## And The \$64 Question Was …

I ran across something interesting — I always do, but this was something I wasn’t looking for — in John Dunning’s On The Air: The Encyclopedia of Old-Time Radio, which is about exactly what it says. In the entry for the quiz show Take It Or Leave It, which, like the quiz shows it evolved into (The \$64 Question and The \$64,000 Question) asked questions worth amounts doubling all the way to \$64. Says Dunning:

Researcher Edith Oliver tried to increase the difficulty with each step, but it was widely believed that the \$32 question was the toughest. Perhaps that’s why 75 percent of contestants who got that far decided to go all the way, though only 20 percent of those won the \$64.

I am a bit skeptical of those percentages, because they look too much to me like someone, probably for a press release, said something like “three out of four contestants go all the way” and it got turned into a percentage because of the hypnotic lure that decimal digits have on people. However, I can accept that the producers would have a pretty good idea how likely it was a contestant who won \$32 would decide to go for the jackpot, rather than take the winnings and go safely home, since that’s information indispensable to making out the show’s budget. I’m a little surprised the final question might have a success rate of only one in five, but then, this is the program that launched the taunting cry “You’ll be sorrrreeeeee” into many cartoons that baffled kids born a generation after the show went off the air (December 1951, in the original incarnation).

It strikes me that topics like how many contestants go on for bigger prizes, and how many win, could be used to produce a series of word problems grounded in a plausible background, at least if the kids learning probability and statistics these days even remember Who Wants To Be A Millionaire is still technically running. (Check your local listings!) Sensible questions could include how likely it is any given contestant would go on to the million-dollar question, how many questions the average contestant answer successfully, and — if you include an estimate for how long the average question takes to answer — how many contestants and questions the show is going to need to fill a day or a week or a month’s time.