> If you dig into the data, you'll see it's closer to "100 marriages this year, 50 divorces -- omg half of marriages end in divorce!" This is only true if there

> If you dig into the data, you'll see it's closer to "100 marriages this year, 50 divorces -- omg half of marriages end in divorce!" This is only true if there were zero existing, previous marriages before this year. Ok, but your objection is only true if there were zero divorces in previous years. If the divorce count is half of the marriage count for the year every year, then it is indeed half.