Beeminder Forum

Voting on what the next blog post should be

This is a companion to the new (meta) blog post, “Happy Now, Beeminder?” – go read that, then come here and vote!

  • Ice Cream Truck Loopholes
  • The Midweek Megafast
  • The Stanford Marshmallow Experiment: A Retrospective
  • Lupine Lachrymosis; Or, The Anti-Wolf-Crying Principle
  • Contra Positive Reinforcement: Why Beeminder Is A Glutton For Punishment
  • Paying Is Not Punishment
  • Upside-Down Support
  • How To Name Your Beeminder Goals

0 voters

Official Tally

Here’s the official tally so far, including all the daily and weekly folks, who, as the blog post explains, are encouraged to vote again here and have their votes counted twice!

(a) icecream truck:   ********************* (21)
(b) midweek megafast: ************* (13)
(c) marshmallows:     **************************** (28)
(d) anti-wolf-crying: *************** (15)
(e) yay punishment:   ************************* (25)
(f) pay != punish:    ************************* (25)
(g) upside-down:      *************************** (27)
(h) naming goals:     ****************** (18)

~60 people voting, including via beemail

1 Like

@dreev would you be willing to give the tldr for the anti-wolf-crying principle? I have this feeling this is something I’m probably violating with TaskRatchet :flushed:

1 Like

Sure! We actually have a whole draft but are hoping to end it with something better than “do as we say, not as we do” before actually publishing it. :slight_smile: I.e., we’re also violating it! I really want to solve this.

Anyway, my off-the-cuff summary is that you need a system to strictly partition things into 2 categories: (1) things that go on your admin dashboard for you to have ambient awareness of, and (2) things that beep at you or text you or whatever – actual notifications. And the 0th rule for the notifications is that you absolutely must take immediate action on every single one. If you get one that doesn’t feel urgent enough to justify that, then the urgent immediate action is to make whatever caused it not generate a notification anymore.

1 Like

Have you done any thinking about how this would apply to alerts sent to end users?

I guess in that case it’s more like the anti-wolf-crying principle that everyone knows. Be careful not to over-alert and make users tune you out.

Relevant excerpt from one of our specs (note: not currently true!):

The new principle for do-less reminders is that there are exactly two reasons to alert you: (1) You need to add data to a mandata goal to prevent a PPR from derailing you. (2) You just moved dangerously closer to your hard cap on an autodata goal. Read on for what “dangerously closer” means. The only time we nag you (send repeated reminders saying essentially the same thing) is like with do-more goals: “OMG DO THE THING WHY HAVE YOU STILL NOT DONE THE THING”.

We never want to be like “DON’T DO THE THING ARE YOU STILL NOT DOING THE THING OK HOW ABOUT NOW??”. That’s wolf-crying.


Alright, marshallows eked out a win over upside-down support, so we published but we’ve got nice draft for upside-down support now as well so expect that one fairly soon as well!


Are the others in the pipeline as well? Or are they … pipe dreams? :wink::wink:

Postscript for the etymological nerds I know are lurking: pipe dream


I’d say it’s a mix (pipeline vs pipe dreams). Upside-down support is definitely happening! The marshmallow one totally wouldn’t have happened without this poll. I had a couple things to say about it but had to read a bunch of papers before I felt confident enough to enshrine it as a blog post. Beeminder blog posts are a big deal, y’all!

Happy to give tweet-sized summaries of other ones that folks are curious about.


It just occurred to me that “Ice Cream Truck Loopholes” is more likely to be about loopholes the size of a truck rather than loopholes related to ice cream, so now I’m less interested in that. lol

1 Like