SoylentNews is people
SoylentNews
Spotted at HackerNews is a link to a multipart series from Pennsylvania's PennLive and several collaborating outlets on the wins of improbably lucky lottery players.
On, Dec. 29, 2016, Clarance Jones did something that most Americans could scarcely imagine: He cashed-in 20 winning scratch-off tickets, collectively worth $21,000, one after the other.
For the average lottery player, that would be the ultimate payday. For Jones, it was practically routine.
In the past six years, the 79-year-old from Lynn, Mass., has won more than 7,300 lottery tickets, totaling $10.8 million. That establishes him, by far, as the luckiest lottery player in America.
But that luck, experts say, is unlikely to be what it seems. And Jones is not alone in raising eyebrows
The three part series goes on to look at the patterns of frequent winners, and the attitudes of lottery retailers to these apparent runs of incredible good fortune.
In Pennsylvania alone, more than 200 players have won at least 50 prizes of $600 or more in the past 16 years.
Statisticians approached by PennLive say many of those wins are difficult to explain by luck. In other states, investigations into frequent winners have sometimes found their wins are rooted in theft and cheating, or schemes relating to tax evasion and money laundering.
"From a statistical point of view it stinks to high heaven," said Ronald Wasserstein, executive director of the American Statistical Association.
The Pennsylvania Lottery, however, has a different view: Its most frequent winners are simply frequent players.
The Original HackerNews Thread and reporting from the Columbia Journal on the FOIA requests which underly the reporting. There are associated articles in The Boston Globe, New York Daily News, Hartford Courant, and the Cleveland Plain Dealer.
(Score: 2) by Virindi on Sunday September 17 2017, @10:13AM (1 child)
Okay. I was just talking about the "20 tickets in a row!" claim.
But also that "statistician" is being a bit misleading as well if that was the way they actually phrased it. The way that sentence is phrased makes it sound like the birthday-problem-effect was not taken into account. That is, there are a significant number of "players" who spend a decent amount on lottery tickets. This increases the probability that some player somewhere will hit a lot of winning tickets, even if "the probability that person x makes this many winning tickets" is low.
The statistic we are interested in is, given how much "serious players" spend on lottery tickets and how many "serious players" exist, what is the probability that one of them somewhere hits this many wins. The one person who does win the lottery always beats the odds by a huge amount! But the odds that someone will win the lottery are extremely good.
It sounds like the probability that any player anywhere will get that number of wins is in fact low, but the statistician's statement does not show this.
(Score: 4, Interesting) by theluggage on Sunday September 17 2017, @12:10PM
The problem is that - if you read TFA and skim the "How did we investigate" link - they're really running on empty when it comes to data. Sounds like the only data they had was on claims (not how many tickets were sold, or how many any individual bought) and in some cases didn't even know the odds of the game (and many of these are pre-printed scratch card 'games' with convoluted rules - not the classical 'pick six numbers from fifty' lottery draw where the odds are probability 101) or the price of the winning ticket. The detailed account is TLDNR - but there is a metric shedload of educated guesswork involved.
So the "would have had to spend hundreds of millions of $ on tickets to have a 1:10000 chance of winning that much" is probably about the best they could do in the absence of data on how much the subject did spend or what the total ticket sales were: it is clearly unlikely that even a "serious player" would spend that much - and even less likely that enough billionaire scratchcard addicts exist to make the appearance of such a winning streak unremarkable.
I think its OK provided nobody claims to have proven anything and the conclusion is simply "more investigation required" or "shouldn't lottery organisers - who actually have access to the data - be monitoring results for suspicious blips?"
I mean, to start with, nobody who can afford to buy $300,000,000 worth of scratchcards got where they are to day by paying for shit with their own money. :-)