90%: Judge Rotenberg Center continues torturing children at least through December 31, 2012.
While I know nothing about the case, given that sentence has the same structure as ‘have you stopped beating your wife’ it may be hard to place a prediction on.
It doesn’t have the same structure at all. “No” is always the preferred response to “do you continue to beat your wife?” The preferred answer of a strict reading of “have you stopped beating your wife?” depends on whether you had been previously (which sets up the trap of the joke).
“Did they continue torturing children?” “No” implies they had in the past.
“Did they continue torturing children?” “Yes” is only true if they did so in the past and are continuing to do so.
What I meant was without assuming a value for ‘they have tortured children in the past’ (which I assume to be at least slightly controversial) you cannot give a probability to it.
Though I suppose if they have not tortured children in the past the correct probability of continuation would be 0% as it is impossible. Same as the prediction “P&¬P.” (Though realistically you’d want to incorporate your assessment of the available evidence, see my comment on kalla724′s post).
That’s why you think it’s unreasonable to accuse them of torture.
I’ll consider myself to have guessed wrong if it comes out that they were really never torturing anyone at all to begin with. I will not, however, use a euphemism when what I mean is “torture”.
I didn’t think I needed to for the same reason that I didn’t think I needed to separately claim that there are OWS protests before predicting their continuation. I thoughtthis wasn’tup for debate.
Do you think you could make a hundred predictions like “Harry Potter and the Methods of Rationality will be updated this year” or “The Winds of Winter will not be released this year” and only be wrong once? Maybe you’re right, but your confidence seems high to me. (Note that 98% resp. 96% allows you two resp. four errors.)
Yes, I really mean that high confidence. The Winds of Winter is the sixth in a series; the fifth was released in July, six years after the fourth, which was released two years after it was due and five years after the third book. The author is slipping, the books are getting longer and less manageable and the author enjoys watching football. He’s also spent a LONG time promoting his latest book so aggressively I’m just about sure he can’t have been writing for months. It’s just barely conceivable he could deliver a manuscript to a publisher in 2012, but if so, it would be late 2012, and it would be published in 2013. It essentially would be fighting the barriers of what’s possible for him to do for him to actually get it done in time for a 2012 release date. Another author might do it, but not him and not his thousand-page doorstoppers.
For a non-abandoned fic like HP:MoR, with 76 updates in 22 months, where Eliezer actually has the next chapter completed already and is just trying to do two at once, it will take a catastrophe to keep an update from happening this year. (Hmm. Given the high likelihood of a catastrophe happening, maybe I did guess too high there.)
I still think you are about 5% too high on both of those predictions, but at least you aren’t being stupid in arriving at your probabilities.
(By the way, if you are wrong, you’ve done your future self a service by writing this comment—explaining in detail your reasons is one of the few known effective tactics against hindsight bias.)
A bit off topic, but you seem to be doing this kind of thing a lot: is there any trick for calibrating high-/low-probability events? I can see how to figure out whether my 50% is 50% or 40%, but I’d need to make a lot of predictions to get a statistically useful number of 1% predictions wrong, even if my 1% is really 2% (a serious error!)
You can know that your numbers were wrong, if many of the 1-2% predictions become true. But there is no way to find out (by looking at the outcome) whether it was 1% or 2% without several hundred predictions.
Okay, sure. Thank you. Actually, you might be right. Maybe I did fail to consider certain possibilities that could keep those things from happening how I assumed. Of course, that would be evidence in favor of my other prediction:
90%: the probabilities in this post are poorly calibrated, but things I think are likely will probably happen, and the converse is also true.
How would you define “end”? Without a coherent leadership OWS cannot formally declare themselves finished. The most likely “end” for OWS will be that most of the protesters go home while a few stragglers will stick around for years.
nothing could possibly go wrong just from my making some educated guesses, right?
Well, there are potential self-fulfilling prophecy effects, but I suspect the only one for which they’re worth even mentioning is #2, and it’s not clear that such an effect would constitute “going wrong”.
Okay, this seems like nothing could possibly go wrong just from my making some educated guesses, right?
90%: the probabilities in this post are poorly calibrated, but things I think are likely will probably happen, and the converse is also true.
10%: I’ll learn to play Magic: the Gathering by 2013.
.1%: Singularity occurs before January 1, 2013.
80%: Occupy protests do not end before May.
90%: Judge Rotenberg Center continues torturing children at least through December 31, 2012.
99% There will be at least one update to Harry Potter and the Methods of Rationality this year.
99% The Winds of Winter will not be released this year.
85% George R. R. Martin will not die this year.
While I know nothing about the case, given that sentence has the same structure as ‘have you stopped beating your wife’ it may be hard to place a prediction on.
Rotenburg Center
I agree that this is torture. However, I recommend adding links for subjects which aren’t common knowledge.
Another school—uses mace rather than electric shocks
I have no strong opinion about when or whether these practices will be stopped.
It doesn’t have the same structure at all. “No” is always the preferred response to “do you continue to beat your wife?” The preferred answer of a strict reading of “have you stopped beating your wife?” depends on whether you had been previously (which sets up the trap of the joke).
“Did they continue torturing children?” “No” implies they had in the past.
“Did they continue torturing children?” “Yes” is only true if they did so in the past and are continuing to do so.
What I meant was without assuming a value for ‘they have tortured children in the past’ (which I assume to be at least slightly controversial) you cannot give a probability to it.
Though I suppose if they have not tortured children in the past the correct probability of continuation would be 0% as it is impossible. Same as the prediction “P&¬P.” (Though realistically you’d want to incorporate your assessment of the available evidence, see my comment on kalla724′s post).
That’s why you think it’s unreasonable to accuse them of torture.
I’ll consider myself to have guessed wrong if it comes out that they were really never torturing anyone at all to begin with. I will not, however, use a euphemism when what I mean is “torture”.
Point was, you didn’t make the accusation first, and then predict its continuation.
I didn’t think I needed to for the same reason that I didn’t think I needed to separately claim that there are OWS protests before predicting their continuation. I thought this wasn’t up for debate.
Do you think you could make a hundred predictions like “Harry Potter and the Methods of Rationality will be updated this year” or “The Winds of Winter will not be released this year” and only be wrong once? Maybe you’re right, but your confidence seems high to me. (Note that 98% resp. 96% allows you two resp. four errors.)
Yes, I really mean that high confidence. The Winds of Winter is the sixth in a series; the fifth was released in July, six years after the fourth, which was released two years after it was due and five years after the third book. The author is slipping, the books are getting longer and less manageable and the author enjoys watching football. He’s also spent a LONG time promoting his latest book so aggressively I’m just about sure he can’t have been writing for months. It’s just barely conceivable he could deliver a manuscript to a publisher in 2012, but if so, it would be late 2012, and it would be published in 2013. It essentially would be fighting the barriers of what’s possible for him to do for him to actually get it done in time for a 2012 release date. Another author might do it, but not him and not his thousand-page doorstoppers.
For a non-abandoned fic like HP:MoR, with 76 updates in 22 months, where Eliezer actually has the next chapter completed already and is just trying to do two at once, it will take a catastrophe to keep an update from happening this year. (Hmm. Given the high likelihood of a catastrophe happening, maybe I did guess too high there.)
I still think you are about 5% too high on both of those predictions, but at least you aren’t being stupid in arriving at your probabilities.
(By the way, if you are wrong, you’ve done your future self a service by writing this comment—explaining in detail your reasons is one of the few known effective tactics against hindsight bias.)
A bit off topic, but you seem to be doing this kind of thing a lot: is there any trick for calibrating high-/low-probability events? I can see how to figure out whether my 50% is 50% or 40%, but I’d need to make a lot of predictions to get a statistically useful number of 1% predictions wrong, even if my 1% is really 2% (a serious error!)
Are there any tricks? Base-rates/frequencies (plus Laplace’s law) and breaking down conjunctions (#2 and 3 in http://www.gwern.net/Prediction%20markets#how-i-make-predictions ).
You can know that your numbers were wrong, if many of the 1-2% predictions become true. But there is no way to find out (by looking at the outcome) whether it was 1% or 2% without several hundred predictions.
Betting time!
Well, unless your or AspiringKnitter’s bid-ask spread is too wide.
A 5% difference isn’t enough to bet on—I don’t make bets that often so gambler’s ruin becomes an issue.
Okay, sure. Thank you. Actually, you might be right. Maybe I did fail to consider certain possibilities that could keep those things from happening how I assumed. Of course, that would be evidence in favor of my other prediction:
How would you define “end”? Without a coherent leadership OWS cannot formally declare themselves finished. The most likely “end” for OWS will be that most of the protesters go home while a few stragglers will stick around for years.
That’s a good point. Maybe there’s a problem with my concept of “end” as it relates to OWS protests.
I expect them to continue to be common and happen a bunch (as much as they’re happening now or more) at least until May. After that, who knows...
Well, there are potential self-fulfilling prophecy effects, but I suspect the only one for which they’re worth even mentioning is #2, and it’s not clear that such an effect would constitute “going wrong”.