Hi dumpli,

Although picking numbers is usually a strategy reserved for the Quantitative section, it can also be useful on number-centric CR Inference questions.

Imagine that 200 books were sold in Michigan in 1980 (way to low to be realistic, but who cares!). That would then require 268 books to be sold in 1989.

Now, imagine half of those books are fiction books. We started with 100, and only had a 16 percent increase, for a total of 116 fiction books sold in 1989.

We also started with 100 nonfiction books in 1980. In 1989, since ALL books are either fiction or nonfiction, and 116 out of 268 were fiction, the remaining 152 were nonfiction--and increase of 52%!

Once we've picked these numbers, the reason why it must always be this way becomes clear. In order to average 34% change overall, the two groups must balance out. If one group is below 34% increase, then the rest of the overall change has to come from somewhere! Thus, we can predict and select the correct answer.

Hope this helps!