Two weeks ago we posted a story on Gliese 581g - a planet that was discovered and said to be in the habitable zone of a star.   Our comment was that reduced chi-squared statistics may mean that, if there was an error, it would not even exist.

That didn't stop one of the researchers from saying, "Personally, given the ubiquity and propensity of life to flourish wherever it can, I would say that the chances for life on this planet are 100 percent. I have almost no doubt about it."  Somewhat shocking for an astronomer to a journalist who was going to repeat it verbatim.

Even more muddy now is that no one else can detect it.  At the Astrophysics of Planetary Systems meeting, a Swiss group said they could not detect a fifth planet in Gliese 581's orbit - but they used their own data.  "We do not see any evidence for a fifth planet ... as announced by Vogt et al.," astronomer Francesco Pepe of the Geneva Observatory wrote Science in an e-mail from the meeting but "we can't prove there is no fifth planet."

Obviously that's true - you can't disprove a negative.  Plus, if we had that kind of precision we would know if Gliese 581g were made of rock or marshmallow creme or had water on it by now also.  

Does it mean Butler and Vogt and crew were wrong?  Not at all, they used an 11 year data set and their study used as much care and rigor as was possible, but we are still dealing with wiggles in the motion of the Gliese 581 star at a distance of 20 light-years away.  So some caution is warranted - and a lot less hyperbole.