• 0 Posts
  • 14 Comments
Joined 1 month ago
cake
Cake day: February 21st, 2026

help-circle

  • In voting theory, there are these voting graphs where every candidates is a node. If you rank every candidate, you can draw directional lines between each node then sum all the ranking from all the voters to find a cumulative ranking.

    Most people oppose this system for the practical reason of no one wanting to rank every candidate at the ballot box; however, I believe I’ve found a clever work around to this complaint. You have a none option (or a lottery option) and you allow people to rank people equally. From there it’s pretty trivial to set up a tablet or something where you can send candidates to the bottom or top and modify the <=> symbols between them. Everyone starts in a random order below the None/Lottery Option. If you want to get fancy you could even give people the option of grouping and moving an entire party on the tablet. In the cumulative ranking, anyone equal or below the None / Lottery option gets tossed. If it’s an election where you need multiple people just start at the top of the ranking and work your way down. Once you hit None/Lottery, your repeat the lottery or go without for any further seats.

    The None/Lottery Option also prevents it from being weak to large numbers of candidates as frankly people will just ignore the vast majority of candidates leaving them below the none/lottery option. In a polarized society people will put the opposing party below the none/lottery option. You can vote [lottery > blues > reds] or [blues > lottery > reds] and it’s the same result for red vs blue.

    There’s a slightly more advanced version of this where you put numbers on each relation then normalize. It gets complaints of not meeting the condorcet criterion, but it’s actually superior. I think this gets too complicated at the voting booth though, so whatever.

    Some people do criticize this because strategic voting can get weird, but since this system has a none/lottery option that argument doesn’t hold water. If the population “strategically” votes [blue > yellow > lottery > red] and [red > yellow > lottery > blue] then [yellow > lottery = red = blue] is the favored result. They could easily swap yellow and lottery and get [red = blue = lottery > yellow]. They made their choice. That’s democracy, we ought to respect it.

    Also, also, if it’s truly equal e.g. [red=blue > lottery ] just flip a coin. It’s unlikely to be truly equal but we’re already accepting some luck in this system.


  • I like the idea of adding a lottery option to some sort of ranked choice. I’m perfectly fine electing good politicians, but if a majority of people think they’re corrupt, we should be able to rank a lottery option above them.

    I’m fine with re-running if the chosen person opts out, but I don’t like the opt-in versions. I’m also not fond of some of the statistical biasing some people advocate with the system – a straight lottery where everyone has equal odds. I’d compromise on including felons, but personally I think including them incentivizes rehabilitation.

    I also worry that this effectively gives power to public servants who are not necessarily good people – wasn’t Stalin originally a secretary? I can see every think tank offering up people with their own agendas to work in a new office but having an established office with entrenched interests also seems super dangerous.






  • While I don’t disagree with the transparency Mozilla is advocates, I think it fails to address the underlying problem then tries to compensate by picking and choosing winners (which arguably is the same as the underlying problem). The underlying problem is the ad-incentivized watchtime algorithm, which isn’t a technical issue but a financing one.

    I’ve been an advocate of endowments for a long time, but this is just another area where they’d be ideal. They supply a small steady income to support a relatively cheap product. As the website grows you can either do temporary ads to grow the endowment or ask for donations. Either way, it’s not that hard to fund operations this small. Add in federated systems like lemmy and each individual operation is even smaller and cheaper.

    Heck, universities who are already accustom to dealing with endowments would be ideal places to host lemmy instances. I can definitely imagine offering to donate 10k to an endowment dedicated to hosting a lemmy and mastadon instance with open to registration to students, staff, and alumni. Maybe coordinate with the computer science and IT folks. Allow some percentage of the endowment income to go to “salary overhead” while the rest just funds the server. Point out that the university would essentially be creating the perfect route to solicit donations and they might do it themselves… Honestly, I’m probably gonna flesh this idea out and email the people at my university because it’s just too perfect of a solution.


  • There was a similar post recently about Cambridge leaving twitter, and it got me thinking that universities are really the ideal organizations to host lemmy servers. They have a vested interest in truth and community building. They have a decent enough sense of free speech to stay federated with most other instances. They have pre-existing communities on topic ranging from clubs to technical subjects. Users can confirm their identities by association with the universities, which will keep things civil. Obviously I don’t think they should be the only instances - anonymity has it’s place and value - but I really think universities should be hosting instances.