Four cardinal errors of Beeminder goals: the leniency/expense coordinate plane

Akrasia
1

Imagine a coordinate plane on which your implementation of a Beeminder goal lives.

The x-axis corresponds to how expensive the goal is - higher (maximum) pledge amounts are further East.

The y-axis corresponds to how lenient or strict you are in your implementation, rules, and fine print of the goal. If you have a lot of fine print that carefully rules out a lot of situations where you might call non-legit, and you follow that, that would be more strictness, and would be further North. No-excuses mode is very far North. Having to submit proof would also be far North. Calling non-legit for anything and everything would be very lenient and very far South.

So, that leads to four “cardinal errors” for Beeminder goals, one for each cardinal direction:

  • Too far North - too strict. Problems from this include losing a lot of money in a real emergency or when you’re sick, reneging on your promise to pay because it’s so much and feels too strict, or giving up because it’s too overwhelming.

  • Too far South - too lenient. This can cause your standards to erode until you call non-legit on everything or find a way to circumvent everything.

  • Too far West - (maximum) pledge amounts too low. This can cause you to just choose to derail more often than you’d like.

  • Too far East - (maximum) pledge amounts too high. This can cause frustration, financial loss, reneging, or giving up, and also lead you to sacrifice your health or other important things to make sure you never derail even when it would be better to derail.

Also of importance, though not included in this 2-D model, is how hard the goal’s setup actually is - for instance, you can always lower the rate or turn weekends off to make a goal easier.

The point here is that there is no perfect setup - any setup is going to be prone to error in some of these ways. Too strict, and you suffer in a real emergency; too lenient, and your standards erode.

One key is to have a mix of different locations on the plane with all your different goals. For instance, I have half my goals in no-excuses mode, so I don’t just call non-legit on everything, but still don’t lose everything when I’m actually sick.

2 Likes
2

I feel difficulty is far more relevant than strictness. I don’t call non-legit very often, really only for technical issues. I also normally run with some safety buffer, so if I need to (briefly) skip something I can do so without derailing. I’d put difficulty on the y axis.

1 Like
3

Given the progressive pledge structure, this is only ever a temporary problem.

1 Like
4

I’m pretty much the opposite! I’m basically maximally akratic, so if I have any safety buffer I won’t do the task! And I’m constantly fighting the temptation to call non-legit, so that’s always something to guard against.

No, because you have a maximum that the pledge can reach. I should have specified that I was talking about the maximum pledge amount - actually I’ll go back and edit that.

I was tacitly assuming that your pledges have already reached their maximum (I just set mine to the maximum at all times anyway - I prefer to start out each goal with the pledge already at its max).

If your maximum is too low, the progressive structure will no longer increase them past that point so it doesn’t help with this problem.

1 Like
5

Yes, I can see this is different for edge-skaters, that leniency is far more relevant if you are always in the red.

Yes, I misunderstood the “pledge too low” thing. Setting your maximum pledge too low makes sense.

I flake out when the goal is too hard AND the pledge is too high.

1 Like