Irony Alert: I got an alert & hide in the Hillary Clinton group for talking about the alert stalking

[Jim Lane]

50. A bit over one percent, I think

Last edited Fri Sep 11, 2015, 03:42 AM - Edit history (2)

Assumptions:
* There are 15% Clinton supporters and 85% Other.
* There are enough DUers available when the jury is picked so that the odds don't change with successive picks. (For example, if there were only 100 eligible jurors, and the first three picks were Other, then the chance that the fourth pick is Other is no longer 85/100 or 85.0%; it's 82/97, or 84.5%. This assumption, by ignoring this effect, is probably false, but probably not by much, and getting a more exact chance would require knowing the number in the pool, which presumably varies by time of day.)
* Each Clinton supporter and each Other has the same chance of being on a jury. (Rates of Star membership, for example, might differ between the two groups. I ignore this factor because I have no way of guesstimating it.)
* We don't care about precisely four Clinton supporters on a jury; we care about four or more.

On those assumptions, I make the chance that any particular jury is Clinton-majority to be a bit above 1.2%.

DU presumably has numerous juries per day but I have no idea how many. If we have 57 juries in a given time period (a day? two days? a week?), it's about even money that one or more of them will have a Clinton majority.

Of course, the big complication is that significant numbers of Clinton supporters and of Other will vote according to their honest evaluation of the alerted post in light of the ToS, sometimes voting to hide a post they agree with or to leave a post they disagree with. "Clinton majority" and "Other majority" don't automatically determine the jury results. I'm guessing that there are a lot of spiteful alerts that get crushed by 6-1 or 7-0 and that we never hear about unless one of the jurors decides to post the results because the alert was so laughable.

ETA: Even if only four jurors were picked, your figure of 50,000 to 1 would be way off. The probability that Juror #1 is a Clinton supporter is .15. The probability that Juror #2 is a Clinton supporter is .15. To find the probability of both of these independent events occurring, you multiply, so the chance of both of the first two jurors being Clinton supporters is .0225. Similarly, the probability that, in picking only four jurors, all four will come from that 15% minority is .15^4, which equals .00050625. The probability that this will not happen is .99949375. If you want to state it in terms of odds, the odds against picking an all-Clinton jury would be .99949375 to .00050625, which is less than 2,000 to 1 (actually, about 1,974 to 1), not 50,000 to 1.

Of course, picking seven jurors greatly increases the chance of getting precisely four Clinton jurors, because they don't have to be the first four jurors picked. Then you also add in the chances of getting five, six, or seven Clinton jurors, which collectively add about a tenth of a percent to the chance of a Clinton-majority jury.