Kepler Team Needs To Take PR 101

They are using lame excuses in order to prevent other people from doing the follow ups, since they think the follow ups (confirmations) are more important than the main statistical result...

The eclipsing binary star argument is BS.

They can analyze the light curve to even figure out how large the planet is (by looking at the exact intensity drop-off profile) so they can rule out eclipsing binaries. They might get a few wrong, but as Sasselov is saying, it is the statistical signal that is the giant news. Not the confirmation of any single planet.

In short, they are saying they can't confirm the forest since there's doubt about some of the trees.

It sounds like he said, with all of these unconfirmed Earth-like planets, odds are there will be lots of Earth-like planets (uh statistically).

As we know, they need confirmation. They have to see a planet-effect twice to know and they're rightfully being careful so earth based observations are used also.

That said, they could make a release with the statistical result (come up with an estimation of how many will be confirmed plus some of their data), rather than with actual confirmed planets. Then planet confirmations can come later. Kepler should still get credit this way.

It is kind of a tricky situation.

Anyway, Kepler kicks butt. Awesome to hear lots of small Earth-like planets!

Keith wrote:

There is no hesitation or equivocation, this is a declarative statement that comes across as a fact i.e. "the Galaxy *IS* rich in small, Earth-like planets". You certainly would need a lot of confirmed "candidates" in order to make such a bold claim in public.

Not true. You could have enough data to know that in the end there will be a lot of Earth-like planets without having any single one confirmed as an Earth-like planet. He said over and over that his claim is a statistical result.

The Kepler team is working extremely hard with extremely tricky measurements. They are entitled to the time and investigative procedures to do it right. The simple truth is that it will take ~2 years at the minimum just to get the basic time-base needed to see earths in the typical habitable zone - which means mid-2011 at the earliest for many stars.

Eclipsing binaries are in fact the strongest contaminant that can fool one into false-positives. Ordinary EBs in isolation are instantly recognized as such. It is the faint EB that is blended with a much brighter foreground star that can make a signal that looks exactly like an earth. Since one further has little S/N margin in such cases, one does not even begin to have the full light-curve diagnostics that one would want to have. People have already been fooled by these in other transit surveys.

I very much look forward to the Kepler results. I trust the team to do it right. I have followed this mission for years before its launch and really don't mind waiting a year or two more for the goodies.

It is just a bunch of scientists, so I don't think you can expect much of a PR direction.

One guy talked about some very limited data he was working on, and he is allowed to do that, so no big deal.

The team as a whole is just more conservative than he is. They'll release stuff when they're confident, just like the LCROSS people did / will do.

There is probably some movement about trying to get government-funded research done more in the open, with more regular disclosures... but for the time being, I think they can talk when they want...

Keith, please remember that the results Sasselov showed in his talk were published last month by the Kepler team (http://arxiv.org/abs/1006.2799). Though I recognize not everyone reads astro-ph in the morning (especially the folks on cable news), this can hardly be considered a leak if the information has been out in public for the last four weeks.

Sasselov just had the awkward misfortune of repeating these results in a much more public setting, and people actually noticed this time.

Let's keep this in perspective here.

1) As tentative and preliminary as it is, this information has never been available before. A year ago, we didn't know within a factor 10-100 what number would be on this PPT chart. (Although everybody had a guess, mostly it was personal optimism or pessimism.) Even with the caveats, this teaser says there will be lots of Earth-like planets in the Kepler survey, and by extrapolation, in the solar neighborhood and in the Galaxy.

2) Sasselov didn't actually reveal much in this talk -- just a preliminary result for how many planets there are in one category: very short-period planets less than twice the diameter of Earth. Believe me, there's a ton of questions we still want answered. What's the distribution of planet size, semimajor axis, and eccentricity? Host star properties? What other planets are present? Is there an infrared excess, indicating a dust disk?

3) The Kepler team has a pretty good estimate of what fraction of candidates are eliminated by each follow-up test. Therefore they can give a good prediction now of how many candidates will be called planets when the journal papers roll out early next year. They can't say yet which ones will turn out real, but they can say statistically.

Contrary to the characterizations of this usually well-measured web site (and other press accounts - Science, Discovery News - that have subsequently caused a minor firestorm), Dr. Sasselov's remarks are based on the first data release that is already PUBLICLY available, and are a /statistical/ assessment of the incidence of signals consistent with rocky ("Earth-like") planets in those data. There's a large number of assumptions here, but in the presence of large statistics, the overall assessment is most likely correct. What Dimitar did /not/ do is 'jump the gun' or 'leak' an announcement of a specific, confirmed system discovery, which is what is being currently heavily embargoed by the Kepler team. In particular, a rocky body in an Earth-like orbit has not been seen and could /not/ have been seen yet, since the mission lifetime hasn't been long enough yet.

I've already seem numerous similar talks by other researchers associated with the project; it's simultaneously amusing and dismaying to see this one presentation picked out for flagellation by the blogosphere.