# A Synthetic *What*? Lady Who Won Lotteries Is Statistician, More

*Last big travel day. Off to the blue crammed into motorized sardine cans. I’ll begin looking at comments and email tomorrow.*

**Medical Science Advances On Bottoms**

Turns out that I was wrong yesterday and that papers *are* getting better. At least if the one put out in *Gastroenterology* by some Wake Forest docs is any indication.

Seems they wielded their beakers and petri dishes in just the right way and were able to brew up a synthetic sphincter.

Yes, just when you thought there were too many in the world already, especially down DC way, Khalil Bitar and pals found a method to grow artificial sphincters on mice. Once fully developed, these can be cut off and grafted onto human, non-mice personages in an operation that can only be described as *delicate*.

Those not up on your *Gray’s* might be interested to know this: “There are actually two sphincters at the anus — one internal and one external.” So there’s double the need for artificial you-know-whats.

**Encounter Books Puts Up Link**

Roger Kimball over at Encounter Books, which issued David Stove’s posthumous *What’s Wrong With Benevolence?* linked to my review of the same on their home page, which I consider quite an honor.

If you have not yet read the review, do so now. And if you have not yet bought the book, what are you waiting for? Also highly, hugely recommended are *Darwinian Fairytales: Selfish Genes, Errors of Heredity and Other Fables of Evolution* and especially *Against the Idols of the Age*.

If you enjoy the material on this blog at all, you will like Stove’s much more.

**Next Stop Vegas**

In what *cannot* be a coincidence, and instead must be an event deeply laden with significance—statistical, cosmological, and certainly theological–last year this same time I was also off into the skies and wrote about the lady who won the lottery four times.

I did a back-of-the-envelope and surmised that the probability at least one person wins four lotteries (of the type the mystery lady won) was about 1 in 100, which isn’t so small.

Now, thanks to the many readers who alerted me to the follow-up story, we learn that the lady who won is a—are you ready?—she is a *statistician*! Her name is Joan R. Ginther, 63, and she is formally from Stanford.

Speculation has run amok. Do statisticians actually have an algorithm that predicts lottery numbers? Are we gifted with random prescience that allows us to amass riches?

Being a member of the guild, I cannot tell you. But I can say that one of my flights stops in Las Vegas, in which I will have a two-and-a-half-hour layover. Do not thus be surprised if the “Hire Me!” link disappears from this website tomorrow.

**Update** My attempts to parlay my grub stake ($1) into a retirement-inducing fortune with the Las Vegas airport slot machines failed. I will thus still be accepting job offers.

Well, if the trip to Vegas pans out, you’ll be able to buy that bridge you’ve been promising yourself.

I predict that, if the synthetic sphincter industry takes off, the market for pig anuses will contract.

Matt:

Numbers make it a possibility that a statistician might have an edge. However, as I understand it this Prof. won with scratch tickets!! Of course, she could always be related to Whitey Bulger.

Off topic. The Washington Post carried the obituary of Paul Meier, of Kapkan-Meier survival analysis fame. The obit was about a third of a page. You come up with a fakey method of calculating a conditional probability and you become famous.

For your delight, I present the story of Mohan Srivastava, a geological statistician:

Cracking the Scratch Lottery Code

* By Jonah Lehrer

* Wired February 2011

http://www.wired.com/magazine/2011/01/ff_lottery/all/1

â€œOn my way, I start looking at the tic-tac-toe game, and I begin to wonder how they make these things,â€ Srivastava says. â€œThe tickets are clearly mass-produced, which means there must be some computer program that lays down the numbers. Of course, it would be really nice if the computer could just spit out random digits. But thatâ€™s not possible, since the lottery corporation needs to control the number of winning tickets. The game canâ€™t be truly random. Instead, it has to generate the illusion of randomness while actually being carefully determined.â€

Another practical use of a good statistics education.

Well if you couldn’t make your fortune with a hundred pulls of the lever you might as well give it up, anyway.

I once worked with an engineer with a strong interest in probability. He would play the lotto, but only when he calculated that the payback was greater than 100%, which he determined could happen when no one had the winning numbers and the pot rolled over to the next week. (The payback calculation was complicated by the fact that there could be multiple winners who would then split the pot, but he felt confident in his calculations.) I remember asking him why he did this when the odds of winning were still so very, very low. He replied “Whenever the payback is above 100%, you should always bet your brains out.”

I bet Ms. Ginther used this approach, too, and I bet that after her first big win, a least, she bet seldom and very big when she liked the payback percentage.

For some reason I can’t really explain, I find it annoying when a group of people get together and buy a single ticket and it wins. There was a group at work doing this, and they would make a big deal out of it and post a copy of the ticket on a wall in their cubicle area. One time when I knew members of that group would see, I walked over to the copy of the ticket on the wall and conspicuously copied down all the numbers. Of course they asked me what I was doing, since I wasn’t in on the ticket, and I told them that I was going to play that number ten times. For some reason they never posted another ticket on the wall.

When the lotto pot gets really big from all the un-won rollovers, I will buy a ticket. But then I’ll usually buy the same numbers at least twice, so that if I win my portion of the pot will still be very big even if I have to split it. People tell me that this is stupid, odds-wise, but I don’t know enough about probability to calculate it out one way or the other.

Ms. Ginther is formally from Stanford, which leads me to speculate: From where does she informally hail?

Milton, I’m no statistician, but I think your reasoning is faulty regarding lottery jackpots and their unwon roll-overs.

For large-jackpot lotteries, the chance of winning the grand prize is about 1 in 135,000,000, Those odds never change. In week 1, when the jackpot is around 15-20 million dollars, the odds are 1 in 135,000,000 that you’ll win the jackpot. Week 20, when the jackpot is fat with all the rolled-over, unwon amounts, the odds are still 1 in 135,000,000.