Punished by Rewards

[charmian]

Sabina and I were talking about points systems on websites, such as Blip.fm's blips, and whether this created negative social effects. She argued that the existence of such systems prompted users to play it as a game/play purely for points, and also explained to me the secret of getting props and being uber-popular on Blip. I was excited because I thought she was going to reveal to me which bands are the hippest or something, but actually the secret is to reblip the popular and prolific people, and blip a lot yourself, and follow the popular folk. However, what really prompts people to play blip.fm, or another service, as a game?

Is it the appearance of metrics? Yet, LJ also has metrics: comment count, and # of people friending you. I certainly use comments as a metric and believe getting them does measure my 'success' in some manner. Probably blips were created to be a game, to get people to use the site more, rather than as a metric of discerning who is the best blipper. Also, while blips are infinite (because they aren't based on putting $$$ into the system), they are also finite (you don't have unlimited blips to give others directly) and transferrable.

This began to remind me of Alfie Kohn's Punished by Rewards, which argued that rewarding students causes them to devalue learning itself and simply work for the reward. Extrinsic motivators diminish intrinsic motivators. I've read the book: the research cited is IIRC fairly convincing, although I'm not sure the theories can be implemented in a large scale in schools.

Yet, I don't think the system I was discussing in the last entry ([this is good]) really falls into the category, because it works more like comments, like a very short, pre-made kind of comment, or like a poll result.

Anyway, in other news, I see that DW is planning cross site authenticated RSS reading. The bug reports says: " you won't need to create a feed account for every account you want to follow. You'll
provide us with your authentication information for the LJ-based site you want
to read the friends page of, and we will take what the protocol returns,
perform magic, and intercut it on your DW reading page." I wonder how this would work... Well, maybe I wouldn't be able to understand it anyhow.

Comments

[zvi]

They're going to use an undocumented API instead of cross-site RSS, so, in order to understand it, you'd have to understand LJ's crazy code.

[charmian]

Thanks for the info!

[verstehen.livejournal.com]

I think the exchange of online communities is fascinating, particularly fannish ones. I'm actually writing a paper this semester on symbolic exchange in flashfic communities.

I really think metrics are the way popularity is ranked online. The hard part is figuring out the metric rankings; are comments considered stronger than website hits? Do website hints trump links? And so forth. We can't use objective measures of value (like dollars) for most online "transactions" so I think the metrics are a useful substitute.

[charmian]

Cool, what is your theory on that?

Yeah.. .Well, I think the SEO people have developed some theories in that direction. XD Do you also study SEO and the google analytics stuff?

[verstehen.livejournal.com]

Well, I looked specifically at sga_flashfic, so I was analyzing the different forms of symbolic exchange for fic through comments and if the narrative conventions within the community changed over time in relation to those measures. I looked at total number of comments by story type/narrative, type of comment (positive/negative), and length of response (range between one-word and comprehensive critique). I basically found that slash (McKay/Sheppard) received the most positive feedback over all measures (and was also the most number of fics produced. Essentially there wasn't really a change over time in narrative types, but the number of M/S fics increased steadily over time. That means either a) the community gained a reputation for being a M/S community or b) people noticed that M/S stories were getting more (positive) response and started writing M/S stories too which may have led to c) a feedback loop between the two. I have no real way to make a determination, without interviews on motivations of writers, but it was an interesting study anyway. :)

[charmian]

That's interesting. What do you mean by 'symbolic exchange'? Is that like 'social capital'? So essentially you mean how people wrote their fanfic changed in response to user feedback?

I'm kind of confused though... so on this comm the slash pairing McKay/Sheppard is the only slash pairing allowed to be posted? Or the only one in which these trends were observed?

Hmm, maybe you should write a post on this and submit it to metafandom. I'm sure they'd be interested in it.

[verstehen.livejournal.com]

Well, basically if an exchange is objective goods/service for another objective goods/service/money, symbolic exchange is basically when the objects being exchanged have no objective value beyond the value placed on the people exchanging them (fics/comments can't have monetary value, for example). And yep, there is a correlation between the pairing and story type and the amount of positive response that pairing/story type received prior to you posting yours.

The community was open to any and all pairings but M/S became so ubiquitous (likely through reputation and response levels) that on one challenge (the "Harlequin" challenge) about four or five months into the community that people asked on the challenge post if pairings other than M/S were allowed. The mods quickly assured people that, yes, other pairings were welcomed and encouraged, but you can tell how much that particular pairing had basically made the community by that point.

I just might! The paper is due in about two weeks, so I could summarize my findings after I'm done. Until then, my focus is writing the paper. ;)

[charmian]

Ah, I see.

Well, I'm not entirely surprised that story type is influenced by what stories have been previously praised. XD What is this effect called in sociology? Is this kind of like the long-tail effect? (er, the opposite part, the forming of the large head)

Wow... indeed, that is pretty drastic. That makes me feel sorry for the other shippers though.

Good luck on the paper.

[verstehen.livejournal.com]

Biased by past experience, I was completely unsurprised by my results too. XD There isn't really any name for this particular phenomenon, that I know of (that doesn't mean it doesn't exist!). Maybe the Pareto principle is the closest I can think of.

[charmian]

I wonder what the mechanism is... Is it that people switch over to the preferred pairing? Those who ship unpopular pairings stop posting? Or simply that more people who like that pairing join?

[verstehen.livejournal.com]

My conclusion is that it's a little bit of four things: a) people switch b) those who ship unpopular pairings keep community membership but stop posting c) people who ship unpopular pairings leave the community and go elsewhere and d) once the reputation is established, people join specifically to write/read that pairing. All of which runs on a sliding scale, because it's not like people ever stay with just one thing, particularly with the ease of joining/leaving on LJ. Or, at least, that's basically what my conclusion says. :D

[charmian]

I guess when things reach that point, what might be better is to create a separte community for non M/S (or whatever the pairing is) pairings.

I wonder how this might pan out in terms of other metrics (views, etc), or whether feedback itself is also affected by other feedback (its prescence)

[verstehen.livejournal.com]

I wish I'd been able to cross-reference things like page views, against comment measures but I no real reliable way of telling those for a LJ community, which is a pity, because I would have loved to be able to do a trackback type link hopping too. See who linked to what stories, how they were described, who linked to the community and how the community was described. Looking at the context in which the community is placed within the fandom would definitely effect the results. Well, that's a study for another semester. ;)

[charmian]

Yeah... there are no page view metrics on LJ unless you install them. Nor is there a reliable way to find who linked to the stories etc.

[sub-divided.livejournal.com]

I think that where you say "blips" in the second paragraph you really mean "props".

Another way people game the system is by giving props to each other. You only need two people (http://blip.fm/profile/sub_divided/blip/3834699) to do it - though there might be a limit? Dunno, never tried.
Someone claimed today, in a blip, that Blip.fm is going to reset all the props soon. Can't remember who or where though and of course it's impossible to search. OTOH you can see prop obsession first-hand here:

http://search.twitter.com/search?q=props+blip.fm

So the etiquette is to say thanks for props? (Oops.) Lots of people have an automatic prop back policy (http://blip.fm/profile/evablue/blip/7169752), as well, which makes other people more inclined to prop them.

You don't get medals for props, though - followers is more important. People try to game that too, by giving @shoutouts to random users from the blip stream, hoping to be followed:

http://blip.fm/profile/yamamiya

Some genuinely great blippers mixed up with all the gamers.

[charmian]

Yeah, props, wuffies, whatever.

What would even be the purpose of resetting all the 'symbolic exchanges'? I don't get it. So people are playing for the medals as well? This is pretty hilarious. XD

[sub_divided]

I don't know, since they don't really measure what they're supposed to anyway (that is how much people like the music you're posting). Maybe blip.fm is getting rid of reciprocal propping, and wants to reset everything. They already got rid of the feature that let you see what other people said about the songs you blip, maybe because they realized that like 90% of messages were "good morning! ;)" or the song title.

[troisroyaumes]

Hey, pretty cool (I've heard of similar behavior on Xanga, which also uses a props system). Now I'm really hoping you'll get around to making that post on blip.fm; I don't use the site often enough to understand its social dynamics.

[sub_divided]

Thank you for the invite! And thank you charmian, too~~~ XD When it rains, it pours.

(To Charmian: I hope that wasn't your only one. ^^; Or that there is some way for me to give it back. Can I pay it forward?)

<3 <3 <3 <3

Twitter and blip.fm social dynamics are both hugely weird to me, since the system puts so many constraints on what you can do - but that only seems to make people more inclined to try. ^^ At the same time it's much easier to connect than on LJ, where you are forced to take chances on people you don't know well (yet). In a way it's almost more like RL than like the blogsphere.

[troisroyaumes]

Yay, you're here! <3

Hm, the point about it resembling RL sort of reminds me of what Scalzi said about Twitter, which a friend linked me to a while ago.

[sub_divided]

John Scalzi is always exactly just ahead of the curve on these things. XD Following on twitter is like dropping into the middle of a conversation with someone you already know well - which is why celebrities/bloggers do well, everyone already "knows" and cares about them. And it IS like texting - or like changing your AIM status to something clever, actually, and hoping that someone will ask you about it.

And it's easy. LJ is less casual. Before you can friend someone on LJ, you have to go through the same awkward first steps you would if you were meeting someone in real life - hang out at parties (communities) and have conversations about mutual interests, or go over to a friend's (journal) and talk about your mutual acquaintances. The relationship may ultimately be deeper, but it's not as effortless.

[charmian]

Hmm... Well, if you have someone that you really want to give it to, go ahead and send it. If not, I have someone in mind that you can send it to.

[sub_divided]

It's your invite, so I gave it back to you. (Assuming that the code will work for anyone, not just the person with the email address you gave DW.)

[charmian]

Okay, I'll give it to that person then.