Random Phone Call From Pollster Saves Woman’s Life
A New York City woman is likely alive today only because of a random phone call from a pollster:
Bobby Berlin was going into diabetic shock in her Upper West Side apartment when she received a call from a Marist College student conducting a public opinion poll about Mayor Bloomberg.
When Berlin answered the phone, the Marist student on the other end of the line sensed something wasn’t right.
“Something just sounded off,” he said. “It was just really heavy breathing and panting.”
He called in his supervisor Daniela Carter, who asked Berlin if she was OK.
“No,” Berlin said.
Carter stayed on the line and called 911. Responders determined the address Berlin had given authorities was incorrect, but the FDNY was able to track down the right address using her phone number.
“The man from the ambulette said I would have died during the night,” Berlin said later.
“It had to be fate,” said Carter. “It just had to be. Because what are the odds? What if we hadn’t called?”
Daniela Carter and the Marist pollster paid a visit Wednesday to Berlin, who wanted to thank them in person.
Berlin hugged the two of them, telling them, “I owe my life to you.”
And the most amazing part of the whole thing is this:
The computer program used for Marist polling makes up random phone numbers using New York City area codes.
Right now there are five area codes covering New York City. The odds that they’d get this phone number at that exact time? Too high for this non-math guy to contemplate.
Or, you could say “they’ve been telephone polling for 40 years, a person in crisis was going to turn up at some point.”
In other words, the odds were not that one call would be to her, it is the odds that a million calls would find someone like her.
(Though from Ms. Berlin frame of reference, she was very, very, lucky.)
Fair point. Indeed, it’s possible similar things have happened before but haven’t been publicized like this case.
It is good PR ;-), “This is the poll company, just calling to see if you are OK …”
@Doug Mataconis:
Why do you think they always have that little “no response” wedge in every poll?
@Doug Mataconis: It’s the grass-on-the-golf-course effect: odds that a blade of grass will be hit with a golf ball are very small, but some blades of grass are clearly going to be hit.
Just doing some back-of-the-envelope calculations here, if you’ve got 40 years of polling ~500 people per day and maybe 10 organizations doing it, that’s about 70 million phone calls, or 4 phone calls per minute per day. (These are clearly clustered during daylight hours, but that complicates the math a bit.) Just consider heart attacks: there are about 3500 heart attack deaths per year, so over the last 40 years there have been at least 2x that number of heart attacks, or about 78 million heart attacks, which works out to about 4 people having a heart attack in every minute of every day.
The odds of having any one of the pollsters calling any one of the heart attack victims in a given minute: 16/280 million. Multiply by the number of minutes (~21 million) and we expect about 1 call over those 40 years to have been from a pollster to a person having a heart attack that initiates in that minute.
Note that this doesn’t even account for durational effects: the poll takes about 10 minutes, the heart attack can probably be identified for about 10-20 minutes…all multipliers. And diabetic shock — which is less common than a heart attack — occurs over a much larger timescale.
(Sorry to ruin the “Wow, that’s amazing!” effect…)
@ptfe: Wow, that back of the envelope calculation is amazing!
_couldn’t resist_ 😉