Wednesday, January 6, 2010
Saturday, January 2, 2010
Holiday Reading
Who would have known that reading a book by one of my favorite authors would provide an example of sampling, expected value, and confidence intervals in the first few pages. It was a great read. The main character is a computer whiz and an artist, who makes his living as a professional criminal. I thought that I had read all of Sandford's books, but must have missed this one.
Excerpts that reference the AP Statistics content are below.
It works like this: the casino advertises (and reports to the tax authorities) a given return on the slot machines. If that return is even a little lower than the rate reported, the income increases sharply. That is, if you report that your machines will return 95 percent to the players, but you really only return 94 percent, and a million bucks a night goes through the slots, you're skimming $10,000 a night. In a few months, that adds up to real money.
Exactly what we did was, we dropped dollars - and quarters and nickels - into slot machines and counted the return, and then ran the results through a statistics package. We wanted 98 percent confidence that we were less than half of a percent off the true return. We therefore needed to take a large random sample of machines and had to run enough coins through each machine that we'd get a statistically accurate return on each.
I'd chosen the target machines the first night, using a random numbers program in the laptop I carried. We'd been at it ever since, dropping the dollars, quarters, and nickels, doing the numbers at night....
Reference: Sandford, John 2003. The Hanged Man's Cross. New York, NY: G. P. Puntman's Sons.
Excerpts that reference the AP Statistics content are below.
It works like this: the casino advertises (and reports to the tax authorities) a given return on the slot machines. If that return is even a little lower than the rate reported, the income increases sharply. That is, if you report that your machines will return 95 percent to the players, but you really only return 94 percent, and a million bucks a night goes through the slots, you're skimming $10,000 a night. In a few months, that adds up to real money.
Exactly what we did was, we dropped dollars - and quarters and nickels - into slot machines and counted the return, and then ran the results through a statistics package. We wanted 98 percent confidence that we were less than half of a percent off the true return. We therefore needed to take a large random sample of machines and had to run enough coins through each machine that we'd get a statistically accurate return on each.
I'd chosen the target machines the first night, using a random numbers program in the laptop I carried. We'd been at it ever since, dropping the dollars, quarters, and nickels, doing the numbers at night....
Reference: Sandford, John 2003. The Hanged Man's Cross. New York, NY: G. P. Puntman's Sons.
Subscribe to:
Posts (Atom)
