Advertisement

Dunbars number for game characters

Started by February 26, 2016 10:31 AM
20 comments, last by Norman Barrows 8 years, 10 months ago

You could use the number writers use for TV shows and novels. Nine characters is easy enough for viewers and readers to feel engaged with.


That's a great idea Tom. They might also have numbers for how many recurring side characters people remember.


I suppose so. Look at how many recurring side characters there are in the Simpsons. Not so many in South Park. Big Bang Theory has 7 main characters, and I counted 13 or fewer recurring side characters.

-- Tom Sloper -- sloperama.com

Anyway, I find the "could die" problematic. I mean, if you really care about a character, and you have an emotional attachment, you should just reload the game if it dies. I mean, if you had the power to do so in real life and you cat died, wouldn't you reload the last save to prevent your beloved cat's death? If you include perma death the player most likely won't care about these characters since it would hurt too much if they die (and they will), so the player will just treat them as replaceable assets OR engange in save scumming.

I find the opposite. In any game with save scumming, I don't get attached to characters as I know they're never in danger. In games with permadeath, every emotion is amplified as danger is actually real and actions have non-reversible consequences. I usually avoid save scumming even when it's possible as it makes for a more interesting narrative and allows for more emotion in the game.
Even in adversarial perma-death this applies -- e.g. in the DayZ mod (before it devolved to deathmatch sniping survivors from hills for no reason), I felt real guilt for the first time ever when killing another player in an FPS, because I knew they might have spent a dozen hours on that now-dead character...

No offense, but, it's kinda strange if fear of losing a character to death is the only reason you feel attached to characters. Don't you notice if some characters are particularly attractive or funny or empathetic in non-deadly situations? Guilt is a strong emotion but I definitely wouldn't want to feel it while gaming, or make my players feel it unless they took a really obviously bad choice.

I want to help design a "sandpark" MMO. Optional interactive story with quests and deeply characterized NPCs, plus sandbox elements like player-craftable housing and lots of other crafting. If you are starting a design of this type, please PM me. I also love pet-breeding games.

Advertisement

I've heard you can only sustain about dozen "deep" relationships at a time. Reading the Dunbar number article (which I've never heard of until now), and this article, it seems the answer isn't a strait "150", but that the 150 is loose relationships,

"Our theoretical circle of 150 is not a homogenous social group, Dunbar explains, but rather consists of four layers, or "Circles of Acquaintanceship," which scale relative to each other by a factor of 3—an inner core of 5 intimates, and then successive layers of 15, 50 and 150. With each successive circle, the number of people included increases but the emotional intimacy decreases."

So that may be worth noting. 15 primary characters would probably be a good guideline - including villains, ofcourse.

However, this may not be a one-to-one relationship with game design. If I already have a deep relationship with 15 people in real life (in my case, more like four people), does that mean I can after 15 real-world relationships and also 15 fictional relationships? Or are fictional relationships so shallow that at best, they are part of the 150 instead of the 15?

Interesting to think about - thanks for raising the subject.

I find the opposite. In any game with save scumming, I don't get attached to characters as I know they're never in danger. In games with permadeath, every emotion is amplified as danger is actually real and actions have non-reversible consequences. I usually avoid save scumming even when it's possible as it makes for a more interesting narrative and allows for more emotion in the game.

Even in adversarial perma-death this applies -- e.g. in the DayZ mod (before it devolved to deathmatch sniping survivors from hills for no reason), I felt real guilt for the first time ever when killing another player in an FPS, because I knew they might have spent a dozen hours on that now-dead character...

No offense, but, it's kinda strange if fear of losing a character to death is the only reason you feel attached to characters. Don't you notice if some characters are particularly attractive or funny or empathetic in non-deadly situations? Guilt is a strong emotion but I definitely wouldn't want to feel it while gaming, or make my players feel it unless they took a really obviously bad choice.

In DayZ, Hodgman killed a real-life person's account, perma-deathing their progress in the game. That was the "guilt".

In some games, like Fire Emblem, I fear losing them (and reset the level), because I'm attached to them because of how useful they are to me. If one character constantly carries the team, it does feel like an attachment forms, because I constantly have that character "save" my weaker characters.

Perhaps it's poor writing, but most attempts in games, movies, TV shows, and animes, where creators try to have "funny" characters just feels hit and miss, or else entirely miss.

Humor works alot better for me when it comes from situations and dialog, rather than making one character "the funny guy". I felt Sokka in Avatar: The Last Airbender was ruined as a character for me, because they took a cool warrior guy with interesting romantic relationships and make him into a bumbling idiot who can't fight, and is a joke.

But an interesting situation and dialog that was funny involving Sokka, was when he's talking to Zuko and says, "My girlfriend turned into a moon spirit" (which was funny because it was a true observation of something that actually happened), and then Zuko looks at him and in a totally deadpan voice says, "That's rough dude."

What's really interesting is when you take two already-established characters who haven't had much interaction with each other thus far, and put them together in an odd situation, and let their different eccentricities interact. Baccano! was full of that. One such amusing situation in Baccano! was when the mass-murdering sociopath, Vino Stanfield, was literally covered head-to-toe in blood and proposes spur-of-the-moment on top of the train to the mute assassin (Chane Laforet) who he just met and, as part of the marriage proposal, offers to help her protect her imprisoned father who is the head of a cult (of whose members' blood is what he happens to be covered with).

(Incase people haven't seen it, Baccano! is a prohibition-era anime that is a brilliantly executed mix of drama, romance, comedy, and action. I strongly recommend it to anyone, whether you're a anime fan or not. There's not a single scene or character that I disliked - it is really masterfully crafted)

Speaking of Baccano!, it has at least 20 main characters, each unique and memorable with their own backplots and personalities and, while it was sometimes difficult to remember the side characters and who was in which mafia family, the characters themselves were easy to get invested in and stay connected to, despite how many there were.

psychology has the golden rule of "seven plus or minus two" as the number of things the average person can keep track of simultaneously.

The SIMs has always had a limit of 8. there may be design as well as technical reasons for that. even with robots its still 8, plus 8 pets i believe.

for many games where the player controls multiple characters, the limit is six for some unknown reason.

Caveman 1.0 had a limit of 10.

In Caveman 3.0 i've been experimenting with large bands. the limit is now up to 50, and i've run bands as large as 35 or so for an entire game year.

needless to say, the number will depend on game type. larger numbers for things like sports games, and smaller numbers for things like RPGs.

not caring because you have so many you cant keep track, and not caring because they are disposable are two very different things.

the important question is how many is too many to care about. if they are disposable, you can never have too many - cause they're cannon fodder. and you can never have too much cannon fodder.

the lower limit of the golden rule and the inner core from Dunbar coincide niecly at five.

add one disposable flunkey slot, and you get the typical six.

add three minor characters to your five major characters and you get a SIMS household of eight.

in Caveman (an open world sandbox stone age FPSRPG) characters' stats and skills define their usefulness to the band, and thus affect how attached you become to them. having played them for a long time also increases attachment above and beyond the attachment due to skills learned over that time. characters who have poor stats, few skills, and/or have not been band members for long are more disposable. in fact, many are recruited into the band as cannon fodder to guard the inner circle of long term band members i do care about.

i have a long term playtest game. i'll play the primary character for about a game year, then export to a new game and play for another game year. i'm now on my third year, and third game, and third band. the first and third bands were about a dozen each. the second one was the 35 member band.

one of the big things i noticed was that the more characters you control, the longer it takes to play, because more time is spent controlling the additional characters.

that was one motivation for the export to the third game, and starting a new band - getting something smaller than 35 members. i mean, imagine trying to keep a household of 35 SIMs busy all the time - its a lot of work! <g>.

in the third game, i'm down to 12 members (limit is still 50), but am considering cutting that number down to a smaller group i can really care about - probably more like 4 to 6 band members. you don't get as many warm fuzzies when a character you don't care as much about accomplishes something, so might as well spend your time on the characters you do care about whose accomplishments thrill you more.

Norm Barrows

Rockland Software Productions

"Building PC games since 1989"

rocklandsoftware.net

PLAY CAVEMAN NOW!

http://rocklandsoftware.net/beta.php

You could use the number writers use for TV shows and novels. Nine characters is easy enough for viewers and readers to feel engaged with.


That's a great idea Tom. They might also have numbers for how many recurring side characters people remember.

I feel like that fits for XCOM 2 also. If your main characters that you really care about get injured, you'll have to go on some missions without them, using characters that you're not quite as familiar with. You get more customization options as the characters survive for longer, so by the end of the game I had a few that I liked a lot and took on missions whenever I could, and others that were like recurring characters that weren't always around. The lower ranking soldiers didn't mean much at all until they survived a few missions and started to get the abilities and customization options to make them feel more unique.

Radiant Verge is a Turn-Based Tactical RPG where your movement determines which abilities you can use.

Advertisement
> How many characters do you think most players will "care" about? Where "caring" means not just being disappointed by the failure and mechanical consequences, but actually feel sad about losing the character?

In games I think, that it is less a question of numbers, but a question of uniqueness. Games, like XCom (I only play the first part), have often very generic characters.

If you consider Game of Thrones, many important characters are easily tagged by some unique features or names, even if you don't remember their name. Most people knowing the series will immediatly know of whom I speak:
- the bastard lord
- the winterfell princess
- the sadistic bastard
- the queen
- the female knight
- the dragon queen
- the spider
- little finger
- the sadistic king
- the fire priestess
- the mountain
- the assassine girl

So I would suggeust to ignore more the numbers and to concentrate on unqiue features instead.

No offense, but, it's kinda strange if fear of losing a character to death is the only reason you feel attached to characters. Don't you notice if some characters are particularly attractive or funny or empathetic in non-deadly situations? Guilt is a strong emotion but I definitely wouldn't want to feel it while gaming, or make my players feel it unless they took a really obviously bad choice.

No offense makes any normal bit of discourse suddenly sound offensive! laugh.pngtongue.png
I think you misunderstood the context a bit. Seeing the OP mentioned XCOM, I figured that the posters above me were talking more about non-narrative based games, where characters are really just a type of strategic resource. In these games, their personality is really down to how well they perform in combat, which is basically down to successes or failures on dice rolls, and your own decisions about how/when they move, and whether you bother to remember these details and attach these useless flavour memories to each unit...
Usually in a narrative based game, where you actually can value characters for their dialogue and interactions, a character's (permanent) death is a scripted part of the storyline that can't be avoided by simply reloading right before they die and having them dodge to the left.

I was saying that in these kinds of tacitcal games, the possibility of perma-death is (for me) what changes characters from simply being a strategic asset - soldier, tank, sniper, helicopter - and into a character:
- Frank the soldier who has never missed with his rocket, makes me yell when he finally does.
- Johhny the tank who was suprises me by surviving for 5 missions, cements his heroism by eventually going out in a blaze of glory.
- Jane the amazing sniper who was stabbed in the back, causes me to make tactical mistakes in my quest for vengeance...
Seeing these tactical games are war stories, a war story almost requires certain character arcs to end in death. You can't be a hero if you're never in danger, which is why Superman needs kryptonite as a plot device. Or why most action flims are predictable pulp, as the cliched drama-through-combat means there is no drama as we know exactly how it's going to play out. Or, A song of Ice an Fire would probably be a pretty terrible series if no one ever died -- great character development goes supernova upon their death or harm, exaggerating and cementing the feelings they've given you up until that point.

But yes -- writing custom dialogue, interactions and attitudes for each unit would certainly make them more memorable and make people care about them.
One of my favorite tactical games is Jagged Alliance 2, where there's a large number of mercenaries you can hire (and one that you can customize to be your own avatar), but each one actually did have a different personality, with lines of dialogue and even different interactions and background interpersonal relationships. This is still a tactical combat game, which is very light on narrative... but had just enough to make it feel like an RPG.

e.g. If you neglected to read their background files before hiring your team, you could accidentally hire two jilted ex-lovers who go on to absolutely destroy the morale of your whole squad laugh.png Likewise you could hire a team who've previously worked together and see them gel.

Other times you would recruit an inexperienced local who you cross paths with, but end up valuing them more than your professional hired killers because you've spent so much time watching them as a student, progressing in their (still mediocre, but much improved) abilities.

A character's manic high might cause them to ignore your orders to only placed well-aimed single shots, and instead go bezerk by stepping out of cover and firing a full magazine in a momentary immortality complex... Other times they may be so depressed at the recent loss of a close team-mate that they refuse to engage in battle properly, making you choose between booze as a short term band-aid for their inability to function, and the long-term consequences of having a drunk on the team.

As for the original question, I think JA2 let you have squads of 6. I found it easy to care about every member of squad #1, but any extra squads kinda became the B-team of faceless helpers.

So, going with the idea that we're talking about mechanics-based character development, which is an adjustment for some: I can vaguely remember getting hung up over the death of a valiant unit in Starcraft back in the day, but it hasn't happened in RTS' I've played since for me. Going with that idea, then the number of characters you care about is limited by the available roles in your head. Going with the examples from television and biology, that number seems to be north of 7 and south of 20. I think where you fall in that range depends on the mechanics and how they're implemented. If there's 12 cool things to do on the battlefield, you generally have 12 characters. You can push it a bit by diversifying the feats within the role, such as 'this sniper was the one that saved the president at 1000 paces, and this sniper was the one who sniped their sniper who was mowing us down' but the more redundancy you have in mechanics the easier it is for someone to feel replaceable both in your mental map and in actual experience of the game.

But then the OP is asking about maximum attachment and for that I think you're talking about that smaller number, somewhere between 4 and 7. That group of people can each have multiple feats and a more dynamic wide ranging role. 'This is the XO who sniped that pilot who was mowing our people down, and if he wasn't there holding the doors open back in Heidenburg, I don't know where we'd be...' So if you're going for maximum attachment, then you're talking about fewer characters. There's a tradeoff there, and it's not clear where the desired balance is, because it'd be different for everyone.

But this topic is incredibly enlightening, I'd never really considered the consequences of large parties in a narrative-driven game.

Jagged Alliance sounds awesome. I love XCOM for the war stories, but I wish soldiers had more personality, character development and relationships too.
Maybe I should share what made me bring up this topic:
I was thinking about how to design a game where you lead a little community of people hoping to survive in a hostile world. Think Fallout Shelter, but with focus on character development, relationships and war stories. XCOM is obviously an inspiration and I wanted to understand the "XCOM-factor". Note that I'm more interested in a generic discussion about the topic than specific advice for myself.
I do also want to clarify that I'm not asking about how many characters the player controls at the same time on a "mission". Some good advice on that in the thread already. 4 - 6 seems to be normal and a lower number might help foster attachments, but it depends so much the type of game.
I guess "maximum attachment" is a little vague. Maybe I should say maximum overall attachment. My goal is to find a sweet spot between having many characters weak attachment and strong attachments to only a few. Also the right balance between high and low attachment to different circles of characters. (5 favourites/20 supporters/125 faceless vs 10 favourites/140 faceless) No hard rules for that, I know.
Say you want to simulate a community of 150 people. How would you find the balance between giving each character depth and narrowing the players focus down to a few interesting characters that matter?
In Banner saga I had around 6 favourites and a group of heroes and supporting characters(20 - 30?). But there were also hundreds of faceless clansmen and fighters only represented by a number. Occasionally some of them will show up once or more in "random" events and even turn into a hero. It's a bit like generated pedestrians in an open world game where people disappear when they walk out of view.
Other games show every single character and give them a name and a unique appearance in addition to their role, but they are still just numbers to the player unless they become interesting or useful. Do you think this approach can distract players rather than allow for more potential attachments? In Fallout Shelter I think you can take any character and put them to any task from cooking to exploring the wastes and fighting raiders and for the really short period I tried the game I felt that it made it harder to "get to know" any of the dwellers. Of course it's not a game where you spend hours commanding your characters through tense and deadly situations so there would be different expectations.
I'm not quite sure which approach I would choose, but if the latter is chosen my gut feeling is that it is a good idea to separate the filler characters from the ones you're "supposed" to bond with. I prefer abstractions and streamlined, tight design over bloated simulation, but I like the idea that everyone has a name and a face.

This topic is closed to new replies.

Advertisement