Monday, July 16, 2007

Here's yer sign!

This morning I got an email from a new reader commenting on my blog profile:

Like you, I am a self-professed overeducated liberal atheist. Imagine my surprise (and disappointment), therefore, to find you touting your astrological sign as a prime aspect of introduction on your web page. Do you think astrology makes more sense than God cults? (I don't.)

Naturally, I had to reply:

Let me set your mind at ease: I didn't tout my astrological sign at all. asks you to enter your birthday when you first sign up, and it automatically computes your sign and proudly displays it for you. I have no more truck with that nonsense than you do. In fact, I'm looking at the blogger options right now to see if I can find a way to turn it off, and I can't.

Let me take an opportunity to tell a funny story about astrology. It's a story I tell often when the subject comes up, but somehow it has escaped getting the blog treatment until now.

When I was but a wee nerd in college, I used to love a local San Diego morning radio show called "Dave, Shelly and Chainsaw." In fact, I loved it so much that I volunteered to do some of their web content for them, and if you poke around the site you'll still find some things I wrote (though uncredited). That experience helped me get a background in HTML, which led to harder drugs like Javascript and CGI programming, etc.

Anyway (tangent!) in addition to the three hosts, there is a guy named Chris Boyer. As far as I could tell, his main jobs were to play drum fills after jokes, make annoying comments, and get ridiculed by the primary folks.

So one day, the DSC show was graced with the presence of A Famous Astrologer. Dave, Shelly, and Chainsaw all seem to have bought into astrology hook line and sinker. Boyer was the one who was skeptical of astrology, and the others gave him no end of grief about it. They kept insisting that this lady was so good that he'd become a believer. In fact, they decided amongst themselves that the first thing they would do was make her do a reading on Boyer.

So the astrologer arrived, they explained the situation to her, and she laughed and said she would do it. She began by asking Boyer for his birthday. Then she proceeded to tell Boyer all sorts of details about his personality.

The other three were just eating it up. They were chortling and punctuating every sentence with "That's Boyer EXACTLY!" "You're getting this absolutely right!" and so on. Chris just sat there and took it like a trooper, politely accepting everything she told him.

Finally, after a few minutes, he asked, "Are you finished?" She said "That's all." Boyer calmly concluded: "That wasn't really my birthday."


  1. I saw someone tell a story of another astrology bait-and-switch on TV, though I can no longer remember the program. (Maybe it was "Bullshit"?)

    One week a professor took down the names, birth dates, and birth locations of all of his students, and told them that he would have a professional astrologer draw up a "chart" for them all.

    The next week, he passed out the readings to students in sealed envelopes, each labeled with a particular student's name.

    He had all of the students open the envelopes, read what their "chart" had to say, and then rate its accuracy. As you might imagine, many thought it was a pretty accurate description.

    Then he had them all switch papers and repeat the process. At which point everyone in the class discovered that everyone got a copy of the exact same description.

    It seems that people tend to believe you when you compliment them. :p

    (If anyone knows what program this was from, please leave a comment!)

  2. For what it's worth, if you go back into your Blogger profile and remove your birthday completely, your profile will display neither Astrological Sign nor Zodiac Year. You may have already figured this out, but I thought I'd suggest it just in case.

    Your blog is a great read, kudos.

  3. Anonymous4:53 PM

    The scenario from nfnitloop was set up by James (The Amazing) Randi and was on one of his programs debunking astrology, dowsing, Uri Geller and various other nonsense.

  4. Question: if you had 100,000 pieces of data (natal charts, which depict exact degrees of celestial objects in relation to your location on earth at time of birth, calculated by a computer), and sorted, re-sorted, dissected, and analyzed it, at what point would things seem atypical of randomness? And I don't mean just signs, like Cancer, Leo, etc, but aspects between planets. And declination. And patterns.

    For example, if you were able to find, say, out of those 100,000 charts, there were 1000 who had some specific "trait" in the natal chart. Doesn't matter the trait - let's say it's Moon and Saturn conjuction in the 12th house (this just means both moon and saturn were just above the horizon, together within a degree) for sake of argument. Then let's say you find out something specific about the people with the charts, and that a lot of these people share that something specific. Again, not that this would be there, but for sake of argument, let's say it's that their father dies or leaves the family within the first 8 years of life.

    So if you were to see that out of the 1000 who had the trait had, say, oh, 80% chance of their father dying in first 8 years of life, and the people WITHOUT the trait had, oh, < 20% chance of father dying, would that be statistically valid? Would it warrant further study? Would it even raise an eyebrow?

    My interest in astrology is not of the cosmopolitan magazine horoscopes, or fortune telling tricks, or blind reading on the radio, but rather of the data that could be mined, and seeing patterns. I want to know if there are real patterns that can be repeated, or used to predict something or a tendency.

    My wife is a professional astrologer and I'm an IT Director; a technologist. I chose to support her in any way I could. I'm learning quite a bit and have gained new respect for those astrologers who choose to view data as more important than gut feel. The pure math involved in generating a chart by hand is fun (to me) and the whole numeric system of how 12 is used and how it works for the signs is also quite intriguing.

    With the advent of the Internet and computer programs to analyze and chart, I foresee a massive database of info where we can actually parse and pivot and find backing substance to what Chinese and Babylonian astrologers "knew". But I will confess to follow the data more than the ancient astrologers.

    I now believe that in addition to our genetic DNA imprinted onto us by our parents, we also have a celestial DNA imprinted by the energies in our world around us. However, I'm only interested in the data we can accumulate about this celestial DNA and determine how it affects our lives.

    Does that count as still being intelligent? I value your opinion.

    PS: learned of your blog from Scott Adams' blog. I'm also in Austin, so thought I should read :)

  5. Whall,

    I took a data mining class as part of my graduate studies, and I'm continuing to do data mining work as part of my thesis. (If you hunt through my posts with the "grad school" tag, you'll see what I'm talking about.)

    Anyway, my data mining professor gave us an analogy about raisins in a cake. Say you have a raisin cake, and for some reason you really like raisins. You would like to cut the cake precisely in half, and do it in such a way that one half has many more raisins than the other half. Could you do this?

    If you had some kind of elaborate supercomputer that can analyze the whole cake and produce a 3D model of it, you could have the computer give you the perfect cut that maximizes your raisin results. But this is the important part: your computer model would be completely worthless for any future raisin cake you might make. The model only tells you details of the current cake, it doesn't impart any profound wisdom about cakes in general.

    A lot of data mining schemes are very similar. Even if what you have is a whole bunch of random noise, if you look hard enough you are bound to find patterns in it which appear significant. So in your case, I'd be less interested in your discovery that 80% of the people in group X had a father who died when they were young, as I would in whether you can use that to predict future results. Can you take more adults who were not in your sample, find out their sign, and guess with an accuracy better than random chance if their fathers are alive? If not, then your earlier discoveries were just noise.

    Even if you could prove that there are real correlations, that wouldn't prove that it was the stars and planets affecting your future. Lots of correlations show up which have unexpected hidden causes. For instance (another example from data mining) it turns out that ice cream sales are strongly correlated with shark attacks. Is there a connection between ice cream and sharks, then?

    Well no, not exactly. It turns out that ice cream sales are high in the summer, and people also tend to swim in the ocean during the summer. And of course, when people swim in the ocean a lot, they're more likely to get attacked by sharks. So the correlation between those events is real, but neither one causes the other; they just both happen to share a certain window of time.

    As far as I can understand, there isn't much in the way of rigorous support for astrology. That doesn't make astrologers unintelligent at all, of course. It just means that people are very good at inventing patterns for themselves, which appear to be important. And certainly I wouldn't deny that people who study astrology can easily come up with patterns that have meaning to those people themselves.

  6. Anonymous9:50 PM

    My favorite part of that response (other than that it was spot on) was the part that goes "for some reason you really like raisins." For some reason that struck me as funny. I take it you're not a raisin fan.

    Sour grapes?

  7. Anonymous12:26 PM

    The funny (funny odd, not funny ha ha) part of the James Randi demonstration was afterwards, when they interviewed the students with the identical horoscopes, many hemmed and hawed and rationalized a belief in anstrology anyway.

  8. nfnitloop,

    I believe the effect you are talking about is termed the Forer effect.

    Check out this wiki link

  9. Anonymous6:26 PM

    I think I heard you mention the email part of your post on either Nonprophets or AE, of both of which I am a big fan.

    I was trying to remove my sign from my profile too and googled: remove+astrological+sign+from+blogger+profile
    and this post came in third.

    As for the story, that was awesome! It sounds like a good glurge story.