Recount, So Far, Challenges Statistical Probablility
No one knows statistical probability better than the folks in Vegas.
Yesterday the Las Vegas Sun reported that the recount in Florida that has narrowed the gap for Gore by as many as 1400 votes so far
does not jibe with statistical probability.
The paper reported that University of Nevada at Las Vegas economics professor Tom Carroll "began running statistical equations
Thursday on the net gains both Gore, who gained more than 2,200 votes,
and Texas Gov. George W. Bush, who added about 700 votes, have made in the recount. He found that the statistical chances for such
large and different totals to occur as a result of random glitches was less
than infinitesimal."
Carroll told the paper that the "The probability of being struck by lightning is about one in a million. The same person would have to be
hit by lightning 30 times to compare with what we've seen in this
recount."
Carroll based his computations on the notion that ballots in Florida "were the result of computer or human mistakes."
He added that "A recount is a 50-50 proposition, so statistically speaking, making up something like 1,700 votes is highly unlikely,"
Carroll said. "For this to have been just random error is statistically unlikely. It wouldn't be unlikely to see some changes in the number of
votes during a recount, but the differences should come much closer to canceling each other out."
Carroll is a Democrat
The paper reported that University of Nevada at Las Vegas economics professor Tom Carroll "began
running statistical equations Thursday on the net gains both Gore, who gained more than 2,200 votes,
and Texas Gov. George W. Bush, who added about 700 votes, have made in the recount. He found that
the statistical chances for such large and different totals to occur as a result of random glitches was less
than infinitesimal."
Carroll told the paper that the "The probability of being struck by lightning is about one in a million.
The same person would have to be hit by lightning 30 times to compare with what we've seen in this
recount."
Carroll based his computations on the notion that ballots in Florida "were the result of computer or
human mistakes."
He added that "A recount is a 50-50 proposition, so statistically speaking, making up something like
1,700 votes is highly unlikely," Carroll said. "For this to have been just random error is statistically
unlikely. It wouldn't be unlikely to see some changes in the number of votes during a recount, but the
differences should come much closer to canceling each other out."
Carroll is a Democrat. |
