Better Than Luck

There is an activity I like to do with my classes when they are in the probability unit. I start by holding up my brown paper bag and telling them that there are colored tiles in the bag. I pull out one tile of each color to show them as I say there are red tiles, blue tiles, and green tiles. And there are twenty tiles total in the bag. Can they now predict how many of each color are in the bag? At this point, some try, but I point out they are just making random guesses. I do encourage them to write down their guess to see how lucky they will be.

Next, I go around the room letting each student pull out one tile from the bag without looking. They must put it back in the bag once we have seen what color it is. Another student keeps a tally on the board. We do this several times to get some data. Then we proceed to turn the data into a fraction of how many picks were of each color. The fraction is then turned into a decimal, and then a percent. Next, we use the percent to make our (educated) guess about how many of the tiles are red, blue and green. For example, if 25% of our picks were blue, we would guess that there are 5 blue tiles in the bag because 5 is 25% of 20.

After figuring out all of our guesses, known as experimental probability, and discussing our final hypothesis and the process used to make it, we are ready for the “big reveal”. We examine how close we came and most always discover that statistical data based upon probability is better that just plain old luck!