Friday, 16 November 2012

Daily Mail out by a factor of about four hundred, as per usual

From The Daily Mail:

Now that's double trouble! Woman, 25, beats odds of 170,000 to one by giving birth to twins for the second time

According to Wiki, about one birth in thirty* results in twins. We also know that some people are more likely to have twins than others, it's genetic as well as being related to age of mother, fertility treatment and diet. So if a mother has already had twins, let's assume that the chances of her having twins again is twice as high, maybe one-in-fifteen?

So the chances of any woman having two sets of twins is very approximately 1/30 x 1/15 = 1-in-450. Not 1-in-170,000, FFS.

But congrats to Mrs Power anyway, and I wish them all the best.

* Seems about right. There was a pair of fraternal twins in my class at primary school, and there was also a pair of fraternal twins in my class at secondary school. Was there a pair twins in your class at school?

15 comments:

Tim Worstall said...

Twins insurance. You can buy insurance against having twins (err, OK, to aid with the extra costs of having twins).

The price/payout is markedly different if the woman has already had twins than if she hasn't. And yes, following the money is often a very good way of working out what the odds of something are.

Mark Wadsworth said...

TW, excellent point.

According to this article, the premium for £5,000 cover is £210, which means the insurance companies know the chance is less than 1-in-24. That article also says 1-in-67 births is twins. Ho hum.

Bruce said...

Had three sets of twins in my primary school class (of about 25-30) - 1970s so all produced the old fashioned way - identical girls, unidentical girls, and a boy & girl. That's got to be quite unusual.

Mark Wadsworth said...

B, very unusual. That actually would be a 1-in-170,000 event.

Sobers said...

Wasn't this the same sort of mathematical failing that Sir Roy Meadows committed over Sudden Cot Deaths, and resulted in multiple women being falsely jailed for murdering their children?

Mark Wadsworth said...

S, yes. But were there ever any twins in your class at school (or one of a pair)?

Sobers said...

My mother is a twin. But can't remember any twins at school. It was along time ago though.....

A K Haart said...

"Was there a pair twins in your class at school?"

Yes - in a class of thirty.

Bayard said...

AFAICR there were no twins at all in my school, out of approx 600 boys, unless one half of a pair had gone to another school.

CIngram said...

A few years ago I taught a class- of only about 12 children- in which there were both twins (boy/girl), and triplets (three girls, two co-ovualar, one not).

Mark Wadsworth said...

Thanks to all. So the current tally is:

Three of us had twins in our class.

Two of us had no twins.

CI, that's a good anecdote, but if you are a teacher your sample is far too large so it's highly likely that some of your classes will have had one or more sets of twins.

Kj said...

I remember twins in the class above me, and we were around 15-20 per class. Also I have twins. Wish I'd known about insurance.

Kj said...

- and in kindergarten they are one of two pairs of twins out of about 50. Works out doesn't it.

Mark Wadsworth said...

Kj, ta. On the basis of the flimsy evidence here, it appears that not all classes of 25 kids have a pair of twins, but probably more than half. So the figure of 1-in-30 birhts = twins is probably too high but hey, it is not far off.

Kj said...

Seems so. As they mention in the Wiki-article, there's also a ethnic genetic component. around 1 in 30 in Europeans, up to double that in central africa, and very little in east asia. So correct for ethnic background in classes.