Polls, General Requests, and Social Activism
hey, hoodie here. if you’re paying attention to the things i post on medium, you’ve probably seen my series of articles on social activism and social media, and how they interact. here’s the previous article just in case you haven’t. https://medium.com/@novemberninerniner/timed-blocks-user-preservation-and-public-norms-20e439048481
today i want to talk about a couple features you are already aware of, at least in theory, and how i think they are important for the future of social media.
polls are pretty simple, but they do have nuance, nuance that is often overlooked in favor of the first-glance assessment
you see, polls are not only a fun thing to do with your friends, but they are used time and time again to influence social change, be it through democratic elections, gauging social opinion on such elections whilst in the process of the election, or even just asking what i should eat for dinner tonight, coz dangit, i really can’t decide
so let’s look at the most basic bit of nuance to address, authorization and ‘fraudulent votes’. this is a big deal for elections, but at least in the most vague sense, it doesn't have as much room for worry here in our little sphere of social media.
someone who takes the time vote fraudulently probably really cares about the topic of the vote, afterall.
there’s a few ways to avoid fraudulent votes, one of which i’m sure you’ve already thought of.
we can authorize votes with an account that voted, leveraging the user paradigm already existent within federated social media, and thus allow one user one vote, but then comes the topic of alternate accounts. i personally have at least five on the fediverse. i can easily vote with each individual account, can’t i?
so that method is out, and that’s sad, because that means we have to get creative. i have a few solutions, and each is it’s own level of ‘security’, if we choose to think of it in that manner. it’s worth noting that no solution is wholly secure, in the context of the fediverse as it currently exists
i could make a whole article on that topic, though.
so let’s say we are a user on the fediverse, and we have friends. bit of a vague concept, but it’s one we’re all familiar with, so i wont spend too much time on it. i have friends A, B and C. i, more likely than not, am following at least 2 out of 3 of these friends. now we have a functional paradigm, though. ‘Followers’ of the account in question, that posts the poll, are more likely to be unique users, by a very large factor, i would estimate.
similarly, ‘Mutual Followers’ are almost guaranteed to be unique pairs of users. as such, we can make three different levels of ‘security’. i want to reiterate here that none is perfect, but each are more secure compared to the last.
- Account Verification — account must be unique to place a vote in the poll
- Follower Verification — account must be following poll poster to vote in the poll
- Mutual Follower Verification — account must follow the poster of the poll, and be followed by the poster of the poll, to place a vote in the poll
now if we look at this from the stance of a traditional poll, we can find the one downside with this approach, and it is worth noting, because it changes the way we look at this data, very meaningfully.
the votes in the poll are no longer anonymous.
it’s a bit of a lose-lose scenario, when you think about it. if your votes are to be anonymous, they will never be unbiased in regards to the level of effort the voters are willing to go to to change the outcome of the vote
if your votes are no longer anonymous, your users will never be truly secure in knowing their votes will not be privately monitored to see what you voted for.
i, like most people, have a bias, so i want to clarify that before i speak on my personal opinion of this problem. i do not personally care all that much for my anonymity, which many people are likely to disagree with
that said, i can see the value in anonymity. i understand why you, proverbial reader, do value your anonymity.
does the anonymity of poll-goers truly matter in a social media landscape?
that’s the real question at hand here, isn’t it?
does voting fraud matter? it depends on the user and the circumstance, so we should give them tools to make that choice
you might disagree with the above statement, so i want to clarify why i believe that.
polls being asked in a space are directly influenced by the person making the poll, the purposes that person is using the poll for, and how likely that person is to abuse the non-anonymous factor in question
as much as we may want to think otherwise, we will never know other people as well as we know ourselves. i do not know you, you do not know me, not like you might want to think you do
so we do not actually know if the people in question will try to take advantage of non-anonymity to leverage that information in some way
we cannot know
alright, we’ve spent a lot of time on polls. but i have two other topics to address, don’t i? it’s a bit odd to spend this much time on polls, really, but they are the most nuanced topic of the three, so it’s unlikely i can work it into a different division of time
let’s outline the final bits of polls real quick, before continuing
there are four types of polls. ‘state’ of the poll, if you will
- ongoing
- timed
- closed
- open to review at all points in time
each of these responds to actions differently. you cannot vote in a poll unless it is ongoing, or timed, and you cannot view the results (in most poll paradigms) unless it is already closed for voting, hence the final option of the four
i have some other things to say on this topic, such as a few examples of how twitter’s polls aren’t really well handled, but that can go in the feature specification for the polls
instead, i will now point out that we have all the rules in our hands for making a feature complete poll implementation
- state of the poll (four possible states)
- security level of the poll (three possible levels)
- vote data already submitted to the poll
- whether votes can be changed after first submission (true or false)
as such, i will only state that timed polls are a good innovation on twitters part, at least in theory, and that when a poll is closed, no matter whether it is closed automatically, based on a timer, or by the action of the poster of the poll, the results of the poll should be returned the the poster of the poll, probably as a notification, since that is generally where users in the social media paradigm ‘receive’ information
General Requests
alright, polls, knocked it out of the park. home run. you could probably use that article as a feature specification, not to toot my own horn
what is a general request? well, looking at it based on definition, we can presume it is a yes or no question, presumably asked of the user in the social media paradigm
what does the user need asked of them? well, ‘can i follow you’ is a pretty recurring question, but there are many, many more we can point out, as possibilities.
- can i put your handle in a follow friday post?
- can i put your handle in my bio, since we’re dating?
- can i direct message you?
- can i add you to a list?
- can i add you to a conversation through adding your handle to my post?
wow, that’s a lot of questions, and that isn't even all of them
hell, i could create questions for this format more than i can answer them. if any of my followers was to ask me this question, right now, i would probably have to think for a few seconds and maybe even get some relevant data like how many followers they have before i could come up with an answer.
this is where my feature comes in.
let’s expand the follow request feature and allow for requests of all types, so that users can ask these questions on their own, and answer them, on their own, without having to make it explicit in a probably-awkwardly-worded conversation.
the functionality involved is rather simple, we need to identify such use-cases, where a general request is necessary, and send one to the user involved.
a #ff post is simply a post with a mention, and we can treat ‘may i mention you’ as an opt-in require-request feature
settings menu can contain a list of things that may or may not need general request authorization, such as boosting posts, mentioning the user, direct messaging the user, and more
i will not elaborate on the feature of a yes or no request the user can approve or deny, since it already exists
i’ve just about said my piece, and i only want to make sure something is clear
this stuff is really, really important
in fact, the polling feature directly affects how users interact with development.
for instance, a developer can struggle to choose what to focus their time on, because things can have blurry amounts of importance
good polling features also encourages activism, and socially interactive development
these are all good things.
credit to Erin for better wording on the section related to voting fraud
thank you for listening,
- hoodie aida kitten.
here are some links to support me if you wanna send me a few bucks or support my work. also now this article is over on dreamwidth where i will be staying and putting my future work. dreamwidth supports RSS and atom, which medium doesn’t.