As a physicist with a blog, I am contractually obligated to do a post on the CDMS almost-a-result. This is that post.
The short version: they expected at most 0.8 events (that's total events, not events per day, or anything-- this is a whole community built on detecting nothing at all), and got 2, with maybe a third that was close to making the cut, but didn't. I think Joe Fitzsimons on Twitter summed it up best, writing:
Isn't that the least informative number of events possible?
It's more events than expected, but not enough to really be meaningful. The probability of this level of signal occurring by chance is around 23%, which is way too high to be significant, but low enough to be tantalizing.
The big live presentation conflicted with SteelyKid's pick-up from day care, so I didn't watch it. If you'd like to simulate seeing the data presented live, there are liveblogging posts you can read at your leisure. If grumpiness is more your thing, Peter Woit has you covered.
Zero would be less informative.
In order for that third event to make it into the candidate box, the expected background would have increased to 1.7 events. It's a pretty robust "about twice the expected background" result. One of the two candidate events also has a somewhat glitchy signal in terms of its timing vis a vis a candidate event's expected timing. It's possible that they're seeing one glitch and one background event.
rehana said: "Zero would be less informative."
Not really. It would give stronger exclusion limits.