To determine the due date, they use something called "Naegele's rule", which basically sets your due date 280 days after the beginning of your last menstrual period (LMP). Presumably, this is because the best science mankind has to offer has determined that human gestation lasts 266 days on average, and the best estimate of the commencement of gestation is 14 days after the beginning of the last menstrual period -- bing bam boom, 280 days. Of course, they tell you up front that few women give birth on their due date but, short of having a fortune teller at your baby shower, it's the best intel you've got.
Science has had a long time to work on this problem. Undoubtedly, one assumes, scientists have conducted extensive empirical research and observed that the median pregnancy lasts 280 days. That is, of the women who don't give birth on the exact date, 50% would be earlier and 50% later than that. There should be something like a normal bell curve around that date, but maybe leaned to the earlier side (since there should be more babies born 3 weeks early than 3 weeks late).
Have you guessed where I'm going with this? Let's check in with Wikipedia and see what they have to say about good ol' Naegele's rule:
Franz Karl Naegele was born July 12, 1778, in Düsseldorf, Germany. In 1806 Naegele became ordinary professor and director of the lying-in hospital in Heidelberg. His "Lehrbuch der Geburtshilfe," published in 1830 for midwives, enjoyed a successful 14 editions.
What the what? The best estimation technique medical science has to offer was developed two hundred years ago? Perhaps this is because it is so accurate that it just never needed to be updated? Nope, that's not it.
Studies of uncomplicated spontaneous-labor pregnancies have shown that this assumption leads to due dates that are premature, relative to the median.
It turns out that the method in common use is kind of bogus, like using the Farmer's Almanac to choose a date for your outdoor wedding. Given how easy it would be to study human gestation length, and how relevant it is to, you know, humans, you'd think there would be extensive data available. There isn't. In fact, there have been very few studies, and these studies produce different results based on time, place, race of the mother, and other factors.
What we do know is that the best data available does not suggest that your due date is 280 days from LMP. 288 days is more likely to be the median for the first time mom in an uncomplicated pregnancy.
It gets worse. Remember that they estimate the date of conception based on your LMP, but that in itself introduces a lot of room for error. Very early ultrasounds may be even more accurate at placing gestational age than LMP, but doctors don't usually bother using the ultrasound to set the due date if it's within a week in either direction of the LMP version. But based on our early ultrasound and the 276 day gestation implied by scientific knowledge obtained in the last 200 years, we should be expecting the baby not on January 5, but on January 18th.
That's a pretty big freaking difference.
Fortunately, V-Train and I do have an alternate method with nearly as much scientific rigor as Naegele's rule: the fortune teller predicted that baby DJ will arrive in the wee hours of January 8, measuring a slight 17 inches long but weighing a healthy 6 lbs, 2 oz. And that, my friends, is as good a guess as any.