Cake
  • Log In
  • Sign Up
    • Is it possible to have a passion for problem solving?

      I certainly do, so heck yeah.

      I mean enough to have regular conversations about it?

      I think so.

      I am throwing up a trial balloon to see if there are enough like minded people here who would enjoy spirited discussions on solving real and hypothetical problems.

      If I get a half dozen emoji thumbs up 👍 or “heck yeah” comments in the next few days, I will create a Problem Solving topic and a first post.

      Why is this your passion?

      I used to travel around the world looking for ways to improve the operations of organizations. I’ve been a project manager troubleshooting system implementations. I have a masters degree in mathematics and occasionally create recreational maths puzzlers for fun.

      I love optimization problems: the discussions in Cake and Cake feedback have been like brain candy 🍭 for me.

    • Heck yeah!

      Here’s an example problem I puzzle often: suppose you run a site like Reddit (ahem, Cake). On Reddit, you can set your feed to new and get a firehouse of posts with zero upvotes because they are new. You have to be a special person to wade through all that, but if you are the community depends on you because what they see is the stuff that gets upvoted.

      What everyone experiences is sometimes something catches fire and gets upvoted. Sometimes the very same thing posted at another time or by a different user gets zero upvotes. And it’s gameable

      How best to solve this? Like Apple does with news by having a human editorial team help curate?

    • How best to solve this? Like Apple does with news by having a human editorial team help curate?

      Hey Chris,

      Let’s start with some relevant data that may help us to make an informed decision.

      Gaming the system is always going to be a problem when you give equal up votes to everyone.

      One approach is to give some participants multiple upvotes to assign to a new post, based on

      - the participants length of time on the platform (quantitative),

      - number of posts with 20+ comment threads (hybrid),

      - and overall awesomeness rating by the topic moderator (qualitative).

      It doesn’t matter what formula you come up with—there is no perfect set of metrics. What’s important is that it doesn’t come down to a popularity contest.

      Why is it so important to avoid popularity contest mechanics?

      - On Twitter there are secret groups that share one Hootsuite account and like and retweet each other’s tweets to go viral.

      - In the Blogosphere, there are mommy blogger groups that boost each other’s content on their blogs, on Twitter and in Facebook groups.

      Note: I know personally someone who is involved in such secret groups.

      I also have a family member who has 20,000 followers without ever resorting to such games. So even in the most gamed up system, it’s still possible for great content creators to get their voices heard.

      Curating Great Comments

      I saw your pic of the New York Times featuring a well-crafted comment. Being chosen for such an honor would be to me the equivalent of having your caption chosen for a New Yorker cartoon.

      From an optimization standpoint, you want to be able to have this process scalable when you grow to 80,000(?) participants. What if your moderator gets sick or you can’t hire more fast enough?

      One approach would be to curate the featured comment. Then assign the commenter featured status for a period of time: you could showcase their profile with all their recent posts and comments. As the community grows that time period could shrink from a week to a few days or even a few hours.

      Executive Summary:

      Use the data analytics built into your platform, as well as moderator judgment, to curate quality content. And provide 15 minutes of fame to the content creators who increase whatever engagement makes both ethical and economic sense for the platform’s long-term success.

    • I believe we need better tools to condition what is expressed and filter what is consumed. I feel we are too reliant on algorithms and majority opinion to get at the heart of what speaks to us.

    You've been invited!