rmiller wrote:


At 11:08 PM 6/8/2005, Jesse Mazer wrote:
(snip)

You should instead calculate the probability that a story would contain *any* combination of meaningful words associated with the Manhattan project. This is exactly analogous to the fact that in my example, you should have been calculating the probability that *any* combination of words from the list of 100 would appear in a book title, not the probability that the particular word combination "sun", "also", and "rises" would appear.

RM: Are you suggesting that a fair analysis would be to wait until Google Print has the requisite number of books available, download the text, then sic Mathematica onto them to look for word associations linked with a target? What limits would you place on this (if any?) Or would this be a useless (though certainly do-able) exercise?

I'm saying that you have to select the possible targets before you actually go mining the data of old stories to see what's there (or at least you have to try to imagine you didn't know what was there when selecting the targets). If your choice of targets is explicitly based on what you find in the data you will get bad probability estimates, for reasons I've already explained (you haven't really responded to these arguments in any substantive way--for example, do you agree or disagree that basing the choice of target on knowledge of the data tends to lead to situations where, even if the correlations are pure coincidence, 1 out of x parallel versions of you would claim to see a 'hit' with a significance of 1 out of y, where y >> x?)




. . . Would it be fair to test for ESP. . .

We're not testing for ESP--only out-of-causal-order gestalts in popular literature that are associated with similar gestalts in literature (or national) events taking place at some future time.

Yes, I was using "ESP" as an umbrella term for any mysterious foreknowledge that can't be explained in terms of currently-known types of information channels. Substitute "foreknowledge not explainable in terms of known science" for "ESP" in that sentence (and any other sentence where I talk about 'ESP') if you like.

Or it might be explained by some of the more offbeat analytical procedures---say, involving exponential or Poisson probabilities as applied to delayed choice events.

I know what "delayed choice" means in the context of QM, but what do you mean by "applying" exponential or Poisson probabilities to delayed choice? According to our current version of QM, it is possible to prove that delayed choice experiments cannot be used to send information backwards in time--are you suggesting a modification of QM, and if so, how exactly are "exponential or Poisson probabilities" involved?



Again, my concern is that scientists are too willing to prejudge something before diving into it.

OK, but this is a tangent that has nothing to do with the issue I raised in my posts about the wrongness of selecting the target (whose probability of guessing you want to calculate) using hindsight knowledge of what was actually guessed.

As a former fed, I would wholeheartedly disagree. There is a grand tradition of avoiding analysis by whatever means are available, including "hindsight knowledge" invalidating the correlation. In other words, you shouldn't ever mine for data. Thankfully, that admonition is routinely ignored by many biostatisticians.

I'm not saying you should never mine the data, I'm just saying if you want to do an actual calculation of the probability that a correlation would happen by coincidence, you can't use this type of hindsight knowledge in selecting the target whose probability-of-happening-by-coincidence you want to calculate. I've given several examples of how this leads to badly wrong answers, and again, you haven't really addressed those examples.


If you don't want to discuss this specific issue then say so--I am not really interested in discussing the larger issue of what the "correct" way to calculate the probability of the Heinlein coincidences would be, I only wanted to talk about this specific way in which *your* method is obviously wrong.

Thank you. (Finally!!!) Whew! That sentence has validated the entire horrid exercise. May I quote you???

Is this supposed to be validating your claim that scientists prejudge issues? Note that I'm not a scientist, and I'm also not prejudging things, I'm just saying I'd rather not discuss this right now, just because I personally am not that interested in it, and also because it's a distraction from the topic that I originally brought up. If I were to use your post as a jumping-off point to talk about some totally unrelated issue like the mechanics of cumulus cloud formation, and you were not that interested in talking about this issue and wanted to get back to the topics you were originally talking about, could I use that to validate my view that you are guilty of prejudging the facts of meteorology?



Like I said before, any method that could be invented by someone who didn't know in advance about Heinlein's story would avoid this particular mistake. . .

. . .another money quote. . .

*although it might suffer from other flaws*.


This one too!!!

Uh, are you trying to mock these quotes, or use them to support some "prejudgments" of your own about my psychology? If so, please come out and say what you mean.

Jesse


Reply via email to