They certainly don't make women or men like they used to. According to entries in some of the family records I have seen recently, women were having children in their late 60s and 70s and men were fathering families at over 100 years of age! If you can believe the records, many of these ancestors were married during early childhood and still managed to produce children into old age. I must come from remarkable stock with children born more than 100 years after the 50 year old wife married the three year old husband.
Do you get the picture? Have you looked critically at the dates in your own database? How many of these dates create improbable or impossible situations for your ancestors? I find this to be so common, that anytime someone brings me a pedigree chart to look at on disk, I automatically look for improbable dates for their ancestors. Interestingly, I find problems with dates in nearly every pedigree I look at (including my own). Commonly, a child will be shown with a birth date before the birth date for one or both of the parents. This isn't rocket science, it is simply grade school addition and subtraction.
The solution seems too simple. Do the math. Go through the pedigree and look at the dates. Were too many children born in the same year? Do the ages that events were supposed to happen add up? Were your American ancestors married in their childhood or having children at an advanced age? All of these indicate either the dates are wrong or the people are misidentified. Surprisingly, I find very few typographical errors. The errors in dates usually reflect basic problems with the entries not just transposed numbers.
This type of problem is especially prevalent in copied or contributed family trees. People are so happy to "find their ancestors" they fail to look at what they find in a critical way.
Sometimes the errors reflect a more serious problem; the grafting on of the wrong family line. This occurs primarily when people assume that a person of the same name as their ancestor, is their ancestor without examining the family members or dates and places to see if they are a reasonable match.
OK, I need to mention calendar changes. If you are fortunate enough and go back far enough in history, there are a differences in dates due to calendar changes. Here is a good summary of the problem from the Connecticut State Library. Most of these are not significant enough to create the types of errors I have mentioned, but they may be significant in identifying the right ancestor from two or more individuals with similar names.
Other problems caused by poor math include losing children from families and skipping generations. One of my ancestors who must have been "of age" in 1680 could not be the same ancestor having children in 1720. Well, it is possible, but unlikely. With these dates it is more probable to begin looking for a skipped generation, perhaps a son with the same name as his father.
We could chalk up dating problems to sloppy research, but unfortunately, it happens in the best of genealogical pedigrees. If you are looking at a pedigree, or part of a pedigree, you haven't viewed for a while, just start adding up and subtracting the numbers and see what happens.