.

Sunday, October 17, 2010

Completely off topic

From YNet:
In an odds-defying incident, Saturday night's state lottery numbers were an almost identical repeat of the lottery numbers from September 21. Ninety-five people picked the lucky combination and won.

The numbers that rolled out during a live studio broadcast this past Saturday, in lottery number 2194 of Miphal HaPayis, Israel's state lottery, were 36, 33, 32, 26, 14, 13, and an additional 'strong' number 2. Moments after the celebration in the studio, curious web surfers were amazed to notice that these same numbers happened to roll out less than a month ago; on September 21, in lottery number 2187, the winning numbers were 13, 14, 26, 32, 33, 36, and a 'strong' number of 1. The order in which the numbers were picked was reversed.

According to Zvi Gilula, a professor of statistics at the Hebrew University, getting all seven lottery numbers exactly correct, under normal circumstances, is one in 18 million. Guessing six numbers correctly, excluding the 'strong' number, is one to 2,250,000.

Gilula, an expert on gambling, estimated the probability of the same set of numbers being randomly picked twice a few weeks apart is no higher than one in 4 trillion, or 0.00000000000025.

"Usually, this is the type of numbers they use to describe the probability of life on Mars," Gilula said.
I highly doubt that the reporter accurately quoted the statistician.

If the chances of picking 6 numbers correctly is 1:2250000, then the chances that any random week's numbers would be exactly the same as the previous week is also 1:2250000. In fact, the chances that a picked set of six numbers would be identical to those of any of the previous four weeks would be roughly quadruple that chance, or 1:562,000.

The chances that a specific set of six numbers would be chosen a few weeks apart is probably closer to one in a trillion (if I did my math right), or perhaps he was saying the chances that the two sets of numbers would appear in exact opposite order a few weeks apart was one in 4 trillion.