Advent 2022: 06. Pessimistic Presumption in Do More Goals (#SELFDESTRUCT datapoints)

Today’s Advent Calendar post reveals a way to make a goal red immediately without permanently losing any of its safety buffer.

Pessimistic presumption is officially a feature of Do Less goals, however it works with other goal types too. To use it:

  • Enter a datapoint that has any value (positive or negative).
  • Include a comment that contains #SELFDESTRUCT
    • must be uppercase
    • must have the # mark
    • can be by itself or have any words around it
    • spaces on either side are not needed (e.g., aaaa#SELFDESTRUCTaaa works)
  • You’ll see the datapoint appear on your graph as normal.
  • Later in the day, enter your real datapoint for that goal.
  • You’ll see that the pessimistic presumption datapoint is automatically deleted.

How is this useful? Well say you have a Do More goal that you’ve worked hard to get to green and you want to keep your safety buffer, but you also want to force yourself to make more progress on the goal today. Enter a pessimistic presumption (PP) datapoint with a large negative value, and you’ll see the graph go red immediately. Then do some work on the goal and enter a new datapoint. Your goal goes back to green straight away, with a larger safety buffer, and with the PP datapoint deleted.

The exact value of the PP datapoint isn’t important, as long as its enough to change the colour of your graph. I usually use -99 because it’s easy to enter and guaranteed to make any of my goals red.

When you enter a real datapoint for the goal, it doesn’t have to match the PP datapoint (e.g., you don’t need to enter +99). Any value will cause the PP datapoint to disappear.

I use this because I love seeing many of my goals green at the end of the day but I find it hard to motivate myself to work on a goal if it’s not red. Of course there’s nothing stopping me from manually deleting the PP datapoint at the end of the day, and sometimes I do that without guilt (especially if it’s been a really busy day) but usually this trick makes me actually work on the goal.

You can enter a PP datapoint for a future day too! For example, if I wanted to encourage myself to work on a goal tomorrow, I’d enter a PP datapoint with tomorrow’s date. It wouldn’t have any effect on the goal colour today (and wouldn’t cause it to derail today), but when I wake up the next morning, the goal is red and begging for attention. :slight_smile:

This doesn’t work well with goals where you’d typically enter multiple datapoints each day, because as soon as the first datapoint goes in, the PP datapoint is deleted. For those goals, I’ll sometimes enter a negative datapoint with the comment “DELETE” and then manually remove it later.


This is so smart! Thanks! I love having buffer but I I’ve found that it becomes harder to take decisions when goals are not red. Maybe it’ll fix it.


Three hurrahs for whatever genius implemented those autodestructing datappoints. :blush:


This, once again, is genius. I had never thought of doing this with do-more goals. Fittingly, this was my commentary in the daily beemail after the previous Advent post:

Alys’s Beeminder Advent Calendar series is mind-blowingly good. I’m learning things myself. (Yes, me, the CEO of Beeminder; I’m not sure whether to be embarrassed about that but I’m going with the story that Alys’s posts just really are that good. And I guess also Beeminder is such a big beautiful bounteous beast that it can’t fit in any one human’s head.)

It’s been on my list to make that less hideous. Possibly just “#PPR” will be the magic string that makes this work. What do people think? It would mean that if you have hashtags turned on, the PPR would be visually indicated on the graph, which maybe is a good thing for transparency and warning the user?

Ha, yes, thank you @philip! :tada:


A bit less hideous might be nice (but I have it saved in a keyboard macro on Android and a clipboard thing on Linux so it never bothers me).

Too simple though and you risk deleting real data. For example someone reading 20 pages of a book about Professional Practice Research and annotating their graph with a hashtag datapoint wouldn’t want it deleted when they decide to read one more page just before bed. :slight_smile:


Ha, yes, ok, we better be pretty careful with this one. How about #SELFDESTRUCT?

(That also has the advantage of making it slightly clearer that it’s in fact a general feature not limited to do-less goals. The autogenerated comment can still also describe how it’s a pessimistic presumption, etc, just that “#SELFDESTRUCT” is the magic string that effects the self-destruction.)


I like that! Feels safe enough and obvious.


I’ve edited the post so that now instead of “PESSIMISTIC PRESUMPTION (autodestructs with new data)” it says “#SELFDESTRUCT”. The original string still works though, and “#THISWILLSELFDESTRUCT” also works. See Subtle change to self-destructing datapoints for details if you’re curious.


Oddly, I think my adverse reaction to Beeminder-created hashtags possibly doesn’t extend to the self-destructing ones — by definition I want to eliminate those on my graph by entering the real data.

Don’t think we had hashtags back when this got implemented. #SELFDESTRUCT is descriptive of what happens to the datapoint but not a clue as to HOW the user might go about triggering that destruction. Arguably the destruction is a delightful side-effect rather than the intended message, so perhaps #ENTER_MISSING_DATA or similar