The mathematics of overspecified fictional taste - a quantitative method (was: some generic Tina Hall thread)

JavaJosh wrote:
> Luna wrote:
> > How can you like any stories? None of them entirely fit what you want.
> > What you want is so specific, that many thousands of well written
> > stories and books won't please you. Interesting characters, a tight,
> > well-crafted plot, smooth and unobtrusive prose, and original ideas are
> > all ruined for you if someone kisses someone somewhere in the story.
> > Since the overwhelming majority, and most likely the entirety, of books
> > and stories have something in them you don't like, you can't much like
> > reading.
>
> Excellent reasoning Luna. If I might generalize a bit: the more
> specific the story criteria, the fewer matches you will get. Here's an
> interesting question: for every new criteria I add, how do the number
> of matches change? I'm going to guess, arbitrarily, that they are
> halved. That is, for each criteria you have, it disqualifies half of
> all remaining stories. (Frankly, I believe this is a very low estimate
> for the types of criteria Tina has. A more reasonable assumption is
> that each criteria reduces the remaining pool by 90%).
>
> Let us assume that there are 10 million stories of all sorts in
> existence. (This is almost certainly way too high). The number of
> stories that match N criteria are: 10^10/2^N. Now, roughly speaking you
> have 40 criteria (I'm guessing low - I didn't count). 2^40 =
> 1,099,511,627,776 (roughly 10^13). 10^10/10^13 = 10^-3 or .001.

I did this wrong. 10 million is not 10^10 it is 10^7. The odds of
finding something that matches 40 criteria goes *down* to .000001.

> In other words, there is much less than 1 story that fits Tina's
> criteria. Indeed, even with only 20 criteria you'd be lucky to find a
> single story matching all such criteria.

That's wrong, too. The number of criteria before becoming overspecified
is now about 16.

To round out this method, adding a corralary about how many criteria a
reader can reasonably have after reading N books. For example, if
you've only read 5 books in your life, it would not make sense to have
40 criteria. Indeed, I would say that one's reading criteria should
come much more slowly, at << 1/book. To do otherwise would be akin to
a child saying that peas are "verboten" when they have never had peas.

.

Relevant Pages

  • What would be thwe correct statistical procedure for this design?
    ... between true and false statements. ... stories, one of the being true and the other being false. ... X number of participants stories 15 criteria of credibility will be ...
    (sci.stat.math)
  • Re: What I really want.
    ... None of them entirely fit what you want. ... > stories and books won't please you. ... Excellent reasoning Luna. ... specific the story criteria, the fewer matches you will get. ...
    (rec.arts.sf.written)
  • Re: Vote For Top-100 Science Fiction, Fantasy Titles at NPR
    ... Criteria are not specified. ... Major omissions include Bujold's _The Curse of Chalion_ and ... _The Paladin of Souls_, McKillip's Riddlemaster trilogy ... Henderson's People stories (and yes, ...
    (rec.arts.sf.written)
  • Re: President Obama, the leader you rightards have always feared
    ... Considering the number of different conflicting stories that have ... You are lying. ... They were just to trusting of bush. ... By your own criteria, you are lying. ...
    (talk.politics.guns)
  • Re: word limits and revision
    ... I am totally convinced that most stories can be cut without harming ... When I was a kid there were still the Reader's Digest Condensed books ... They'd be a hardcover volume that now would fit one Steven ... King novel, but back then would have held two unabridged regular ...
    (rec.arts.sf.composition)