Newcomb

[packbat]

Reposted from my Facebook:

Imagine the following scenario (a variation on the classic dilemma known as Newcomb's Problem):

About six months ago, a crack team of psychologists came up with a brilliant new device, and decided to run a curious experiment to test it. The experiment takes the following form:

1. Each subject, chosen by lottery, is provided with the money to purchase two identical plain manilla envelopes.
   
2. They and their envelopes are given free transportation to the lab, where they (but not the envelopes) fill out a survey.
   
3. They wait approximately one hour, and then are ushered into the experiment room.
   
4. In that room, they are permitted to examine three stacks - one containing twenty U.S. fifty-dollar bills, one containing twenty fifty-dollar-bill-sized pieces of blank U.S. fifty-dollar-bill stock, and one containing one thousand U.S. one-thousand-dollar bills.
   
5. An attendant removes the stack of thousand-dollar bills. They are instructed to privately place one of the remaining stacks in each of their manilla envelopes, so that they would have two apparently-identical envelopes, and then signal.
   
6. On the signal, the attendant returns with a case, which either does or does not contain the million dollars. The subject then gives either of their two envelopes in return for the case.

There is only one catch in this procedure: the case either contains blank bills or the million, as follows. If the psychologists predict the subject would return the envelope with the thousand dollars, the case contains the million. But if the psychologists predict that the subject will return the envelope with the blank paper, the case contains blank paper. And in each of the one hundred trials so far, the psychologists have always gotten it right. Everyone has either left with the thousand or left with the million.

(Edit: Well, not quite. A few clever people thought to randomize the envelopes so that they didn't know whether they lost the thousand or not. About half of them walked away with a thousand, the other half with nothing.)

The experiment is valid - it has been tested by dozens of experts in experimental protocol, sleight of hand, hypnotism, and every other relevant field. They neither coerce your choice nor switch out the million if you choose to keep the thousand.

You are in the room, with your two envelopes, and the attendant is before you with his case.

Do you give him the thousand dollars or the blank paper?

Comments

I mean, it might grow on me, but it's not doing so thus far.

[the-zaniak.livejournal.com]

Blech. I'm going to be honest, I hate it. I really liked the format of the first one. Which are we using for Game 3?

Maybe it just needs better fertilizer?

[packbat.livejournal.com]

We haven't decided. I bet if you wrote a ruleset people liked, you could get in as Head and enact it, though!

Nifty icon!

[the-zaniak.livejournal.com]

Remind me what's wrong with the original ruleset?

Swiped it off DevArt - http://Mrcrapinson.deviantart.com/art/Diablo-Suelto-71385814

[packbat.livejournal.com]

The original ruleset is very general - it's not optimized for Internet play, and indeed it works very badly for Internet play, judging by almost every other LJ nomic that was attempted.

As far as the critical differences go, I think the main thing is the deadlines and the loosening of the everyone-must-vote criterion. If I were going back towards the original version, those are the changes I'd ask to keep.

[active-apathy.livejournal.com]

That all the LJ nomics to try it died quickly and horribly; using default Nomic on LJ, your nomic dies the moment someone either doesn't propose, or doesn't vote. This is a Bad Thing - so our base ruleset was designed around post-mortems of the other LJ nomics we could find, and I think we can be fairly confident that it worked.

On the other hand, if by 'original ruleset' you mean the one for the first game: there's nothing at all wrong with it, and I'd happily start with it again.

I didn't realise that Packbat was referring to a different one, in fact.

[the-zaniak.livejournal.com]

That is indeed the one I meant!

Ah, I should have made that clear!

[packbat.livejournal.com]

Incidentally, if you're curious as to the 'default Nomic' i.e. Peter Suber's original game (http://www.earlham.edu/~peters/writing/nomic.htm), it's online.

*nodnod* Thought that might've been the case

[active-apathy.livejournal.com]

Shiny! Well, in that case, I'm all for the way we played the first game - i.e., slowly adding in complexity rather than having rules that were moderately tricky upfront.