Will we (Beeminder) deem it prudent to make it harder to turn No-Excuses Mode back off after you opt in to it?
Basic
18
Ṁ2581
resolved May 11
Resolved as
5%
So far we've just made it show in the interface and in the legit check / derailment notification emails how long you've had No-Excuses Mode turned on or when you last turned it off. Presumably that adds a bit of disincentivizing shame to the loophole of quickly turning it back off when you're about to derail. This is a prediction about whether that will suffice. Background: https://blog.beeminder.com/noexcuses/ To make it slightly less subjective, let's factor out development cost. We'll call it "prudent" if, in the Platonic ideal of Beeminder (still per our opinion), it'd be harder opt back out of No-Excuses Mode than it currently is, even if it's too small a problem in practice for us to deem it worth implementing that. Close date updated to 2022-05-18 11:59 pm
Get
Ṁ1,000
and
S3.00
Sort by:
predictedNO
I discussed this with our Support Czar, @shanaqui, who's inclined towards NO. The status quo strikes a pretty nice balance. There has been the occasional frustrating person who's had No-Excuses Mode on and wanted to argue anyway and we kind of view such people as a lost cause and it's way easier for us to just say "weasels gonna weasel" and let them turn it back off. For nearly everyone else, it's enough to know how frowned-upon it is to abuse the loophole, and so they don't. (But I'm not fully confident of that so I'm still inclined to resolve to 5%. I.e., I'm 95% confident in the above reasoning, I guess!)
predictedNO
I think the current consensus is that this is a non-issue. But the "Platonic ideal" part of the resolution criteria make this a tough call! I'm inclined to just resolve-to-PROB using the market's probability, 5%, so mostly NO but not fully NO.
Agree with ian that it seems "inconsistent with literally every other aspect of Beeminder." On the other hand, it doesn't _really_ seem urgent or even like a problem.
(I'd say the probability is naturally going down just from Bayesian updating on the fact that we haven't seen a clamoring for this so far...)
I love how you think and I'm absolutely amenable to that. We might even calculate it retroactively. But why does that feature (essentially paying users to use No-Excuses Mode) imply that No-Excuses Mode would have to be akrasia-proofed? Maybe you just earn that premium credit each time you have a derailment with No-Excuses Mode in effect. So the incentive scheme is structured to discourage you from turning it off, without preventing you from doing so.
I'm less confident than I was before... I've since tried it out myself and, to my surprise, it doesn't really matter psychologically that it can be easily turned off. The following was basically the first thought that popped into my head when I heard about no-excuses mode. It seems fitting to post about it here. Would it make sense to pay out users in premium credit based on the percentage of their goals with no-excuses mode turned on, per day (obviously fractions of a dollar, such that it would take months to reach a substantial amount)? The reasoning being of course that a user with all of their goals set to no-excuses mode costs $0 (barring bugs) in support and will probably have more profitable goals than the average. It would serve as a forcing function to push more users to the no-excuses mentality. A scheme like that would definitely need the akrasia horizon for turning it back off though ;-)
I'm still pretty clueless about this. It's making more sense to me that it should be akrasia-proofed in principle but the evidence that it doesn't matter in practice is also increasing.
Thanks Ian! The "just add the 7-day delay for consistency" argument is compelling. We do say that anything that makes your goal easier, including ending it altogether, takes effect with a one-week delay. Aka akrasia-proofing. Turning off No-Excuses Mode would seem to count as making it easier. On the other hand, it's been a few days and we've seen zero instances of weaselishly turning No-Excuses Mode back off right before derailing. Maybe if you have the presence of mind to do that then you might as well just do the thing Beeminder's telling you to do, aka dispatch the beemergency. If it's never abused in practice then might as well not restrict the toggling of it. I think. Unless we still should on principle.
Signalling I'm expecting you to change your mind about this. 1. While the feature is ostensibly not meant for weasels, they will also enable it for (some) of the same reasons as before. When derailing it will be very easy to claim non-legitness with a straight face. "Oh, I'd been planning to turn the setting off for a while and just got around to it today. And then this totally unexpected and unrelated thing happened! (...)". This will be a pointless support burden and would be reason enough to make it harder to turn off. 2. It's inconsistent with literally every other aspect of Beeminder. Want to do x? Wait 7 days. But want to turn off this feature? Go right ahead. 3. There's not as clear a demarcation between weasels and "pure"/hardcore/... Beeminder users as you think there is (I'm guessing). Besides, I see this as a feature to enable for a specific goal not for a specific period of time ("today excuses are anathema, but tomorrow they're A-OK?!"). 4. It might even become impossible to turn this off when no-excuses mode becomes some kind of default. At least adding an akrasia horizon would prepare for that direction, instead of going against it.
I dumped in my remaining balance. I guess my true probability is... I'm gonna say 33% that there'll be enough abuse of no-excuses toggling that we'll make a gissue for addressing it. And I guess that probability goes down each day that we don't see it happening.
(see next comment)
(I just bought a bunch of both YES and NO in order to increase liquidity aka up the ante aka make it more worthwhile for others to participate.)

Related questions

© Manifold Markets, Inc.Terms + Mana-only TermsPrivacyRules