Hacker News new | comments | show | ask | jobs | submit login
The Lottery Hackers (huffingtonpost.com)
337 points by sergeant3 5 months ago | hide | past | web | favorite | 41 comments

James Harvey and Yuran Lu co-founded QuicklyChat (YC S12 - https://web.archive.org/web/20120816231511/http://www.quickl...) https://blog.ycombinator.com/quicklychat-yc-s12-brings-push-...

I guess they had no trouble with question 3 on the YC application: Please tell us about the time you most successfully hacked some (non-computer) system to your advantage.

For the record, plenty of people came to my Super Bowl party in 2005.

Reading this article I was thinking, "I really hope someone from Boston who participated in this is on HN". Check the comments and there it is :D

Care to share any additional thoughts/experiences on this whole thing?

> - I have a very large collection of losing lottery tickets.

Why do you still keep them around? Edit: or for how long does the IRS make you keep them, I guess.

I would assume its the standard 3 to 6 years for auditing, but for something as crazy as hundreds of thousands of tickets, might be worth keeping around longer.

Yeah, that's basically what it is. The tickets that are still within the 6 year audit window have all been digitized; at this point I mostly keep them as a souvenir.

How did you digitize that many tickets?

High speed scanner, and an unreasonable amount of manual effort.

I imagine one at a time at the time of purchase.

Do you still keep an eye on new lottery games?

A little, but mostly just as a casual interest these days.

lol <3

That was a fantastic long read! Wow.

A lot of intrigue in it, too, especially with the emergence of the MIT cohort, computer programmers, and electrical engineers.

Excellent decision on the part of the investigative reporter Jason Fagone to tell the story from the vantage of the Selbees.

The article is fun even as it smudges the Selbees (and other bulk ticket purchasers) with a dab of questionable ethics as is appropriate for a (minor-ish) vice such as lottery gambling.

Thanks for sharing.

EDIT: noun verb agreement

I guess you could argue on the ethics due to the fact that these "smart" players are never contributing to the jackpot, and only playing when the payout is higher than the probability (an over), but you need the mugs in order to fill the pool.

If the world was filled with smart gamblers, no one would bet because nothing would be an overlay.

Its like when I see a horse racing win pool go up from $0, to $2. Someone has had a bet. Whats the most they could possibly gain by that bet (given the current pool)? Its about $1.70, so why bet in the first place? I don't know. Maybe they liked the colour of the horse.

The game was delibratlly designed to divide the players into "smart" and "dumb" groups.

Most lotto games are (though not normally to this extreme where large groups can guarantee a profit). It's a good way of attracting players. As the jackpot gets larger, more and more 'smarter' players play the game. This video gives a good overview of why: https://www.youtube.com/watch?v=pVKTiXdCDyQ

"The article is fun even as it smudges the Selbees (and other bulk ticket purchasers) with a dab of questionable ethics"

Showing this "bulk buyers with questionable ethics" point of view made this article especially entertaining and educational.

>They’d had an incredible run: in the final tally, they had grossed nearly $27 million from nine years of playing the lottery in two states. They’d netted $7.75 million in profit before taxes, distributed among the players in GS Investment Strategies LLC.

>The lottery had worked how it was designed to work. In fact, as one financial reporter for Reuters would argue in the days after the report’s release, Cash WinFall was possibly more fair than other lottery games, because it attracted rich players as well as poor ones. Instead of taxing only the poor, it taxed the rich too.

If you're making a profit, you aren't being taxed. Winfall is, in fact less fair - because not only do poor people lose money, but rich people gain money.

Yeah, it makes no sense, I can't believe that the report said this too:

>The large groups had bought some $40 million in tickets, $16 million of which was revenue for the state.

No, actually it was negative revenue, since the state payed out more than $40 million in prizes.

The payout comes from the other bettors losing bets. The state never loses in a lottery.

My bad, you are absolutely right, the payout comes from previous weeks' undistributed payout, not from the state. The more they played, the more revenue for the state.

This is great. I wrote a personal app that monitors (scrapes) a few state lotteries websites and if the jackpot goes above my personal threshold (it's below pot odds and pot odds minus taxes) then it sends me a text message in the morning so if I feel like it I can buy a ticket on my way home from work.

P.S. a few years ago Megabucks (Wisconsin lottery) was above pot odds even after taxes and with the cash payout option!

> a few years ago Megabucks (Wisconsin lottery) was above pot odds even after taxes and with the cash payout option

This happens more often than you'd think. For example, eyeballing it, it looks like Indiana's Cash 5 lottery has a positive expected value right now, but I haven't broken out my detailed calculator to confirm.

The other major things you need to account for are the likelihood of splitting the jackpot with other winners, the fact that most people have non-linear utility functions, and the cost of logistics. (This last one may not seem too important, but if you buy more than 10 or so tickets, it becomes readily apparent that the time spent checking whether a ticket is a winner can add up.)

Seems like a lottery ticket scanner (+OCR +scripting) would be the next thing to hack together. And then I realized how/why your most recent tickets were digitized.

Split jackpots is why I never bother when Powerball or mega millions get huge because they are heavily played. Some of the state lotteries not so much... They take many months to grow because fewer people pay attention and play them.

You can do better if you cut a deal with the people you are buying the tickets through to split the commission, it can be an extra few % on your turnover.

As alluded to early in the article, he was selling 300k in tickets per year, making 20k profit, which is around 6% commission or so.

So when you plan on buying a lot of tickets, choose one agent, and tell them their total paid commission is about to get a significant boost because of you, and you should split it.

If you find a lottery that looks to be an over that is.

My favorite part was the indignation over forcing the rollover.

There is another long story on this subject, focused on the MIT 'gang' which suggests they'd worked out some very advanced math to identify specific sets of numbers that would most efficiently span the payoff space. So the MIT people probably had better results than the Indiana gang's purely random ticket selections.

Where is this article?


I think that's the video where he tells the story (which comes from his book)

I know that specific story from the book "How not to be wrong", by Jordan Ellenberg. It was nice to read it from a different angle in this article

> A ticket cost $1.

> a 1-in-54 chance to pick three out of the six numbers in a drawing, winning $5, and a 1-in-1,500 chance to pick four numbers, winning $100.

> winning three-number combination would put $50 in the player’s pocket instead of $5, and the four-number winners would pay out $1,000 in prize money instead of $100, and all of a sudden, the odds were in your favor.

1 in 54 wins you $50, 1 in 1500 wins you $1000. I feel really dumb right now, as I must be missing something, but how are those winning odds?

If you buy 40500 tickets. 750 of them will win $50. 27 of them will win $1000. As a result, you spent $40500, and gained $64500. That's a nice guanranteed 59% gain assuming you have enough cash on hand to pull this off (in less than a week's time and effort).

Taxes will probably wipe out most of that profit though.

Fair enough, unless you can deduct your cost against your winning. I don't know whether you can.

The simplified version is that if buy 1500 tickets one of those will win you 1000$ and 1500/54=27 tickets will win 50$ each for a total of 2350$

Of course, stacked odds! Thank you

A question to self: Would I have done the same had I noticed the same kind of exploit? Probably not. I need to get off my ass.

An interesting question. Given a limited number N of lottery tickets you can buy, what is the minimal N for which you maximize your expected winnings? The answer maybe requires enumeration of particular tickets and cannot be derived just by formulas (my conjecture is that problem is #P-complete).

To me, there's something intoxicating about the spirit of schemes like this. Obsessions are thrilling.

Minor nitpick:

> he bet even more on the next roll-down, $8,000, and won $15,700, a 49 percent margin.

Winning $15,700 on an $8,000 wager seems like a 96 percent margin to me, not 49. (Disclaimer: I'm not familiar with betting jargon; might well be that you express margin as (your winnings minus what you paid) as a percentage of your winnings, but seems weird to me...)

It’s not betting jargon. It’s just simple business jargon. Look up how profit margin is expressed. It’s profit divided by total revenue. In other words, it’s the percentage of your total revenue that is profit, which is why you can never have a profit margin over 100%.

In this case, that means $7,700/$15,700, or 49%.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact