Odds of Twins Being Born in Different Years

This cute story is going around Facebook. TLDR: one twin was born shortly before 2013 ended, the other shortly after 2014 began.

A friend has asked “What are the odds of this happening?”

I’ve come up with some relevant data (from 2011), but I can’t crunch the numbers, mainly because I suck at math.

Here’s the relevant data:

Number of twin births (that is, number of times twins are born, not the number of twins born) in the US (2011): 131,269.

Total number of births (multiples included) in the US (2011): 3,953,590.

Cite for two data points above: http://www.nomotc.org/index.php?option=com_content&task=view&id=66&Itemid=55

Average interval between birth of twins: 15 minutes (for simplicity’s sake; cursory googling has revealed numbers all over the place).

Number of 24-hour periods in a calendar year: 365 (rounded down for simplicity’s sake).

So break out your calculators and tell me: what are the odds of any one American mom who is pregnant with twins giving birth to them over the course of two calendar years?

Here’s how I’d do it (without actually doing it):

There are X number of minutes in a year, or, more precisely, X times in a year when the minute changes from N to N+1. One of those times is the one we’re interested in, where N is in one year, and N+1 is in the next year.

So if you have a 15 minute interval, you have a probability of 15 * (1/X) that one of those 15 minutes is the magic one.

This ignores factors like variability in the interval length, distribution of births around the calendar and around the clock, and, you know, stuff.

If I’ve done it right, there are on average .25 twins born in each 15 period. There are 35040 15 minute periods a year. So, 1 in 140160. I’m not that great at these sort of things, but figured I’d take a shot at it.

If we assume that when the first twin is born in the final 15 minutes of a year the second will arrive next year, that gives the 1 in 35040 chance calculated by Antinor01 for this to be true of any one twin birth.

If there are indeed over 130,000 twin births a year, we should expect that typically several will qualify: 130000/35000 ~= 3.7

I think I heard regarding the reported case that the babies were born only four minutes apart. If that’s more typical than 15 minutes the odds will be much longer.

Aren’t twins more likely to be cesarean births? Casareans for twins have a smaller interval time than a regular delivery of twins would have. They’re also more likely to be scheduled ahead of time. So they’d be more likely to happen when the doctor wants to be working. I don’t think that includes new year’s eve.

But if there happened to be an unscheduled casarean late on new years eve, it would be relatively easy for the team to sort of shoot for straddling the years, just to make things interesting. So there could be a much larger than 15 minute time window that could be manipulated into a split-year birth.

My daughter gave birth to twins that were almost born on different days last month. It was an emergency C-section (everyone is OK now). The second boy was born at 11:54 PM. She was suppose to deliver them next week, so being born in different years could have happened in slightly different circumstances.

You don’t want to calculate 15 minute periods. Just how many occasions where 15 minutes separates two times within the same year. So you would do:

365 days x 24 hours x 60 minutes = 525,600 minutes in a year

Of those minutes, each value x has a value (x+15) in the same calendar year except the last 15. For example, Jan 1 at 12:01 am has Jan 1 at 12:16 am. So you would need to knock off the last 15 of those 525,600 minutes since anything after Dec 31 at 11:44pm in 2013 has a counterpart in the next calendar year (2014). So we have 525,585 possible first birth times that qualify, and 15 that are desirable. Plus, you would need to include 15 birth times from the previous year. So our chances of this happening are 30/525,585 or .0057%. Since we have 131,269 twin births each year, this should happen 7 or 8 times a year.

In reality, I would presume it’s a decent amount lower as a due date of 12/31 means you conceived around the first week of April, which doesn’t seem to strike me as particularly busy time to get busy. Additonally, I cannot imagine anyone inducing on New Year’s Eve if they could avoid it. Given that summer months are more popular for babies, I would think the chances of this happening are slightly lower.

Thanks. I figured I’d get it wrong.

No you double counted. For any given year you want to count how many first births lead to a second birth in a different year. You don’t want to add the 15 first births from the previous year. So your 7 or 8 is 3.5 or 4 which is about the same answer Xema gave.

Back in the stone age, my grandma gave birth at home to twins born with a day between them (one born almost midnight of day one, the second a few minutes past midnight on day 3. She never saw a doctor or knew she had twins. After boy no. 1 was born and she still felt really sick for nearly a day she went to the hospital and delivered the second boy.

Before that she had a daughter born the morning of Jan 1. I bet there a few twins alive born on different years.

Good point.

There’s also a fudge factor for recording the times due to the coolness factor of having births in different years. Perhaps up to a few minutes added to whatever you decide the average birth separation time is.

A birth is a little bit of a continuous process, and there’s a teeny bit of wiggle room. “Let’s call it 11:59PM/12:01AM…I think that clock’s about a minute fast…” etc etc

Same thing with the First Birth of New Year contest.

Before clicking on the link I thought it might be about the twins born in the town I live in. It wasn’t, so there’s two sets at least this year.

Cool ?

How cool will it be because twin 1 is split from twin 2 ?
“Sorry, twin 1 can do enroll, because she is 6 this year, twin 2 cannot, because she is 5 this year”… Well they may have some sympathy and ignore the two minutes difference. But the line has to be drawn somewhere. Best to keep the the DoB on the same day…

Funny you brought this up-- I was curious how this was going to play out when it’s time to enroll the little tots for school in five years. Maybe we’ll never know, maybe the parents will homeschool just to keep things simple.

I work at 9-1-1. A couple years ago we had a childbirth call and the baby
(a single - no twins here) was delivered in the ambulance on the way to the hospital. Of course it was right around midnight.

The ambulance crew normally calls for it’s run times which we keep track of. (What time were we dispatched? What time did we arrive on scene? and so on…). With this call they also asked for the time of delivery which would establish the baby’s birthday. The answer depended on which clock I was looking at. One on my computer console said before midnight. The one on my radio console said after midnight. So, yes, the clock matters.

It would be interesting to see how they developed if this was done.

Shouldn’t be an issue - I think most schools set their cut-off date to around August 1 (one cite). Now, of course you would have the same issue here for twins on July 31/Aug 1 - but I assume most schools would waive the one day difference, or the parents would hold the ‘older’ twin back one year to get them in the same class.