Woman who wins lottery 4 times under media microscope, investigation

[AZRedhawk44]

http://www.dailymail.co.uk/news/article-2023514/Lucky-woman-won-lottery-times-outed-Stanford-University-statistics-PhD.html

Yeah, she's a PhD.  Yeah, she knows stats and math.

Even if she DID figure out the mystical algorithm used by TX to distribute the winning tickets, I contend it is the State's fault for USING the SAME algorithm for over 15 years.  Or using an algorithm at all.

Dumb bureaucrats, catering their stupid lotteries to the poor unwashed masses too dumb to realize that the lotto is a revenue raiser rather than "a great equalizer," deserve to get sideswiped repeatedly by someone savvy enough to take 'em for all they're worth.

The only fault I could find with this woman is if she had some sort of inside information from someone in the dotguv lotto commission.

[Boomhauer]

The only fault I could find with this woman is if she had some sort of inside information from someone in the dotguv lotto commission.

This. If she's figuring it out by using math on her own, without insider info...SO WHAT? Good on her...

[CNYCacher]

On scratch-off cards though?

[mtnbkr]

...poor unwashed masses too dumb to realize that the lotto is a revenue raiser rather than "a great equalizer,"...

Who thinks that?  I've never met anyone who thought it was some kind of equalizer.  I have met many too dumb to realize their chances of winning are very tiny, even if they purchase 20+ cards at a time.

On scratch-off cards though?

Card counting.  There are set numbers of winners in each roll and she figured out the system.  Not sure how she determined when to buy hers though.

Chris

[birdman]

This has been done before in both Canada and the US.  It has to do with a statistical error in "randomizing" the wins vs the serial numbers of the cards.  Once that can be determined, you simply buy one card at a variety of stores until you find the right range of SN's, then buy the roll.  It's happened enough (usually games that have been around for a while) that it was even the plot of an episode of Numb3rs.  It's also not illegal.

[Tuco]

Not sure how she determined when to buy hers though.

That's the theme of a decent article in the latest "Harper's" if you're interested.

According to the rag, the state doesn't design the algorithms, one of a few national lottery consultants design them.

[birdman]

Increasing your win fraction based on math to make something profitable isn't illegal, just like card counting isn't illegal--the cool thing is, unlike a casino who can ask you to leave, the state can't prevent someone from playing the lottery...good on her! 

IIRC the most common exploitable information has been scratchers that have numbers printed on the scratch-off material itself--and then statistically exploiting a non-randomness in the correlation of these numbers to the underlying "match these" (or whatever game it is).  The problem with exploitation of that error is it only allows you to determine if a card is likely to be a winner--not determine the if a winner is present in a ticket you don't have.  Fortunately, due to customer protection laws, most states allow you to return (either exchange or for cash) unscratched tickets--so you buy 100, or 1000, and return those not likely to be winners (stupid state).  This has been done numerous times. 

Now people say "just don't make those games"--problem is, the type of "game" that is vulnerable to this exploit is also the most popular type--the "poker" or "tic tac toe" or other types that give the "user" (face it, it's a drug) a feeling of both sequential game-play (not just scratch and randomly win), and the illusion of being able to affect the outcome through choice (which isn't true, but hey, if you already are taxing people who are bad at math, why not really run up the profit margin by addicting their addled brains).

Other more complex (and profitable) methods would allow you to exploit any weaknesses in the distribution algorithms, but require far more samples, and are extremely difficult without insider info.

Anyway, I go with "go her", "lotteries are a tax on people who are bad at math", and "dumb state". And leave with this interesting opinion--lotteries are one of the oldest ways to appease the unwashed masses--they are bread and circuses--by imposing an extraordinarily regressive tax on the poorest, by giving them a hope in pure chance that the state will make all their wishes come true--while "doing a good thing by helping {schools, etc}" it's just a social control method--why do you think winners are so publiscized, (well, the poor ones), and the wealthy winners (I know a family with millions who has won 250k TWICE) or the long term results of winner are never shown (it's like behind the music on crack).

[CNYCacher]

face it, it's a drug

QFT

Some of the users are pretty sad too.  They act exactly like addicts.  Buy the tickets and they can't even get out of the store before they have to hunch over and use.  Hell you are lucky if some of them can sidestep the cash register first.

[HankB]

Not going to go into detail, but I like contests that have exploitable loopholes.   

. . . "lotteries are a tax on people who are bad at math" . . .

I generally agree with this sentiment - though I will sometimes break down and buy a ticket when the jackpot is big enough that they start talking about it on TV; usually by then, the jackpot is larger than the odds against winning.

(Simplistic argument, as it doesn't consider taxes, reduced payout with cash value option, or the possibility of multiple winning tickets, the holders of which will  have to share the jackpot.)

But I figure for a day or two - until they draw SOMEONE ELSE'S numbers - I can dream big.

And besides, my total wagers are probably on the order of $20 - $30/year . . . maybe less. I can afford that. 

[zxcvbob]

Not going to go into detail, but I like contests that have exploitable loopholes.   
I generally agree with this sentiment - though I will sometimes break down and buy a ticket when the jackpot is big enough that they start talking about it on TV; usually by then, the jackpot is larger than the odds against winning.

The jackpot is sometimes larger than the odds they say on TV.  Calculate the probability yourself and see what happens

[Jamisjockey]

The jackpot is sometimes larger than the odds they say on TV.  Calculate the probability yourself and see what happens

You always have the same odds of drawing a winning ticket.  When more people play, the odds of the jackpot being split are higher. So when there is a huge jackpot and everyone buys a ticket, you the odds of the pot being split are higher.

[zxcvbob]

You always have the same odds of drawing a winning ticket.  When more people play, the odds of the jackpot being split are higher. So when there is a huge jackpot and everyone buys a ticket, you the odds of the pot being split are higher.

They'll say something like the "74000000 to 1" when the jackpot is $125 million.  So you think the odds are finally in your favor because of all the previous lotteries where there was no winner and the pot just accumulated.  But if you calculate it (depending on the game) the actual odds of a winning ticket might be (50^6)/2 or 1:7812500000.

I've never seen a $8 billion jackpot, have you?  If you have to account for splitting the pot, that just makes it worse.

[seeker_two]

This. If she's figuring it out by using math on her own, without insider info...SO WHAT? Good on her...

Probably violates the "Barbie" Law.....

 

[birdman]

They'll say something like the "74000000 to 1" when the jackpot is $125 million.  So you think the odds are finally in your favor because of all the previous lotteries where there was no winner and the pot just accumulated.  But if you calculate it (depending on the game) the actual odds of a winning ticket might be (50^6)/2 or 1:7812500000.

I've never seen a $8 billion jackpot, have you?  If you have to account for splitting the pot, that just makes it worse.

For conventional (6 numbers, same pool of N numbers). The odds are 1: N! / (720*(N-6)!  Not N^6 since you can't draw the same number twice.  For ones like power ball where there is a separate pool of M, it's one in (M * N!) / (120*(N-5)!)... which is usually worse odds...  Did I miss something, where did the divide by two come from in your equation?  That's why the resulting odds are usually in the 1:50-100million range, not 1:8 billion.

[slugcatcher]

Is she single? 

[HankB]

They'll say something like the "74000000 to 1" when the jackpot is $125 million.  So you think the odds are finally in your favor because of all the previous lotteries where there was no winner and the pot just accumulated.  But if you calculate it (depending on the game) the actual odds of a winning ticket might be (50^6)/2 or 1:7812500000.

Odds aren't quite that bad - various discussions of computing lotto odds are on the web, one of which is here:   http://members.cox.net/mathmistakes/rawdata.htm

You always have the same odds of drawing a winning ticket.  When more people play, the odds of the jackpot being split are higher. So when there is a huge jackpot and everyone buys a ticket, you the odds of the pot being split are higher.

I did mention the possibility of multiple winning tickets . . .  

Reminds me of an ad I once saw in the local "free" community ad paperL "Old lotto tickets, ten cents each! You save 90% of the cost and still have almost the same chance of winning!"   

[zxcvbob]

For conventional (6 numbers, same pool of N numbers). The odds are 1: N! / (720*(N-6)!  Not N^6 since you can't draw the same number twice.  For ones like power ball where there is a separate pool of M, it's one in (M * N!) / (120*(N-5)!)... which is usually worse odds...  Did I miss something, where did the divide by two come from in your equation?  That's why the resulting odds are usually in the 1:50-100million range, not 1:8 billion.

I thought with Powerball you pick 6 numbers; five of them from the same pool of 1-50, then a 1-25 powerball.  I didn't throw out the duplicates (d'oh!) so the real odds should be 50x49x48x47x46 (254251200) instead of 50^5 for the regular numbers, then multiply that by 25 for the powerball gives you 1:6356280000 odds.

1:6.4 billion is better than 1:7.8 billion, but...

[birdman]

Kind of, there is the divide by 120 part (5!). Because 50!/45! (the 250 million ish one) is the number of combinations not permutations, meaning the 250 millionish is not only the number of different sets of 5 numbers, but it's the number of different orders as well.  The factor of 120 corrects this since lottery doesn't care about the order.  That's why a 50 number, 5-pick main with a 25 number power ball has odds around 50 million:1, not 6.4 billion to one.

[zxcvbob]

So I *should* buy a Powerball ticket when the jackpot is over $53 million...

[Jamisjockey]

They'll say something like the "74000000 to 1" when the jackpot is $125 million.  So you think the odds are finally in your favor because of all the previous lotteries where there was no winner and the pot just accumulated.  But if you calculate it (depending on the game) the actual odds of a winning ticket might be (50^6)/2 or 1:7812500000.

I've never seen a $8 billion jackpot, have you?  If you have to account for splitting the pot, that just makes it worse.

I disagree.  The odds of you pulling a one time draw of a set of numbers should be the same every time. 

[wmenorr67]

At one time if I could just win $100,000, I would be able to pay off all my debt and have a little left to save.

[Scout26]

Yep, when I buy gas, I look to see what the lotteries are up to.  Anything more then $100million and round my purchase up to nearest $5 with lottery tickets.

Which means I also spend about $20-30 a year to daydream for a couple of days about funding the APS Pirate Hunting Expedition. 

[CNYCacher]

I disagree.  The odds of you pulling a one time draw of a set of numbers should be the same every time. 

He's not saying you have a better chance of winning, he's saying that the payout eventually gets high enough that the payout divided by the odds of winning is greater than one.

Or actually he is saying that the lottery promoters are trying to claim this, but it's really not true.


It's like if I offered you 5 to 1 you couldn't roll a 4 on a 6-sided die.  You wouldn't take that.  But if I offered you 7 to 1, you should play that all day.

[Brad Johnson]

All I ask in this life is a chance to prove that money won't make me happy.  If it doesn't, well, at least the misery will be affordable.

Brad

[gunsmith]

Anyway, I go with "go her", "lotteries are a tax on people who are bad at math", and "dumb state". And leave with this interesting opinion--lotteries are one of the oldest ways to appease the unwashed masses--they are bread and circuses--by imposing an extraordinarily regressive tax on the poorest, by giving them a hope in pure chance that the state will make all their wishes come true--while "doing a good thing by helping {schools, etc}" it's just a social control method--why do you think winners are so publiscized, (well, the poor ones), and the wealthy winners (I know a family with millions who has won 250k TWICE) or the long term results of winner are never shown (it's like behind the music on crack).

So your wealthy winners are also unwashed masses bad at math?

there was a guy who won a few months ago, won two million & after tax got 800 thou, he notified the food stamp folks and was told he could still use his benefits.

Glen Beck interviewed him and gave him a hard time.

a few yrs ago some guy who owned his own construction biz won a huge fortune & due to his drinking & family dysfunction it ruined him, his niece ( or daughter or something) even killed herself.

I buy a ticket when its 100 million, what I do is get a quick pick & play advanced play - like 8 to ten bucks - this way if it goes higher I'm still in the game, but I never ever play scratch offs. ... well I did 25 yrs ago & never won more then a buck. When I moved to Reno and went to casino's the odds were much better-back when I was a cab driver I won 1000 bucks one time, 400, another and a bunch of times won between 40 & 200 ... I'm pretty close to even I think.