A Math Problem; Probability of Winning a Radio Promotion

This may sound like more of a General Question, but my last math problem provoked a healthy discussion in MPSIMS, so I’ll start this one here and see if it sticks.

The Cubs radio network/WGN radio have this promotion called the “Cub Foods Attendance Game.” Basically they ask for the seventh caller, who then must guess which of the two Cubs play-by-play announcers will be closer in guessing the attendance at tonight’s game, either Ron Santo or Pat Hughes. During the game, Ron & Pat will write their guesses. Whoever is closer wins, and if the lucky caller guessed the right guy, he/she wins two tickets to an upcoming Cubs game. I won a couple of weeks ago, by guessing Ron Santo.

The kicker is that all of the winners get entered into an end of the year drawing for $10,000. The worst-case scenario is that 162 people guess correctly (162 games), thus making the odds of winning the $10,000 1 in 162. Best-case scenario is that I’m the only winner and my odds of winning the $10,000 is 1 in 1. Of course, neither option is going to happen, so can we determine what the actual odds are likely to be?

Randomness alone would say that contestants are likely to pick the right guy 50% of the time (since they are choosing from two guys), making the odds 1 in 81. BUT, randomness would only rule here IF both men (Ron and Pat) were closest in guessing the attendance in equal numbers. As it stands, however, Pat Hughes tends to guess correctly about three times for every two times Ron does. IOW, Pat dominates by a 3-2 margain.

However, contestants in the game seem to pick Ron or Pat more or less 50/50; if anything they favor Ron slightly, perhaps out of pity.

So, how many people are likely to end up in the end of the year drawing, given a 50/50 likelihood that contestants will pick either guy, and given Pat Hughes’ 3-2 dominance in the game?

Assuming: for any given caller, that caller is equally likely to choose Ron or Pat.

Assuming further: for any given game, the odds of Pat being closer are 3 in 5.

So, in any given instance of the game, fans have a 3/10 chance of picking Pat and being right, and a 2/10 chance of picking Ron and being right. 2/10 + 3/10 = 5/10 = 1/2, so on average we’ll have 81 winners total over the season.

You’re among those winners, so your odds of winning the ten grand are 1 in 81.

It may seem odd that the probabilities “balance out” to 50/50, but it makes sense statistically. 3/5 (Pat’s chance to pick accurately) * 1/2 = 3/10, and 2/5 (Ron’s chance to pick accurately) * 1/2 = 2/10. The fact that people aren’t picking intelligently is compensated for by the fact that the people who pick Pat are more likely to win.

Just for fun, let’s assume that people favor Ron, and will pick him 3/5 of the time. Then we have:

Chance to pick Pat and be right = 2/5 * 3/5 = 6/25
Chance to pick Ron and be right = 2/5 * 3/5 = 6/25

6/25 + 6/25 = 12/25, which does not reduce, but is very slightly less than 1/2.

It still comes out to 50 percent of them.

Pick Pat and Pat right is 1/2 * 2/3 = 2/6
+
Pick Ron and Ron right is 1/2 * 1/3 = 1/6
= 3/6 = 1/2.

It doesn’t really matter which announcer picks right more often, if people choose the announcers 50-50 then it will end up at a 50% expectation.

say X is fraction announcer1 gets right. then 1-x is fraction ann2 gets right.

Pick ann1 and ann1 right is 1/2 * x = x/2
+
pick ann2 and ann2 right is 1/2 *1-x = 1/2 - x/2= x/2 +1/2 -x/2= 1/2

Hehe screwed up my lines.

the last “=x/2 +1/2 -x/2 = 1/2” should be on a new line as it is adding up the two conditions.

My guess: Probability caller picks Pat and is right: 1/2 * 3/5 = 3/10
Probability caller picks Ron and is right: 1/2 * 2/5 = 2/10

So probability of predicting the correct predictor is 3/10 + 2/10 = 1/2.

Makes 81 contestants in the final drawing.