Vocabulary & Concepts

Law of Large Numbers and Loss Sharing

MR. WATSON: Tell me if you agree with this statement, guys. The more times I attempt a certain thing, the more predictable the outcome. True or false?

STUDENTS: True.

MR. WATSON: Let's be more specific. How old are you?

STUDENT: 36.

MR. WATSON: We want to know how many 36-year-old men are going to die. So we watched a group of 1,000 36-year-old men and a group of 1,000,000 36-year-old men. We watched this group for 100 years. Which prediction would be more accurate? Watching a group of 1,000 men or a group of 1 million men to determine how many of them would die? The larger group. The million man group.

MR. WATSON: Now if we watched them over a ten-year period of time or a 100-year period of time, which would give us the more accurate prediction? 100 years would be more accurate.

MR. WATSON: How old are you?

MAN: 33.

MR. WATSON: How old are you?

WOMAN: 37.

MR. WATSON: We couldn't mix a 33-year-old man with a 37-year-old woman in these tables. Men have their tables, and women have theirs. More men age 37 died than 37-year-old women. These are called mortality tables. And these mortality tables range from age zero all the way to age 100. Keep that in mind. At age 100, everyone is presumed to be dead by then. Why? Because the number of people who live beyond age 100 is not a statistically significant portion of the population.

MR. WATSON: What if we want to find out how many people became sick or disabled? We'd watch a large group of people at different ages over a long period of time. Does that make sense? That's called a morbidity table.

MR. WATSON: So a mortality tables tells us how many people can be expected to die. A morbidity table would tell us how many people could be expected to get sick or become disabled.

MR. WATSON: More men die in any given age group than women. How about the morbidity table?

MR. WATSON: Women have a higher morbidity rate than men. But it equals out at age 50.