Most Beeminder goals are open-ended. But we haven't implemented that, we just default to 10 years in the future as an approximation to an ongoing goal. Which is bad and we have a bigger overhaul planned, but this is about a tiny change that might help a little.
Namely, we can just change that "now + 10 years" to the Schelling-like "2099-12-31" to make it slightly more clear that it's a proxy for "forever" or "ongoing". (Thanks to Clive Freeman for the idea!)
Of course, someone confused or frustrated to see that they're seemingly committed for 10 years (they're not) probably isn't going be any less so if it's 77 years. But also probably not any more confused.
And it's a small step in the right direction. For example, the UI could treat "2099-12-31" as a special case where it displays "this goal is open-ended; click to set an explicit end date!" We can't do that with the current "now + 10 years" scheme.
This resolves NO if we get a credible complaint about increased confusion or frustration. Otherwise it resolves YES. In case of ambiguity, I intend to resolve to the market price, assuming the market price stabilizes.
PS: If anyone objects to either a YES or NO resolution -- i.e., has a reasonable argument for why YES or NO would be incorrect -- that would count as ambiguity.
PPS: Note the clarification about weak vs strong Pareto improvement in the comments but I don't think this is turning out to matter because (not counting bugs introduced -- see the comments for that as well!) it is turning out to be a strict improvement. Now we're just waiting for any possible reports of it being worse in some way we didn't anticipate.
1,000🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ250 | |
| 2 | Ṁ1 | |
| 3 | Ṁ1 | |
| 4 | Ṁ0 | |
| 5 | Ṁ0 |
Hmm... I have no horse in this race, but I don't think the necessary/sufficient conditions for the resolution of this market fit the stated bet in the title. Receiving a new complaint might indicate that it's worse, but not receiving any complaint doesn't indicate that it's an improvement.
Here, the "yes" side seems to have a pretty big advantage since it wins in the "it didn't matter even a little" case. There's no room for the neutral stance in the success conditions, and the "it's better" side takes the win even if it turns out to just result in an entirely neutral outcome (when really it should be set up to be the opposite, since it's a claim that it will be better rather than a claim that it won't be worse).
Full disclosure: I prefer the 2099-12-31, myself, to the random date in the 30s, so this isn't an indication of a preference to not change the date; it's just an indication that I'm done my dissertation and need something else to nitpick.
@Essy Yeah, but that was in fact the spirit of the original prediction, namely, is it safe to do this / are we sure enough that nothing bad will happen? That's also what I mean by "Pareto improvement" -- not necessarily an improvement but definitely no worse at all in any way.
@dreev 👍 Just needed a title change, à la "won't be worse," to solve the mismatch, then.
Like I said, though, I have no horse in the race itself!
@dreev Oops, I didn't register the "what I meant by..." (I'd still disagree with that, just out of sheer pedantry, though, since a Pareto improvement has to be an improvement and BEING PEDANTIC IS FUN! Until people stop inviting you to their parties, anyway.)
@Essy Oh I love pedantry! Technically I think the solution here is to clarify weak or strong Pareto improvement and now I'm genuinely unsure which should be assumed if that isn't specified...
I still say the spirit of the prediction was "weak Pareto improvement" all along but I'm definitely listening if anyone made bets under a different assumption.
@MartinRandall How so? Also I should make you read https://blog.beeminder.com/pdp before answering that.