THE TIME HAD come to look inside the box. On November 5, 2009, scientists at
sixteen institutions around the world took their seats before their computer
screens and waited for the show to begin: two software programs being run by
two graduate students—one at the University of Minnesota, the other at the
California Institute of Technology—simultaneously. For fifteen minutes the two
scripts would sort through data that had been collecting far underground in a
long-abandoned iron mine in northern Minnesota. Over the past year, thirty
ultrasensitive detectors—deep-freeze cavities the size of refrigerators, shielded
from stray cosmic rays by half a mile of bedrock and snug blankets of lead, their
interiors cooled almost to absolute zero, each interior harboring a heart of
germanium atoms—had been looking for a particular piece of the universe. The
data from that search had sped from the detectors to offsite computers, where,
following the protocol of a blind analysis, it remained in a "box," out of sight.
Just after 9 A.M. Central Time, the "unblinding party" began.
Jodi Cooley watched on the screen in her office at Southern Methodist
University. As the coordinator of data analysis for the experiment, she had made
sure that researchers wrote the two scripts separately using two independent
approaches, so as to further ensure against bias. She had also arranged for all the
collaborators on the project—physicists at Stanford, Berkeley, Brown; in
Florida, Texas, Ohio, Switzerland—to be sitting at their computers at the same
time. Together they would watch the evidence as it popped up on their screens,
one plot per detector, two versions of each plot.