What are the odds?
The wife likes to veg out in front of the tube, and usually clicks over to the same few shows. "Family Feud" happens to be on her list. That's the one where two hyper families compete on contests involving public opinion and the ability to think laterally about the same. They also try to out-hyper and out cheerlead their opponents, which is why it grates on my nerves.
I caught the final segment the other day, where the most hyper and lateral family gets a chance to win the Big Cash Prize, and found myself faced with an interesting statistal situation.
The whole game revolves around polls, where 100 people are asked for responses. The received answers are then put in order of descending popularity, with the most common responses on top. There is no set number of answers. "Name a popular flavor of ice cream" would probably fetch eight answers, while "Name a gender" would probably receive less than three. If in the first case I, as a contestant, answered "chocolate", and 67 people polled also answered "chocolate", I would get the 67 points.
In the final game, two members are given time to answer five questions. The first player sets the baseline. The second must provide different answers to the same questions. If the point total reaches or excedes 200, the family gets the Big Money.
I just happen to watch when a family pair got the number one responses in all five categories. I was struck because all the number one answers except one scored in the low 20s (22, 25, 23, etc.). That's when it hit me: the odds of winning this game would be very easy to manipulate.
Let's say the producers of this show are running a little short on the budget. They could easily compile five questions for the final game with a) 5 or more possible polled answers, and b) number one answers as close to 20 as possible. Increasing the number of responses spreads the points. The chances that the spread is even are miniscule. One could easily have a response over 50, and 5 remaining responses hovering near 10. However, if I remember correctly, the minimum response is 3; absolutely unique responses can be homogenized. For example, if the single answers are "refridgerator," "stove," and "dishwasher," the final response on the show might be 3 points for "kitchen appliance."
I figure all answers must have number one answers over twenty, but that is really just an assumption. What is to prevent them from throwing in a ringer or two, and making the entire game contingent on hitting the one question with the highest scoring number one? In fact, that was the case on the show I happened to see. If the first player hadn't scored 56 points for one of the questions, I doubt the other player -- with 4 out of 5 of the number one answers, could have clinched the prize. His number one responses were so close to 20 I doubt strongly the second or third responses in those questions would have exceeded the minimum.
I'm curious. I wonder what selection criteria are used to choose questions for each final game section. I've seen games where the baseline contestant scores 189 on the first five questions, and others where, even with three of the five top answers, they still fail to total 200.
I assume the laws state that the producers aren't bound to provide each game with the same chance of winning; but I would be curious to see if anyone had correlated total scores and number of top responses received with, say, the ethnicity, hair color, female to male ratio, or even attractiveness of the contestants.
That might be very interesting indeed.
Damn, am I cynical or what?