Genii Weblog

70 sextillion and counting

Tue 22 Jul 2003, 02:51 PM

by Ben Langhinrichs
On today, the headline reads: Astronomical: Study counts 70 sextillion stars

If you read the story, it is not until the fourth paragraph that you get a vague hint that the counting wasn't really counting at all, when the phrase "was calculated by" is used instead of "was counted by".  It is not until the fifth paragraph that the article clarifies that "the number was drawn up based on a survey of one strip of sky, rather than trying to count every individual star".  Furthermore: "Within the strip of sky some 10,000 galaxies were pinpointed and detailed measurements of their brightness taken to calculate how many stars they contained. "

OK, so first, only a survey of one small strip of sky was done, and everything was extrapolated from that.  Second, the brightness of a galaxy was used to calculate the number of stars, with no particular evidence that it wasn't twice as many dimmer stars or half as many stronger ones.  Finally, this is extrapolated not just across the sky, but "multiplied again out to the edge of the visible universe", whatever that means.  I am not a scientist, but the likelihood of serious error has exceeded any possible value by this point.  On top of that, studies show that most people don't read beyond the fourth paragraph of any given news story, so nobody even knew this was a wild guess at best.

"So what?" I hear you mutter under your breath.  With stars, it hardly matters, but this is exactly the way Microsoft counts seats when it claims to be dominant in collaboration/e-mail/pick-some-other-category.  Use a lot of fuzzy math; pile the estimates on estimates;  discount other reasonable interpretations, and then declare a number, with all caveats buried on a website seven clicks from nowhere.  And guess what, people believe them, just the way they believe the astronomers!

Copyright 2003 Genii Software Ltd.

What has been said:

27.1. Tom Duff
(07/22/2003 04:37 PM)

Yes, but the seat wars are over... Just ask Ed.

And FYI... I find it rather disconcerting that when I post a comment, your picture ends up right next to the comments box. I feel like I'm being watched! :-)

27.2. Ben Langhinrichs
(07/22/2003 05:21 PM)

You and I may listen to Ed, but there are a few others out there who list to Bill Gates as well.

As for the photo, I just want you to remember who you are writing to. lol.

27.3. Steven Alwich
(07/25/2003 05:58 AM)

It's not a "wild guess". I can make wild guesses. They used valid statistical techniques in this case, sampling. The important thing about statistics is that any statistic can be skewed into a favorable or unfavorable light. So when Microsoft makes these claims, they're correct. And when others make counter-claims, they're also correct.