Cool project! The results of all the archived votes made sense to me, but I was most surprised by this one:
> The U.S. was right to withdraw from the World Health Organization (WHO)
> 62% Agree - 38% Disagree
I didn't know that the WHO had such a negative reputation. We are quite fond of such international institutions in the EU at least (ranging from a force for good to fairly harmless). What's the context? The rest of the votes seem quite liberal leaning otherwise.
One of the reasons is that the US pays for a whole lot of those international institutions, creating policy and governance issues that end up beyond local accountability. The "force for good, or fairly harmless" rubric changes when it's your money. Then it becomes "why are they spending my money on that bullshit when we have fires to put out at home?"
Covid era politicization and the fallout from that has a lot to do with it as well.
Thank you that makes sense. I did a bit more research for context:
The US only left in 2020 and then rejoined in 2021, I suppose that’s why I didn’t remember it as a big thing.
The US was also just paying ~15%. It was the biggest governmental funder, with Germany at ~9% as the second. But the WHO is apparently mostly funded by charity donations, the Bill & Melinda Gates Foundation was paying ~5% for instance.
(it’s awkward to list sources on the phone but should be easy to verify)
I do get the sentiment though from the perspective of the US, I don’t mean to argue your points.
Most of my voters have come from Reddit and Bluesky so far, which is primarily where the left leaning is coming from. My X account was unfortunately suspended lol. I put an appeal in, but was originally flagged I think do to a new account, political content, and lots of links to the polls. I use an OG dynamic card generator so if I post a link to that poll or result, it creates a card for it on the fly. I think X didn't like that since I wasn't established.
That one was interesting, I am not really sure why that one skewed so far the other directly. I did not have the discussion section open yet (and just slowly getting a few active users), but that was the original reason I added the discussion. I don't know who users are, what their demographics are, etc.. (and I don't want to store that info), so hopefully in the future polls like that people will explain the "why".
I need to build some more analytics into the site (both frontend and backend) so I can analyze the data and visualize it, and so users on the frontend can get better ideas on what is happening.
This is very much in the same space as what I’ve been building, so it’s interesting to see the overlap in instincts here!
I started working on Civie (https://civie.org) 2 months ago, which also focuses on anonymous, low-friction daily civic polling with time-boxed questions and an emphasis on privacy by default. Different implementation choices, but the same core belief: participation should be easy, frequent, and not require turning your identity into the product.
Would love to chat sometime if you're interested in comparing notes and implementation!
Hey absolutely! I checked out your website and looks great, I guess our ideas are validating each other! I would definitely be interested in learning more about your choices and implementation on demographics/verification, as I started down that path before shying away.
I had originally considered using protected session storage in the browser for storing demographic information so it never got stored on my server and stayed with the user, but decided for now to go as simple as possible with two big buttons. My main concern was, even if the vote was anonymous (especially when starting up without a large user base), if I had the users demographics stored on their account and if there were not many votes from say their region, you could in theory narrow it down to who that person was (email, ethnicity, age, location). If you had one user who was from Wyoming with those sets of demographics and you had one vote from Wyoming, well... I could make an educated guess who that was who voted and what they voted. I may have been overthinking it, but I just didn't feel personally confident enough to move forward with that, at least at the time.
I do also really like how you give the users context during their voting (pros/cons). I went with context after voting to get the gut reaction then drive discussion based on context. Really interesting to see!
Update: The HN traffic exposed a bug I hadn't caught... static caching of user-specific data in Blazor Server was causing session bleed under load. One comment got attributed to the wrong user. I am pushing a fix now, so the site will be down for a few minutes while it deploys.
If you posted a comment about "people find themselves in situations that shape their fate" and it's not showing under your username, you can email me at hello@polliticalscience.vote and I'll fix it.
Dang, perfect timing, ran into an issue with GitHub actions being down. Then the container timed out in Azure after manual publish. Should be back up now.
This is cool. My main question is just, what is its purpose, if not just a coding experiment?
Like you say, it's not scientific or representative. Is it for entertainment, do you want to gamify it? Is it pedagogical? Why is it anonymous? Do you want it to get picked up by the media? Are you trying to demonstrate something about public opinion or polarization? Do you want it to become popular? Do you want it to become more accurate? Or is it just a toy?
I have so many questions just because it could be so many different things, and the idea of a single daily poll on the main current event feels like it could have legs. (Though I don't know what today's poll about the death penalty has anything to do with today's or yesterday's news cycle?)
I am not really sure honestly. I thought of the idea a couple years ago and thought it was interesting. I follow politics and current events, but never really participated in online discussions of it. I have a background in data analytics and have loved stats since I was a kid. I always followed sports stats and election poll stats. I have been working on an enterprise application for ~ 3 years now, that I hope to beta test this year and it has been pretty heavy. I decided to take a short break and just get this idea I had out there and see how it goes.
The reason it is anonymous is I do not want to tie users to votes. A couple reasons being liability. If I know who you are, how you are voting, and your demographics, that is pretty powerful, but also a ton of liability. If something happened and that data got leaked out, that could be awful. I also don't think users are as likely to create an account, give away their information, just to hit two big buttons. The goal was no barrier to entry, sign up if you want more and not to farm political data from users.
Of course I would like it to become popular, be a place for thousands to discuss hot topics, and get enough votes that it washes out any abuse and grabs enough diversity to see real sentiment. I don't know if it will ever get to that point though. I don't plan to make it scientific as that would require removing the anonymous nature of it. I have thought about it as a free tool for universities or high schools to use for current events polls and discussions.
The short version, I have no idea haha. It really is a fun side project for now, it was fun to code and get something out there, but I am interested to see where it goes.
I think that if you keep it anonymous like that, then it's limited to basically being a toy, since there's zero accuracy, nothing to prevent "ballot-stuffing", etc. Which is fine, if you just want to have built it for fun.
But if you wanted to, I do think this is an idea that could take off in a popular way, the idea of America's "question of the day" that people answer every morning after they do their Wordle (or their enclose.horse, ha). But it would require creating an account linked to probably a phone number, that you log in with via a code, and answer basic demographic questions after your first vote that you'd use for balancing. I don't think you need to worry too much about liability if you follow best practices around database security? (And if it really took off, you could put it behind a cheap LLC to protect you personally.) Obviously you don't want people to think you're selling their data to political operatives, so you might want to partner with a political scientist at some well-known university to give some kind of "academic seal of approval" that this is used for research and public information, not for selling data. I think there could really be something here, and let people suggest and vote on what tomorrow's question should be.
Just something to think about, if you did want to try to turn it into something popular that could become part of the news environment, the way FiveThirtyEight has. But if you value the anonymous aspect the most and just want to keep it as something smaller for fun, then that's cool on its own. :)
Thanks for the thoughtful ideas! You make some great points and laid out a few good plans honestly. I do have an "account" feature right now, where you just enter your email to login. It is passwordless. Logged in accounts can only vote 1 time. I still have the fundamental difficulty of not knowing what they actually voted to maybe have two counts, the anonymous vote and the logged in vote.
Maybe if this somehow takes off, I can re-evaluate if there is a better way to maybe have the anonymous voting as well as a more "scientific" version where users opt-in to provide their information or a way to verify it before anonymously storing their vote.
I think it would be cool to track how votes differ depending on where a user was linked from. Being able to see e.g. "x% from hackernews support death penalty, y% from x". You wouldn't just be polling but also showing differences between users of different sites.
That would actually be really interesting since most of my users so far are from Reddit and Bluesky. Seeing how that breaks down would be telling of how sites primary users feel.
The only issue is, while Plausible analytics that I am using is really nice since it is privacy focused and doesn't track people, I have noticed it doesn't do a great job of understanding what links brought someone there. I am sure Google Analytics or others would do a better job, but it sort of circumvents the anonymous idea. Especially since this is a side project mostly for fun/interesting, I don't really want to be responsible for linking people to political choices.
You should probably say somewhere that the questions are US-centric. For example, "the death penalty should remain legal" doesn't make sense in places where it isn't.
Hey, you are not wrong. That is technically true. So the way I format the questions are always a statement (not a question) otherwise the response is really Yes/No more so than Agree/Disagree. I prefer agree/disagree because it makes the user take a stance more so than answer a question.
I also always word them in the affirmative. X should Y or X was right to Y or X is Y. This is so users understand the flow and are not tripped up by X should "not" Y.
It doesn't always look clean depending on the topic, but it is focused on sentiment of the statement rather than the specifics (the discussion section could be for that).
feature request? dumb question, but can you add underneath or after you click your vote and before you show it. can you ask "What do you think most people answered" for that question.
It is cool to see the distribution of yes/no. But maybe when you do that you can do a kind of....how far off was i type result that lets people learn about their...biases? or just a fun surprise.
Haha so like "guess" which way you think public sentiment leans before you submit your vote? Were you thinking like "what percent agreed" or something? You type 55% then it shows you actually 22% agree sort of deal?
That could maybe be a little optional thing to make it more "interactive". The original idea was to be dead simple, two big buttons. The issue is attracting people to come back since they hit the big button, go "that's interesting", then forget it ever existed.
user clicks Yes. new one pops up
Show Results (button) or Guess which was more popular: Yes/No. This is more simple than guessing a percentage. but both could work.
So you then show the results and then you show the "meta" results which is like a "line" that is labelled where it shows what others thought of other people. The dead simple idea is a nice draw, this like...ah hah learning moment to learn or see more about the world I think is maybe a nice feature to add. to get more people to come back?
I'm based in the US and trying my best. I'll have some more broad polls at some point. At one question per day, I'll have a hard time getting everything and everyone included.
A lot of the statements are "general" sentiment as well. Social media, drinking age, phones and children at school, death penalty as a whole (even if it does say remain legal), etc...
Any particular international topics you'd find interesting as a poll statement?
That is a great idea! So I actually have two "meta" tags that will be used as I get going more. One is "revisited". This one will tag statements that have been asked before and I'll then add in more analytics to see how sentiment shifts. The other one is "recurring". I'll likely ask some of the more boring straightforward "The country is headed in the right direction" statements. Then set up some charts to see how it tracks over time.
I am hoping the general public finds this interesting as a lot of the major polls (who to be fair do it scientifically) only ask a couple thousand people. Many never get to participate. While this is not scientific, anyone can participate and eventually with enough people, we can maybe gauge and see how sentiment shifts after time.
The problem with binary responses vs continuous or discrete responses is that the way the question is worded has a significant impact. If all you can do is agree or disagree, then you likely only get responses from those at either extremity of the opinion line.
You're right on that. That is one reason I am avoiding scientific polling and calling it "public sentiment".
The binary, with all its flaws, is dead simple. I see a statement, do I agree or disagree with that statement. If I don't fully support one or the other, the discussion section I intended for them to make their case.
I am trying to keep the same similar format for each poll. X should Y or X was right to Y in hopes people get used to the wording. On the backend I have a bias rating I give each poll as well as a temperature. The bias is LeansLeft, Center, LeansRight. The temperature is Cool, Warm and Hot. For most warm to hot polls, the statement has to be worded in the affirmative towards one side or the other. I try to keep the wording bias somewhat balanced so I don't post too many in one direction.
I am playing around with the idea of having two steps though, Agree/Disagree (required) then an optional follow-up to rate where on the scale you fall. I want to keep the simplicity as the primary target, but it could be interesting to see where people fall on a range. It would still be anonymous but rather this vote was agreed and they leaned 75% towards that side or something.
Website says "no tracking" on the frontpage. I look at ublock origin, it mentions one blocked domain called "plausible.io". I go to plausible.io and see that "Easy to use and privacy-friendly Google Analytics alternative Plausible is powerful, lightweight analytics. No cookies, just insights. Made and hosted in the EU, powered by European-owned infrastructure. "
"No tracking" is a different concept than "Google analytics alternative".
Maybe it was changed quickly, but I can't find anywhere that is says "No tracking". It specifically says "We don't track you around the internet." and "doesn't track you across sites" in the terms and about pages.
Also, you kind of have to "track" users to some extent for a site like this - otherwise it would be simply for someone to stuff votes.
The Plausible does count raw statistics without "tracking" specific users. That is just used for general website analytics. The first-party functional cookie that I am using (very similar to the auth login cookie) is used to prevent duplicate anonymous votes. Neither of these track the user and both are for on-site only. The functional cookie works much better than the fingerprint (actually less invasive too), but isn't full proof. You can switch browsers, go to incognito mode in some browsers, etc.. to bypass it, but it works for most casual users. Since it isn't election level polling, I figured it is fine. I do have an in memory rate limit to prevent excessive voting spam.
Actually no, I'm rate limiting per individual IP address right now. Good catch... I should probably normalize IPv6 to /64. I was originally thinking about not blocking universities or large groups that share IPs, but I guess that is more of an IPv4 NAT concern. Thanks for pointing it out! I didn't really think about a user rotating through IPs. I didn't add the rate limiting on voting until I removed the fingerprint, so that is for sure a valid concern.
It could make sense to lightly rate limit at /48 in addition to /64 (this is generally the largest subnet size given out by ISPs), otherwise it will be easy for people to multiply your /64 rate limit by 65536.
Hey thanks for the recommendation. That makes sense. Layered rate limiting at both /64 and /48 with different thresholds. Appreciate the explanation, and I'll be adding this to the list! This is my first time dealing with a public facing app where this type of rate limiting is needed.
So the site does use Plausable analytics. I chose that one because it does not use cookies, does not identify the user in any way, and does not follow you across the internet.
It counts raw aggregate statistics and is compliant with GDPR without requiring a banner. While it is "tracking" I suppose, it doesn't "track you". Do you think my wording doesn't work? I am open to suggestions.
I guess I will have to dive into this a bit more. I don't want to make false claims on the site. From the research I did before choosing Plausible, they do not collect any user data. They collect statistics of things that happened, but no actual user data.
From their website:
"By using Plausible, you don’t need to have any GDPR, CCPA or PECR prompts and you don’t need a complex privacy policy about your use of analytics and cookies. With Plausible, you are not tracking any personal data after all. Your visitors can enjoy your site without any annoyances and distractions."
You can't even tell if the same person comes back on a different day.
Honestly, I don't even really need them. I may just remove it entirely. I am not a B2C or B2B website. It doesn't really matter to me that much to have the stats, but it is nice in general to see how it is doing. The votes submitted sort of count the users for me anyways.
Cool project! The results of all the archived votes made sense to me, but I was most surprised by this one:
> The U.S. was right to withdraw from the World Health Organization (WHO)
> 62% Agree - 38% Disagree
I didn't know that the WHO had such a negative reputation. We are quite fond of such international institutions in the EU at least (ranging from a force for good to fairly harmless). What's the context? The rest of the votes seem quite liberal leaning otherwise.
One of the reasons is that the US pays for a whole lot of those international institutions, creating policy and governance issues that end up beyond local accountability. The "force for good, or fairly harmless" rubric changes when it's your money. Then it becomes "why are they spending my money on that bullshit when we have fires to put out at home?"
Covid era politicization and the fallout from that has a lot to do with it as well.
Thank you that makes sense. I did a bit more research for context:
The US only left in 2020 and then rejoined in 2021, I suppose that’s why I didn’t remember it as a big thing.
The US was also just paying ~15%. It was the biggest governmental funder, with Germany at ~9% as the second. But the WHO is apparently mostly funded by charity donations, the Bill & Melinda Gates Foundation was paying ~5% for instance.
(it’s awkward to list sources on the phone but should be easy to verify)
I do get the sentiment though from the perspective of the US, I don’t mean to argue your points.
Thanks for checking it out!
Most of my voters have come from Reddit and Bluesky so far, which is primarily where the left leaning is coming from. My X account was unfortunately suspended lol. I put an appeal in, but was originally flagged I think do to a new account, political content, and lots of links to the polls. I use an OG dynamic card generator so if I post a link to that poll or result, it creates a card for it on the fly. I think X didn't like that since I wasn't established.
That one was interesting, I am not really sure why that one skewed so far the other directly. I did not have the discussion section open yet (and just slowly getting a few active users), but that was the original reason I added the discussion. I don't know who users are, what their demographics are, etc.. (and I don't want to store that info), so hopefully in the future polls like that people will explain the "why".
I need to build some more analytics into the site (both frontend and backend) so I can analyze the data and visualize it, and so users on the frontend can get better ideas on what is happening.
This is very much in the same space as what I’ve been building, so it’s interesting to see the overlap in instincts here!
I started working on Civie (https://civie.org) 2 months ago, which also focuses on anonymous, low-friction daily civic polling with time-boxed questions and an emphasis on privacy by default. Different implementation choices, but the same core belief: participation should be easy, frequent, and not require turning your identity into the product.
Would love to chat sometime if you're interested in comparing notes and implementation!
Hey absolutely! I checked out your website and looks great, I guess our ideas are validating each other! I would definitely be interested in learning more about your choices and implementation on demographics/verification, as I started down that path before shying away.
I had originally considered using protected session storage in the browser for storing demographic information so it never got stored on my server and stayed with the user, but decided for now to go as simple as possible with two big buttons. My main concern was, even if the vote was anonymous (especially when starting up without a large user base), if I had the users demographics stored on their account and if there were not many votes from say their region, you could in theory narrow it down to who that person was (email, ethnicity, age, location). If you had one user who was from Wyoming with those sets of demographics and you had one vote from Wyoming, well... I could make an educated guess who that was who voted and what they voted. I may have been overthinking it, but I just didn't feel personally confident enough to move forward with that, at least at the time.
I do also really like how you give the users context during their voting (pros/cons). I went with context after voting to get the gut reaction then drive discussion based on context. Really interesting to see!
Update: The HN traffic exposed a bug I hadn't caught... static caching of user-specific data in Blazor Server was causing session bleed under load. One comment got attributed to the wrong user. I am pushing a fix now, so the site will be down for a few minutes while it deploys.
If you posted a comment about "people find themselves in situations that shape their fate" and it's not showing under your username, you can email me at hello@polliticalscience.vote and I'll fix it.
Dang, perfect timing, ran into an issue with GitHub actions being down. Then the container timed out in Azure after manual publish. Should be back up now.
This is cool. My main question is just, what is its purpose, if not just a coding experiment?
Like you say, it's not scientific or representative. Is it for entertainment, do you want to gamify it? Is it pedagogical? Why is it anonymous? Do you want it to get picked up by the media? Are you trying to demonstrate something about public opinion or polarization? Do you want it to become popular? Do you want it to become more accurate? Or is it just a toy?
I have so many questions just because it could be so many different things, and the idea of a single daily poll on the main current event feels like it could have legs. (Though I don't know what today's poll about the death penalty has anything to do with today's or yesterday's news cycle?)
Very clever domain name btw.
I am not really sure honestly. I thought of the idea a couple years ago and thought it was interesting. I follow politics and current events, but never really participated in online discussions of it. I have a background in data analytics and have loved stats since I was a kid. I always followed sports stats and election poll stats. I have been working on an enterprise application for ~ 3 years now, that I hope to beta test this year and it has been pretty heavy. I decided to take a short break and just get this idea I had out there and see how it goes.
The reason it is anonymous is I do not want to tie users to votes. A couple reasons being liability. If I know who you are, how you are voting, and your demographics, that is pretty powerful, but also a ton of liability. If something happened and that data got leaked out, that could be awful. I also don't think users are as likely to create an account, give away their information, just to hit two big buttons. The goal was no barrier to entry, sign up if you want more and not to farm political data from users.
Of course I would like it to become popular, be a place for thousands to discuss hot topics, and get enough votes that it washes out any abuse and grabs enough diversity to see real sentiment. I don't know if it will ever get to that point though. I don't plan to make it scientific as that would require removing the anonymous nature of it. I have thought about it as a free tool for universities or high schools to use for current events polls and discussions.
The short version, I have no idea haha. It really is a fun side project for now, it was fun to code and get something out there, but I am interested to see where it goes.
Interesting, thanks.
I think that if you keep it anonymous like that, then it's limited to basically being a toy, since there's zero accuracy, nothing to prevent "ballot-stuffing", etc. Which is fine, if you just want to have built it for fun.
But if you wanted to, I do think this is an idea that could take off in a popular way, the idea of America's "question of the day" that people answer every morning after they do their Wordle (or their enclose.horse, ha). But it would require creating an account linked to probably a phone number, that you log in with via a code, and answer basic demographic questions after your first vote that you'd use for balancing. I don't think you need to worry too much about liability if you follow best practices around database security? (And if it really took off, you could put it behind a cheap LLC to protect you personally.) Obviously you don't want people to think you're selling their data to political operatives, so you might want to partner with a political scientist at some well-known university to give some kind of "academic seal of approval" that this is used for research and public information, not for selling data. I think there could really be something here, and let people suggest and vote on what tomorrow's question should be.
Just something to think about, if you did want to try to turn it into something popular that could become part of the news environment, the way FiveThirtyEight has. But if you value the anonymous aspect the most and just want to keep it as something smaller for fun, then that's cool on its own. :)
Thanks for the thoughtful ideas! You make some great points and laid out a few good plans honestly. I do have an "account" feature right now, where you just enter your email to login. It is passwordless. Logged in accounts can only vote 1 time. I still have the fundamental difficulty of not knowing what they actually voted to maybe have two counts, the anonymous vote and the logged in vote.
Maybe if this somehow takes off, I can re-evaluate if there is a better way to maybe have the anonymous voting as well as a more "scientific" version where users opt-in to provide their information or a way to verify it before anonymously storing their vote.
I think it would be cool to track how votes differ depending on where a user was linked from. Being able to see e.g. "x% from hackernews support death penalty, y% from x". You wouldn't just be polling but also showing differences between users of different sites.
That would actually be really interesting since most of my users so far are from Reddit and Bluesky. Seeing how that breaks down would be telling of how sites primary users feel.
The only issue is, while Plausible analytics that I am using is really nice since it is privacy focused and doesn't track people, I have noticed it doesn't do a great job of understanding what links brought someone there. I am sure Google Analytics or others would do a better job, but it sort of circumvents the anonymous idea. Especially since this is a side project mostly for fun/interesting, I don't really want to be responsible for linking people to political choices.
Plausible is probably just looking at the Referer header.
You should probably say somewhere that the questions are US-centric. For example, "the death penalty should remain legal" doesn't make sense in places where it isn't.
I should probably add that. I do plan to add some more international statements, but yeah, right now it is very U.S.-centric.
A better question might be "should the death penalty be legal", for that reason. After all, it isn't even legal in about half of the USA.
Hey, you are not wrong. That is technically true. So the way I format the questions are always a statement (not a question) otherwise the response is really Yes/No more so than Agree/Disagree. I prefer agree/disagree because it makes the user take a stance more so than answer a question.
I also always word them in the affirmative. X should Y or X was right to Y or X is Y. This is so users understand the flow and are not tripped up by X should "not" Y.
It doesn't always look clean depending on the topic, but it is focused on sentiment of the statement rather than the specifics (the discussion section could be for that).
feature request? dumb question, but can you add underneath or after you click your vote and before you show it. can you ask "What do you think most people answered" for that question.
It is cool to see the distribution of yes/no. But maybe when you do that you can do a kind of....how far off was i type result that lets people learn about their...biases? or just a fun surprise.
Anyways fun idea!
Haha so like "guess" which way you think public sentiment leans before you submit your vote? Were you thinking like "what percent agreed" or something? You type 55% then it shows you actually 22% agree sort of deal?
That could maybe be a little optional thing to make it more "interactive". The original idea was to be dead simple, two big buttons. The issue is attracting people to come back since they hit the big button, go "that's interesting", then forget it ever existed.
The death penalty should remain legal. Yes/No
user clicks Yes. new one pops up Show Results (button) or Guess which was more popular: Yes/No. This is more simple than guessing a percentage. but both could work.
So you then show the results and then you show the "meta" results which is like a "line" that is labelled where it shows what others thought of other people. The dead simple idea is a nice draw, this like...ah hah learning moment to learn or see more about the world I think is maybe a nice feature to add. to get more people to come back?
Interesting how the website is called "PoliticalScience" but mainly cares about US Politics
I'm based in the US and trying my best. I'll have some more broad polls at some point. At one question per day, I'll have a hard time getting everything and everyone included.
A lot of the statements are "general" sentiment as well. Social media, drinking age, phones and children at school, death penalty as a whole (even if it does say remain legal), etc...
Any particular international topics you'd find interesting as a poll statement?
It would be cool to be able to see how sentiment changes for a specific issue over time. Maybe you could recycle questions every so often?
That is a great idea! So I actually have two "meta" tags that will be used as I get going more. One is "revisited". This one will tag statements that have been asked before and I'll then add in more analytics to see how sentiment shifts. The other one is "recurring". I'll likely ask some of the more boring straightforward "The country is headed in the right direction" statements. Then set up some charts to see how it tracks over time.
I am hoping the general public finds this interesting as a lot of the major polls (who to be fair do it scientifically) only ask a couple thousand people. Many never get to participate. While this is not scientific, anyone can participate and eventually with enough people, we can maybe gauge and see how sentiment shifts after time.
Awesome! I'm excited to see how this goes.
The problem with binary responses vs continuous or discrete responses is that the way the question is worded has a significant impact. If all you can do is agree or disagree, then you likely only get responses from those at either extremity of the opinion line.
You're right on that. That is one reason I am avoiding scientific polling and calling it "public sentiment".
The binary, with all its flaws, is dead simple. I see a statement, do I agree or disagree with that statement. If I don't fully support one or the other, the discussion section I intended for them to make their case.
I am trying to keep the same similar format for each poll. X should Y or X was right to Y in hopes people get used to the wording. On the backend I have a bias rating I give each poll as well as a temperature. The bias is LeansLeft, Center, LeansRight. The temperature is Cool, Warm and Hot. For most warm to hot polls, the statement has to be worded in the affirmative towards one side or the other. I try to keep the wording bias somewhat balanced so I don't post too many in one direction.
I am playing around with the idea of having two steps though, Agree/Disagree (required) then an optional follow-up to rate where on the scale you fall. I want to keep the simplicity as the primary target, but it could be interesting to see where people fall on a range. It would still be anonymous but rather this vote was agreed and they leaned 75% towards that side or something.
Website says "no tracking" on the frontpage. I look at ublock origin, it mentions one blocked domain called "plausible.io". I go to plausible.io and see that "Easy to use and privacy-friendly Google Analytics alternative Plausible is powerful, lightweight analytics. No cookies, just insights. Made and hosted in the EU, powered by European-owned infrastructure. "
"No tracking" is a different concept than "Google analytics alternative".
Maybe it was changed quickly, but I can't find anywhere that is says "No tracking". It specifically says "We don't track you around the internet." and "doesn't track you across sites" in the terms and about pages.
Also, you kind of have to "track" users to some extent for a site like this - otherwise it would be simply for someone to stuff votes.
The Plausible does count raw statistics without "tracking" specific users. That is just used for general website analytics. The first-party functional cookie that I am using (very similar to the auth login cookie) is used to prevent duplicate anonymous votes. Neither of these track the user and both are for on-site only. The functional cookie works much better than the fingerprint (actually less invasive too), but isn't full proof. You can switch browsers, go to incognito mode in some browsers, etc.. to bypass it, but it works for most casual users. Since it isn't election level polling, I figured it is fine. I do have an in memory rate limit to prevent excessive voting spam.
Are you rate limiting at the subnet/prefix level for IPv6?
Actually no, I'm rate limiting per individual IP address right now. Good catch... I should probably normalize IPv6 to /64. I was originally thinking about not blocking universities or large groups that share IPs, but I guess that is more of an IPv4 NAT concern. Thanks for pointing it out! I didn't really think about a user rotating through IPs. I didn't add the rate limiting on voting until I removed the fingerprint, so that is for sure a valid concern.
It could make sense to lightly rate limit at /48 in addition to /64 (this is generally the largest subnet size given out by ISPs), otherwise it will be easy for people to multiply your /64 rate limit by 65536.
Hey thanks for the recommendation. That makes sense. Layered rate limiting at both /64 and /48 with different thresholds. Appreciate the explanation, and I'll be adding this to the list! This is my first time dealing with a public facing app where this type of rate limiting is needed.
On the landing page, right underneath the "disagree" button it says "no tracking"
So the site does use Plausable analytics. I chose that one because it does not use cookies, does not identify the user in any way, and does not follow you across the internet.
It counts raw aggregate statistics and is compliant with GDPR without requiring a banner. While it is "tracking" I suppose, it doesn't "track you". Do you think my wording doesn't work? I am open to suggestions.
GDPR is not about cookies it is about you giving user data to another company, and by including plausible.io you are doing that.
I guess I will have to dive into this a bit more. I don't want to make false claims on the site. From the research I did before choosing Plausible, they do not collect any user data. They collect statistics of things that happened, but no actual user data.
From their website:
"By using Plausible, you don’t need to have any GDPR, CCPA or PECR prompts and you don’t need a complex privacy policy about your use of analytics and cookies. With Plausible, you are not tracking any personal data after all. Your visitors can enjoy your site without any annoyances and distractions."
You can't even tell if the same person comes back on a different day.
You can see their full privacy statements here: https://plausible.io/privacy-focused-web-analytics.
Honestly, I don't even really need them. I may just remove it entirely. I am not a B2C or B2B website. It doesn't really matter to me that much to have the stats, but it is nice in general to see how it is doing. The votes submitted sort of count the users for me anyways.