Wednesday, July 29, 2009

Only Strange If 1 in Pi Are Like You

I was on the Wikipedia looking for the exact values for various confidence intervals when I saw this for one sigma:
ratio outside CI: 1 / 3.1514871
That looks a lot like this ratio:
and in fact agrees better than 1%. This has given me my new rule of thumb:
Attribute X about you is weird if only one in pi people share X with you.
(Number of people that find this post interesting: < 1 in Pi.)

1 comment:

  1. I was just investigating this too for an application I'm writing. I'm going way overboard, but I'm trying to get statistical improbability as a geometric distance, so I've decided to use the inverse of the cdf of the normal distribution to get it: 
    public double geometricDisagreementTo(Double x){ double toMean = Math.abs(x - distribution.getMean());
    double inverseArea = 1.0d / (2 * distribution.cumulativeProbability(distribution.getMean() - toMean));
    double normalized = (inverseArea-1)/(3.15148718753437704787380015 - 1); return normalized;}problem is, 3.15148718753437704787380015 is wolfram alphas number, and Java's number seems to be slightly rounded, meaning in my testassertEquals(1.0, z_Wave.geometricDisagreementTo(-1.0));

    fails: "Expected <1.0>, but was <1.0000000000000004>"

    chance that you read this + chance that you care + chance that you understand < 1 in pi.



To add a link to text:
<a href="URL">Text</a>