Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  

Author Topic: Some science about dwarven relationships  (Read 2303 times)

Matoro

  • Bay Watcher
  • if you drive alone you drive with hitler
    • View Profile
Some science about dwarven relationships
« on: June 09, 2013, 11:49:09 am »

I've done some science about lovers - how to effectively make lovers and which kind of dwarves will fall in love more easily.

I'm playing with fortress without mountainhome (=no migrants), so after I've got everything running smoothly, I decided to make some couples for population growth. At first I just locked my every dwarf (there's 20 of them) into legendary dining room with food and beds. It really didn't work, because everyone socializes with everyone. It just makes giant amount of "passing aquitances", not real friends or lovers.

Then I built some rooms for two dwarves with lots of food and booze in them and burrowed three pairs into their own rooms; two of the pairs were friends and one had just "passing aqusintance" to each other. Because wiki says that lovers are picked from friend pool, I was of course excepting those two friend-pairs to became lovers fast.

I locked them into the rooms during the late winter. At the spring the pair 2 - the ones with just "passing aquaintance", not "friends" - had become lovers. Pairs 1 and 3 haven't made any change to thei relationships, they were still just friends.

I waited some more. It's now late spring, no changes in the relationships. I began to check their personality traits. I have two pairs of lovers in my fort, other on is from fortresses first years, so it's "natural". I found one trait which is shared by 3/4 of dwarven lovers - "he/she is candid and sincere in dealings with others".  Those pairs don't share any preferences.

Then I found that the other one from pair 3 has this trait also, and she didn't generate a lover from her friend. Se, however, has a "She doesen't experience strong cravings or urges". None of the test dwarves is particulary antisocial.

I begin to question if Wiki's "Lovers develop from a dwarf's pool of friends" is correct at all. Pair 2 went directly from "passing aquaintance" to lover (or the friend phase was so short that I didn't notice it, I checked their relationships about once in a week.) but the friend-pairs don't do anything.

I'll let more time roll (I'm going to keep them in those rooms at least to autumn, maybe whole year) and see if anything happens.

EDIT: It's midsummer and happy dwarves of pair 2 have married.
« Last Edit: June 09, 2013, 11:58:47 am by Matoro »
Logged

Telgin

  • Bay Watcher
  • Professional Programmer
    • View Profile
Re: Some science about dwarven relationships
« Reply #1 on: June 09, 2013, 12:09:41 pm »

I don't know if it's been changed since I updated it, but I added some of that to the wiki based on a response from Toady to a PM I sent him on the matter.  It's possible I misinterpreted his PM, and I deleted it at some point and no longer have it as a reference.  :(  He didn't tell me to keep it secret, so I'll paraphrase it to the best of my memory.

The way he described it working was based on "chats", where idle dwarves that are nearby will have those chats.  He told me that there was a compatibility calculation that determined if dwarves would form grudges, whether or not they'd ever make friends with another dwarf or just acquiantances, or whether they'd become lovers.  He told me that after a certain number of chats a dwarf would decide whether to become a friend or acquiantance, after a few more they'd decide whether to become lovers or not, and after 50 I believe they'd decide if they should get married or wait.

The compatibility calculation got a large bonus if they had a skill in common (Novice or better?), a large bonus if they shared a specific thing they liked (like cats, or whatever), and the base value was based on personality traits.  Those bonuses could easily overcome the base compatibility, so even conflicting personalities wouldn't preclude dwarves becoming lovers.

Based on my understanding of it and my observations with DFHack (I believe the strength of the aquintances (sic) array is the chat counter he was talking about), I don't think the number of chats is a literal number of times that they talk with each other.  I think it's influenced by social skills and possibly compatibility, and they might gain more than one "chat" per chat.  If that theory is correct, that would explain why some dwarves jump from acquaintance to lover directly.

Nevertheless, this is interesting and more hard data is good to have.  There is no better source than the game itself for these things, as the emergent behavior can sometimes work out differently than the theory behind it.  Or reveal bugs or other misunderstandings.
Logged
Through pain, I find wisdom.

anzelm

  • Bay Watcher
    • View Profile
Re: Some science about dwarven relationships
« Reply #2 on: June 09, 2013, 01:11:29 pm »

[...] The compatibility calculation got a large bonus if they had a skill in common (Novice or better?) [...]

This implies that it might be worthwile to divide dwarves into guilds and burrow the guilds in separate meeting areas.
Logged
And they raise their wooden pints
And they yoik and sing
And they fight and dance till the morning!

Matoro

  • Bay Watcher
  • if you drive alone you drive with hitler
    • View Profile
Re: Some science about dwarven relationships
« Reply #3 on: June 09, 2013, 02:21:55 pm »

Okay, I found what my problem was. I forgot to check the ages of the dwarves. That's the reason why pair 1 and pair 3 didn't become lovers. Now I have succesfully married all my female dwarves with 2x3 room with bed, table, chair and food stockpiles. None of the pairs has many common preferences or skills. Just being roughly same age makes sure that they will become lovers, if they have nothing to do but just chat with each other. 
Logged

Deepblade

  • Bay Watcher
  • Tholtarmid
    • View Profile
Re: Some science about dwarven relationships
« Reply #4 on: June 09, 2013, 03:43:07 pm »

I bet the conversations go something like this.

Urist McMan: "So, how old are you, what do you like, and how likely are you to carry our baby into battle?"
Inod McWoman: "I'm 74, I like Steel,  penguins for their black and white feathers, and I prefer to drink Dwarven Rum. Also, I am very likely to carry a baby into battle."
Urist McMan: "Hmmm, I'm 77, I like Bronze, Tentacle demons for corrupt intentions, and I prefer to drink Gutter Cruor. I think babies taken into battle is a very important thing to do. 2 out of 4 points for my perfect woman. Want to get married?
Inod McWoman: "Of course I do. Wait, we're not related are we?"
Urist McMan: "I don't think so."
I
Logged
Deepblade's Standardized Creature Parts, for when you're pissed about all the different types of animal products there are.

Saraias

  • Bay Watcher
    • View Profile
Re: Some science about dwarven relationships
« Reply #5 on: June 09, 2013, 05:22:46 pm »

I am keenly interested in dwarf genealogy, and family/bloodline management is focal to my forts. I sort unmarried migrants for profession but also age and bloodline; as children grow to young adults, I note their maturity year, sex, and lineage at the same time I sort them for profession based on preferences. (In fact, I just paused in the middle of that sentence to add a young lady to the "now eligible" pool.) I assiduously arrange marriages and don't have eligible individuals with potential partners single for long. Organically developed romances are celebrated in the fort, of course, considering that it can be difficult for singles to meet and get acquainted in a bustling - often sprawling - settlement, although it can result in some suboptimal pairings. (e.g. In my last fort I had two couples who had no less than six of their children marry each other. It was great that the families were so close, but the next generation who were cousins on both sides had to be managed carefully into as many different other families as possible when the time came, to suit my "broad, strong, bloodlines" objectives.)

The way he described it working was based on "chats", where idle dwarves that are nearby will have those chats.  He told me that there was a compatibility calculation that determined if dwarves would form grudges, whether or not they'd ever make friends with another dwarf or just acquiantances, or whether they'd become lovers.  He told me that after a certain number of chats a dwarf would decide whether to become a friend or acquiantance, after a few more they'd decide whether to become lovers or not, and after 50 I believe they'd decide if they should get married or wait.

The compatibility calculation got a large bonus if they had a skill in common (Novice or better?), a large bonus if they shared a specific thing they liked (like cats, or whatever), and the base value was based on personality traits.  Those bonuses could easily overcome the base compatibility, so even conflicting personalities wouldn't preclude dwarves becoming lovers.

Based on my understanding of it and my observations with DFHack (I believe the strength of the aquintances (sic) array is the chat counter he was talking about), I don't think the number of chats is a literal number of times that they talk with each other.  I think it's influenced by social skills and possibly compatibility, and they might gain more than one "chat" per chat.  If that theory is correct, that would explain why some dwarves jump from acquaintance to lover directly.

This, plus the quintessential ten-year eligible partner range, is consistent with my observations of organically developed romances or those formed among couples who mingle in group Socialization Chambers. Regarding the "skill in common" part, I observe it with "grouped profession in common": "grey" metalsmiths, "yellow" woodworkers, and so on pair off more often. The strength of acquaintances list you mention is handy to track and probably is related to the "chat counter," and the rapidity of progression from Passing Acquaintance-->Long Term Acquaintance-->Friend-->Lover appears to relate to sociability, probably to profession, and maybe to likes. Except for the fact that antisocial dwarves usually take longer, I can only say empirically that likes/personality may matter a bit especially when opposed, but profession preference shows a distinct trend. I am, however, amused to think that I've been feeling that trend to be pretty dispositive, but that Toady may have coded it to specific skills and I've been making judgments based on coincidence.

Subsequent genealogical management has reaffirmed (for me) all the observations I reported in http://www.bay12forums.com/smf/index.php?topic=117274.0 (An experiment in dwarf romance). Here are some of my other observations since then:

Spoiler (click to show/hide)

Some related observations to genealogy management are in "Heredity Science" http://www.bay12forums.com/smf/index.php?topic=111624.0,  "Dwarven Eugenics" http://www.bay12forums.com/smf/index.php?topic=125188.msg4181925#msg4181925, and  "Fortress-born dwarves & animals are not growing to adult size!" http://www.bay12forums.com/smf/index.php?topic=126558.0. I'd love to hear others' insights and experiences on this subject.
Logged

Swonnrr

  • Bay Watcher
    • View Profile
Re: Some science about dwarven relationships
« Reply #6 on: June 09, 2013, 06:08:12 pm »

[...] The compatibility calculation got a large bonus if they had a skill in common (Novice or better?) [...]

I confirm this, and it may be the first parameter for couple forming.
My miners ALWAYS marry each other, very fast, in a few month they encounter each other, sometimes weeks.

My theory is that they can chat when working, and i'm almost sure it's influenced by LOS.
So that would mean that dwarves working close to each other (example: miners, always in the same room mining, eating, sleeping or idling) get a huge bonus from skill factorised by the exposure they have to each other.
My carpenter/mason are also few and works as slave, but their workshops have walls and doors, so they basically see no-one, and almost never marry.
Logged

Saraias

  • Bay Watcher
    • View Profile
Re: Some science about dwarven relationships
« Reply #7 on: June 09, 2013, 06:17:02 pm »

I'm playing with fortress without mountainhome (=no migrants),

I've been pretty enthused generally about the ongoing history, births/deaths, fortress retirement and related features in the upcoming release. Building a civilization and its distinguished lineages will just get better with that, I think. But now I'm seriously considering the merits of taking the exact start you describe and making an attempt at fully rebuilding a dead civilization from the ashes.

I wonder how certain things will work in this case. Will new forts decimate the population of retired forts as they draw historical migrants? Will a founder of the first fort spontaneously rise to the throne? Hmm. The family lines will certainly have potential to get tangled. :) Sounds fun.
Logged

VerdantSF

  • Bay Watcher
    • View Profile
Re: Some science about dwarven relationships
« Reply #8 on: June 18, 2013, 12:20:13 pm »

Great thread.  Thanks for the ideas.  I have two dwarves that have been stuck as lovers for ages.  Time for that marriage burrow!

evictedSaint

  • Bay Watcher
  • if (ANNOYED_W_FANS==true) { KILL_CHAR(rand()); }
    • View Profile
Re: Some science about dwarven relationships
« Reply #9 on: June 18, 2013, 03:54:50 pm »

For lineage, do physical traits pass down to children?  If both parents have ebony hair, will the child have ebony hair, etc?

Also; are cousins the absolute closest two dwarves can be related before marrying?

Saraias

  • Bay Watcher
    • View Profile
Re: Some science about dwarven relationships
« Reply #10 on: June 18, 2013, 04:43:43 pm »

For lineage, do physical traits pass down to children?  If both parents have ebony hair, will the child have ebony hair, etc?

Also; are cousins the absolute closest two dwarves can be related before marrying?

Second question first: Cousins definitely marry. Siblings definitely not. I've never had aunt/nephew or uncle/niece pairs occur in fort mode, and the family trees I've looked at in legends mode haven't had any that I recall. I wouldn't gainsay Broken's reply in your other thread, nor any other more expert persons than I, that say they do; and I've never tried to force such a union.

I've seen some interesting ties between cadet relations, highlighting interesting characteristics of a race with fertile 170-year lifespans. My favorite example is a fellow who married his great-great-aunt, who was 5 years younger than him.

First question: For the most part, yes, skin/hair/eye color do seem to pass on. Hair color information seems to get lost sometimes if one parent has "his/her hair is clean shaven." The Heredity Science (and to some extent Dwarven Eugenics) threads I linked up-thread have some interesting things on that, especially the first one. Slink's list of children with divergent eye colors from parents in Heredity Science intrigued me, and I've been looking for patterns of randomization/blending/recessives in my population without enough examples of deviance to report anything. (I rely on in-game text descriptors for my observations, defer to anyone with knowledge of the underlying numbers.)
Logged

evictedSaint

  • Bay Watcher
  • if (ANNOYED_W_FANS==true) { KILL_CHAR(rand()); }
    • View Profile
Re: Some science about dwarven relationships
« Reply #11 on: June 19, 2013, 12:52:33 am »

Thank you. The reason I was asking was because I want to make a fort of 20-30 dwarves locked entirely underground, away from migrants and invaders.  The raws were modified so that they produced multiple children per birth, but only had a lifespan of 10-12 years.

I was curious whether leaving these dwarves alone for a century or two would result in a race of near-identical homogenized dwarves, with their memory of the surface dead for dozens of generations, their entire world confined to their tiny vault, pre-cut into the bedrock.

Also, I was afraid that after a few generations no one would marry again, because they're too closely related to one another.  This has pretty much cleared up these worries, though.