitions (I'd call em flags or something), Factions, and reward systems or guages.

Mike

[Shimeran]

Here's a quick list of what seems to define alignments in general:
  - Personality Classification: Alignments attempt to sort people into personality types.
  - Mandatory: Everyone has an alignment.  The classification system is set up so that any target can be placed in a category.
  - Exclusive: No one can have more than one alignment.  Being in one category excludes membership in others.

If we compare these to idioms, we get a few key differences.
  - Breadth: While alignment tries to classify entire personalities, idioms focus on aspect of one's personality.
  - Measurement: Alignment is an all-or-nothing deal while idioms can have a range of values.
Note that idioms also tend to: 1) have mechanical effects, 2) be optional rather than mandatory, and 3) allow more than one to such trait.

An alignment is basically character classifier (supposedly based on the characters personality, beliefs, and/or behaviour), a way of grouping individuals into categories.  By itself, they do nothing.  After all, you could as easily use things like Myer-Briggs personality types to do the same thing.  However, many games also attach mechanical effects to alignments.  These effects tend to either enforce or rely on alignments.

Alignment enforce helps force the character into certain types.  Note that these systems tend to focus more on enforcing alignment consistant behaviour rather than thoughts or beliefs.  The assumption being that people act in accordance with their personalities.  A system with strong alignment enforcement supports archetypical personalities.  On the other hand complex personalities that don't fit any categories particularly well are hindered.  In short, alignment enforcement is good if you favor simple, archetypic character types over more complex and potentially contradictory personalities.  This enforcement can be negative (punishments and penalties) or positive (awards and bonuses).  Negative enforcement leads characters to avoid actions that "break" the classification while positive enforcement encourages certain actions.  In short, penalties limit actions while bonuses promote them.

Alignment targetting is basically a modifier for other abilities.  In essence, the designer slaps a "only works on this character category" limitation on the ability.  This makes these classifications far more important, at least to players with such abilities.  After all, it tells them who their targets are.  For example, consider a character with "Heal Good" the other with "Smite Evil".  He's more likely to ally with Good character since he's better at playing with them but no better at fighting them.  On the other hand, he's more likely to fight Evil characters since he's better at it.  The more prevalent alignment targetting abilities are the more important alignment becomes and the bigger it's impact on player reactions.

Alignment dependant abilities are modifiers that apply alignment enforcement.  In short, the designer says "you must fit this personality type to use the ability".  Most such abilities work on negative enforce (they fail to work when alignment is "broken"), though an ability could be set to trigger when character's exemplify their alignment.  When combined with alignment targetting, these abilities can do a lot to set the relationship between alignment types.  If these linking abilities are positive, the groups will tend to cooperate (since aiding is boosted but harming isn't), while if they're negative opposition and hostility will be more common (as those harm abilities can only be used on certain targets).

Alignments are usually fixed in number, though this is not required. For example, White Wolf's Nature and Demeanor fit all the other marks of alignment.  However, there's not a fixed number of them.  While characters in WW can have only one of each, they can invent their own nature and demeanor or find extra options in supplements. Variable alignment sets tend to be based on the character's strongest personality traits.  Alignment enforcement can apply to fixed or variable alignment sets, so long as the guidelines for that particular character's alignment are set.  Alignment dependant abilities can affect fixed vs variable sets.  If the dependancy is also variable, them it works fine with either, but tends to weaken type distinctions.  If the dependancies are fixed, then each such ability makes that alignment that much more apealing.  Alignment targetting abilities lean even more strongly towards a fixed alignment set.  After all, fixed sets have a finite (usually fairly small) size, which means people with the target alignment will show up more, thus allowing that ability to be used more.

That's my break down on alignments so far.  Hopefully you'll find it helpful.  By the by, the design patterns book is very interesting and certainly helps break things down.

[brightstar]

I think those that claim that Storyteller Factions (Clans, Traditions, Tribes) are a form of Alignment are making an enormous leap of logic. 

With the chief system function of alignment you can detect it, you can form protections against it, it restricts character options (as listed in the Book of Vile Darkness) etc.  A Storyteller Faction are not controlled by this mechanism. You cannot detect a Brujah.  As a Brujah you are not power restricted based on Clan, Tribe, etc.  They do have favored disciplines, gifts etc, but they do not have to follow them.   

This next point some have disagreed with here, but I went back and read the D&D 3.0 Alignment descriptions in my hardback.  Putting aside my own subjectivity it becomes clear that yes they are designed for the purpose of "morality" to enhance roleplay.  From here you can make an assumption that Storyteller Factions deliver a similar result with dictating morality and roleplaying choices through their splats (i.e. Brujah believe in Anarchy - chaos, Ventrue believe in Order - Lawful).  However this is where Alignment and Storyteller Factions part ways.  Within a Storyteller Faction, the design tool is a quick interface into that particular group, but one can deviate from the norm as much as one wants without penalty to the character or retribution from the system.  I.E. a Ventrue who wants to commit Anarchy - not be lawful can do so of their own free will.  Yet in D&D breach of Alignment code (strictly a roleplay decision, for those who think it is not part of roleplay) can have disastrous effects on your character.  For example A Paladin must be Lawful good.  Committing acts against their alignment will cost them their class special abilities.  This function is repeated throughout various classes in the D&D manual.  Being a Ventrue committed to Anarchy rather than social order costs nothing to a character, ever, except in roleplay where he might run into trouble with his elders.  But that is not an inherent system penalty, rather an "in game" penalty that can be chosen or not chosen by those involved in the game. 

My third point of the failures of Alignment in the context of D&D (which is a very system heavy game) is that it is the only rule left arbitrary.  It spends no time clearly defining what would be a failure in law, chaos, good, evil, etc and what would be a success.  There are no rewards for playing ones alignment.  There is really no delineation between the terminology.  Therefore it becomes the only part of the system left to completely arbitrary decisions on the part of the Game Master and player.  These terms, of course, are then defined by a particular gaming group.  By the inherent nature then, of alignment being a debatable point, it goes against the very nature of the rest of the system. 

What makes d20 so strong is that nothing is debatable.  There is a page xx with a clearly defined rule for just about everything that players and Game Masters can think up so arbitration on the part of the Game Master becomes a moot point.  There is a rule that has already made those decisions for him all he has to do is see page xx when a dispute comes up.  But with alignment, there is no page xx.  He is abandoned by the system and left to his own devices to finally make a decision on his own, which normally, as I've seen time and time again, leads to arguments about what determines good and what determines evil.  The only time this does not occur is when players of similar religious/philosophical backgrounds are playing together.  Then alignment itself becomes clear. But if a Christian and Cannibal were playing together, one would define eating human flesh as evil, the latter good. 

Looking at my third point and using the actual definition of anti (opposite, against) I would say that alignment is an anti-pattern to D&D.  It is not cohesive to the rest of the system.  But before I can say for sure, I must look at this example from Nobilis, which was the most convincing argument for its existence.           

Code of the Heaven
1.) Beauty is the highest principle.
2.) Justice is a form of beauty.
3.) Lesser beings should respect their betters.

Code of the Wild
1.) Freedom is the highest principle.
2.) Sanity and mundanity are prisons.
3.) Give in kind with a gift received.
(sorry for the color instead of a quote, couldn't find it on the list and I'm running out of time). 

These two descriptors are meant to allow quick interface into what it means to be a God.  These are then defined as Alignment in Nobilis. 

As a roleplay tool, yes, they are decent defining aspects for playing a God.  As a system to base how one receives Miracle points they work. 

But there is an alternative all together that has been left out thus far in this argument.  The Storyteller System employs four clear methods for creating this type of morality without confusion and within the simplicity of the system. 

The first part of these being Vampire The Masquerade had Paths and Kindred of the East: Dharmas.  These came with an overall description of the ideology of the path (like Alignment).  They were reinforced with ten levels, each with their own descriptor.  Now, the function of these two ideas parted ways in the way they played out in the system. 

In Vampire, being above a certain rating meant you behaved a certain way and obeyed the rules of all the lower points on the track.  Failure to follow ones path meant you slid downward on the track.  Once a character fell to the third level, they were quick to frenzy and needed to begin to reform themselves or chose a new path. 

In the Dharma mechanism, how good one was in one's dharma withheld access to higher rated powers.  Failure to meet the demands of your level of Dharma or below could lead to a temporary loss of power (much like the Paladin's failure to be Lawful Good). 

The Personality Archtypes of the old storyteller system also gave a quick interface into a moral and roleplay system mechanic.  These are similar to Alignment since they do not have numerical values attached to them and a player simply chooses a Nature and Demeanor and went with it.  However, they differed because a clear reward system was put in place for roleplaying ones nature and demeanor in the form of Willpower regain and Experience Points and an exact description of how those points were awarded.  Thus saying, this is how you play x archetype.  They also differ since Archetypes could not be detected by magical means and that loss of one's archetype had no dire consequence for certain characters.     

Lastly, Storyteller created in Werewolf: The Apocalypse Glory, Honor and Wisdom.  Based on ones deeds in a given session one gained points that determined advancement in your power rank.  These take on a similar trapping as alignment because they assert that what it means to be a good Garou was to do these type of actions, thereby creating a moral code for Garou "to obey".  Again, where it part ways with alignment like the storyteller factions is that it is still open ended for free choice. 

Any of the four above systems are superior to the Alignment's of Nobilis.  Why?  Because they offer a broader range of possibilities and could set perfectly clear ways of gaining the Miracle Points.  For instance the Glory, Honor and Wisdom system could be transcribed and broken down into the major aspects of the dieties.  (there is a previous post here concerning those categories, don't have time to re-post them now but they were things like wind, good, evil, etc).  the Personality archetype mechanic could also be used to define the different aspects of a God with things such as hospitality, happiness, luck, fate, etc.   A player chooses a number of them which would then define their godly outlook and way of being.  A Moral Path such as the Dharma of Kindred of The East and The Paths in Vampire would not only show how good of a God you are (because God's do make bad calls and can go against their natures, it has happened before) but it would also allow a clear delineation of how one gains Miracle Points (as well as possibly, lose them or be hampered in some other way). 

Similarly, this path structure could be used for Paladin's code of conduct to clearly transcribe what could cost them their powers.     

In conclusion, based on this analysis, Alignment is an Anti-pattern. 

As a final thought though, let us look at Alignment in terms of the history of gaming.  It was the first attempt at imposing Roleplay upon the Roleplaying universe, originating with D&D.  But over the years there have been a number of advancements in how to treat function of morality with a character.  I used storyteller as my example, but it is not the only example of this change in technique.  I don't feel it is wrong to say it is an Anti-pattern and that it is inferior to more recent developments and that it should never be used.  Hand Cranked cars were once the norm for people, but we don't use them anymore.  Because it was a by far inferior mechanic to the automatic start engines we have now.  Therefore it was made obsolete. 

My problem with Roleplaying in general is there is too much nostalgia for archaic out of date concepts and little focus and promotion (in the mainstream) on progress and innovation.

[John Kirk]

Shimeran and Brightstar,

Welcome to the Forge!

It is obvious to me that the Alignment pattern needs to be re-named to cut through any preconceived notions of what the pattern is about.  Something Shimeran said made me think of a term that just might fit:

Alignment targetting is basically a modifier for other abilities. In essence, the designer slaps a "only works on this character category" limitation on the ability. This makes these classifications far more important, at least to players with such abilities. After all, it tells them who their targets are. For example, consider a character with "Heal Good" the other with "Smite Evil". He's more likely to ally with Good character since he's better at playing with them but no better at fighting them. On the other hand, he's more likely to fight Evil characters since he's better at it. The more prevalent alignment targetting abilities are the more important alignment becomes and the bigger it's impact on player reactions.

That is a good point.  Characters often tend to group with characters of the same category.  So, what we need is a term that describes the categorization of characters based on behavior and/or physical characteristics.  Environmental effects may be different based on a character's category and, based on Shimeran's comment, the term also needs to convey the point that characters of a particular category may tend to "flock together".  There is a word that fits all of these requirements: Demographics.

A game with "Good" and "Evil" categories may contain a concept of Heaven and Hell.  Hell's demographic is that it is populated with Evil spirits, such as devils and demons.  Heaven's demographic is that it contains Good spirits, such as angels.  It is possible that a game world has a mixed demographic overall, but individual cities may have a primarily Good or Evil demographic.

Modern society has a smoking and non-smoking demographic.  Depending on which of these demographics you fall into, you may have very different physical reactions if forced to smoke a cigarette at gunpoint.  Smokers will have no physical reaction while non-smokers are likely to cough and wheeze and may even become ill from the experience.

What do you think?  Should the Alignment pattern, as listed above, be renamed the Demographics pattern?

If we did this, the "Alignment" systems of several games could be classified as combinations of the Demographics and Faction patterns as follows:

D&D: Demographics + Faction
Legendary Quest: Demographics
Nobilis: Faction
Rifts: Demographics
Warhammer Fantasy Role-Play: Demographics + Faction
The World of Darkness: Faction

[brightstar]

I disagree with the demographics idea.  Keep Faction as you state, because yes, factions have a tendency to group, but alignment does not necessarily have the same function.  But evil does not necessarily get "along" nor group with each other.  Nor does good.  The terms are so relative they could be defined as anything.  Like I said before, eating human flesh in one culture is bad, another is good.  They both would have the alignment good if played within their basic culture.  Of course, the players could set a bias for the moral interplay of their genre, but that is up to them alone. 

The other word I was kicking around was archetypes.  Good and Evil are purely archetypes.  They relate closer with the true definition of archetypes.  This way the Nobilis system, which is working off an archetype can be included as well as old storyteller nature and demeanor, but still inclusive to their new Virtues and Vices.  Basically, the purspose of these things is to influence roleplay choices bottom line.  That's their core function.  The second layer of that function is to sometimes perscribe a system mechanic stemming from them such as detect good/evil, will power regain, gain miracle points.  This is a mechanical archetype where type two would be purely aesthetics (but come to think of it, I can't think of an archetype in any system that does not have a system side note which is key to the function of the character's success or failure.)  I just think its a stronger word then demographic, which is broad, vague, and somewhat confusing.  Technically fried chicken eaters and non fried chicken eaters are demographics...therefore, assumably, friend chicken eater could be the fourth D&D alignment and all fried chicken eaters cannot be lumped together and many will never, ever get along.   

 

[alejandro]

well i'm new here, and i have to say that i have always hated alignment. i think that alignment is an anti-pattern that imposes artificial control over a PC, now this is coming from a dnd background, i can appreciate what mj was getting at earlier (sorry that i didn't copy his post, but i don't really know how, it almost convinced me,) but i agree with mike holmes (?) in that it is a good idea but poorly implemented, again as related to dnd.

the problem i have with the dnd alignment system is that 'good' and 'evil' is relative to the individual defining the terms, though i don't agree with either hitler or lucifer they probably would have had different examples of 'good' and 'evil' had they written an rpg.

now i apologize, but i didn't actually read ever single post, but i do tend to agree with the demographic notion of handling alignment, if it is a system that needs to impose this type of role playing control over a PC, planescape for example. other then this i think that alignment is only best used for a gm to better understand NPC's.

that said, i think the best way to handle alignment is to encourage the players to develop a 'good' background, from which any alliances, factions, demographics, etc can be extrapolated and thus the gm can deal with 'alignment' shifts through game play, and not through direct xp penalties (if you screw an important contact, that contact won't trust you much thereafter, etc.)

if a character is encouraged to develop a background that is a) consistant to the setting, and b) strong enough that the character cares about it, then no abstract mechanics that don't make sense need to be imposed on a player.

so to finish, i believe that elements that impose mechanical penalties on acting is generally a detriment to the game, those elements should be dealt with through the effects it has on an unfolding story.

thanks and good night, alex

[Shimeran]

Let's say we've got a pattern called "Classifiers".  A classifier is basically a descriptor for some aspect of the character.  Unlike more open-ended descriptors (such as one's name), adding a classifier involves picking a single option from a given set.  Player's usually only get one pick per set.  Each set in turn covers a certain aspect of the character, such as their personality, background, origins, hair color, ect..  By themselves classifier have little game effect.  However, there are two ways to increase the importance of a given set of classifiers.

If there are abilities that rely on the classifier, it suddenly becomes much more important.  For example, some abilities might only work if the user has a specific classifier, while others might only target characters with that classifier.  Combining these modifiers tends to imply certain relationships between classifier members.  For example, if followers of Blorg have a damaging effect that only works on followers of Glat, it's probably safe to say the two sides aren't too fond of each other.

Another option is to attach templates to a classifier.  Since each option costs the same in a given set, it's generally a good idea to make sure they're of roughly equal value if you want an even distribution of choices.  Unlike normal templates, the player loses all attached characteristics if they drop this classifier at a later date.  However, since classifiers are more often changed than just lost, this simply means trading the old characteristics for those of the new classifier.  Note that some classifiers have attached effects that make it harder or less desirable to change classifiers.

From here we could model alignment as a personality classifier.  D&D style alignments are basically behavioural classifiers with the flaw "experience penalty if alignment code is broken".  Similiarly, D&D contains effects that target or depend on one's alignment.  The system also encourages opposition along the good-evil and lawful-chaotic axises by giving each side effects that make them more combat effective against the other side.

Note that things other than alignment can be modelled with classifiers.  The key thing is to provide a selection from a set of values where each option has a similiar value and properties.  For example, classifier could work well for "elemental attunement" where character get a +1 to use and protection from their favored element, but -1 to resist their opposing element.

Classifiers are useful when you want:
- a key word to cover the target characteristic (providing simplicity and focus)
- selection from a set of options (usually predefined rather than custom traits)
- the ability to differentiate characters later based on that characteristic