> Even if Mandel were to win, there was the possibility of multiple winners — a scenario that could significantly dilute the jackpot.
In fact Mandel was not able to calculate the odds of multiple winners. With publicity about an outsized jackpot, the general public is also expected to buy tickets in larger numbers, making it hard to estimate the expected winnings.
For that reason, I would not consider this to be gaming, or cheating the lottery. There was no guaranteed profit, and a moderate chance of large losses. So he took a big gamble, and was lucky to be the only winner in the end.
An interesting question is whether lotteries should
publish the number of tickets sold when it implies
expected winnings for every single ticket sold (as was the case in Mandel's lottery). On the one hand that should sell more tickets, but on the other hand it will deter some people from buying once the numbers stop being published...
True, but he followed the cardinal rule of successful traders everywhere: gamble with other peoples' money.
The life insurance front sounds like a lot more work than either the algorithm or the ticket buying logistics! I wonder how much of the payouts ended up with lawyers.
But in his case he paid himself a commission and in the case the jackpot was split I’m sure he’d still get paid even if the investors got less.
It makes me think of the other lottery system where you bet numbers that are less likely than average to be bet by others so your chance of sharing the jackpot is less. Certainly there are some numbers that people over bet, and in some lotteries you could avoid some of them in principle by betting on numbers that couldn’t possibly be part of a date.
The upside for the scheme is limited though because a large majority of lottery tickets are sold as ‘quick picks’ where the machine picks a random number for you which is not only easy but a sound min max strategy.
> An interesting question is whether lotteries should publish the number of tickets sold...
Don't many State lotteries publish a "where your lottery dollar goes" breakdown - showing the percentages of ticket sales revenue going pay winners, going to support State programs, going to reseller commissions, and going to Lottery administrative overhead? With that breakdown, a history of the jackpot sizes, and some number-crunching, you should be able to get pretty good feel for ticket sales for each drawing.
UI can't wait until we hear of the next level MIT geeks in vegas armed with AI glasses doing all the card counting for them, across many tables, decks, days, tourneys etc...
If there is one industry that should be afraid of AI nefariousness - it would be gambling.
All gambling except for poker is basic statistics. You are a few decades too late to trying to beat blackjack with computers. The advantages of card counting are negated anyway by multiple decks and shuffles before all the cards are dealt.
Like with trading systems, the thing that made it work here was not being smart about the maths. In fact anyone with high school maths would be able to work out when it was worth doing.
The thing here was the audacity to try to put together a business around getting it done. Things like getting investors, figuring out how to print the tickets, and delivering them.
I’ve come to believe that this is a universal principle: disbelief regarding feasibility prevents actualization of vision; once one sees that it is done it is as if a mental barrier has been removed.
However, even after the mental barrier is removed - Majority of us do not have the motivation to actualize our visions even after we know and believe that they are feasible.
I suspect that it takes a special human emotional/energy makeup to overcome enormous obstacles in pursuit of actualizing a vision. Thats what makes entrepreneurs/hustlers special.
> Majority of us do not have the motivation to actualize our visions even after we know and believe that they are feasible.
Much of the time the decision not to pursue the vision is the right one, because so much sacrifice is required. A lot of successful entrepreneurs have tragic personal lives because in the pursuit of their vision they sacrifice their families. For those of us who are perhaps older, with a lot of personal commitments, and perhaps are also economically comfortable, those sacrifices are too great.
IMO this needlessly aggrandizes entrepreneurs. The way I see it, certain people end up in situations where they can take the risk, and some take the risk. For instance if you already have some wealth saved up you can take a couple of years to try your idea. Or if you work in a field where you are very likely to be able to get back in. On the other side, you might have relationships that are likely to come with you on your journey, whether that's customers or collaborators.
It just comes down to being able to give it a try, mainly from having a combination of low failure cost and high success probability.
I hate lotteries and gambling in general. Take from the hopeful, give to the lucky. And the house (almost) always wins. Yes, I have some respect for those who can game the system, because I like the house losing, but their winnings are still taken from the hopeful (or worse, the naive and stupid).
> I hate lotteries and gambling in general. Take from the hopeful, give to the lucky.
Sorry, but you've just described the underpinnings of our current winner-take-most societal-wide economic systems. Every day, more and more of our systems are rooted in luck. Being born to the right family, getting into a prestigious university, getting a good job, timing the housing market, investing in the stock market, starting a business... Everything is "take from the hopeful, give to the lucky".
But even if you are correct, lotteries and gambling have more of the essence of "take from the hopeful, give to the lucky" than any of the things you described.
That's why in most countries the government has the monopoly and no one else is allowed to run lotteries. It's like an opt-in tax with a gambling element, you can choose to give the government even more money and most likely won't get much in return.
Even if you have an angle, the house still wins. If you buy all the possible tickets for a large progressive jackpot, it doesn't hurt the house at all. They get their commission on all those tickets, the jackpot money comes from the players, and the house has to pay out the required % one day or another.
If you’re playing against the house yes, statistically you’re gonna lose over time. But in games where your winnings come directly from beating others rather than pure luck (poker is a popular example), you can indeed beat others based on your skill.
That said, there are still a lot of factors you cannot control (like the skill of your opponents, luck of the draw) which mean your approach is absolutely the wisest. Anyone who walks into a casino with money they can’t afford to lose is setting themselves up for failure (both at the table and potentially in life).
Voltaire made a fortune in more or less the same way in 1730 [1].
> This system worked best if a consortium of players bought tickets together and split the winnings, so, writes Pearson, Voltaire, de la Condamine and 11 compatriots teamed up and by June 1730, all had made a tidy sum.
The the lack of effort being additionally apparent with the outdated quote about the ban in the United States as lotteries are currently run by 48 jurisdictions: 45 not 44 states plus the District of Columbia, Puerto Rico, and the U.S. Virgin Islands.
It took me a while to figure this out but I think YC are particularly looking for answers to that question that result in a win-win, ie where you hacked a system to your advantage, and to everybody else’s advantage.
The lottery is an explicit zero-sum game. So long as he didn't actually break the rules, I don't think YC would care. That said, the post-lottery financial maneuvers seem like a red flag.
But I guess this guy meant more like, some sort of "sophistication" like in the way that brute-force is an attack on SHA3, but no one has said publicly any better way to hack it!
Although for brute force to yield a successful hack, it must be sophisticated, by definition, otherwise it would be infeasible. And in this case, I think the guy qualifies, because he: 0) was successful, 1) reduced the brute force possibilities, and 2) rose to the logistical challenge of implementing them. Haha! :)
There was no contract. The result of Leonard v Pepsico makes lots of sense read in the context of something like Carlill v Carbolic Smoke Ball Company (the classic English contract law case every student learns).
In Carlill the Carbolic Smoke Ball Co. claimed it had never intended a contract to arise, that the advertisements were just "puffery" and no reasonable person should have expected (as Carlill did) that if they do what it says in the advert and yet they get Influenza anyway (Carbolic Smoke Balls were a bogus cure for flu) they can claim the £100 stated (then a considerable sum of money) like it says in the advert. The judge said well, why does the advert explain that the company has £1000 in a bank account? If there's no intention to pay people money, why did you make a bank account, put money in it and tell them about it? No, this is clearly a contract, you offered, she accepted by performance, now you need to pay the nice lady.
But Pepsico clearly doesn't expect to give out fighter jets. It doesn't have any fighter jets, and the depicted scenario is ludicrous, which is exactly the opposite situation.
We can see that - at least after the initial hype - John Leonard saw this as a profile opportunity more than a serious attempt to get Pepsico to give him a jet, when he's like oh, this shouldn't be a Federal trial, this should be in front of a jury of my peers, and I define "My peers" to be "The Pepsi generation" that's clearly not serious any more.
According to the article, the scheme depends on this seekrit algorithm for picking all the combinations, called "combinatorial condensation".
That term links to an article in Romanian, which I can't read. But honestly, why's it supposed to be hard to list all possible combinations of 6 of the first 50 integers? Isn't that the same as listing all 6-digit integers in base 50?
[Edit] ...Yeah, you also have to filter-out all those integers with repeated digits, but that hardly needs a super-seekrit algorithm.
We call this "roll-over", when no-one wins so the pot accumulates. This week's draw doesn't need to be profitable, the overall scheme needs to be profitable.
Say $2MM in tickets are sold, putting $1MM in the pot. And no-one wins. So next week, $2MM in tickets are sold, putting $1MM in the pot - which already contained $1MM from last week. Continue this over a few weeks.
After 5 weeks you have $5MM in the pot - and another $5MM in profits. On week 5, $2MM in tickets could buy $5MM in winnings, but that doesn't mean the scheme is negative. It just means this week's draw looks like a good deal.
States love lotteries because the house cannot lose - their share is taken before it goes in the pot, so the pot is never a liability.
Honestly, I prefer it. But for one reason or another, MM is the accepted suffix when dealing with money (it comes from mille in latin meaning thousand, not million.)
They're not, generally, but if the stars align correctly, nobody will win the jackpot until it's gotten so large that buying every possible combination will - if you're the only winner - pay you more than the base odds. But then you're betting that you will have the only winning ticket.
I'll usually buy a few tickets if it hits that point. Cheap fantasy.
They’re progressive jackpots, the prize pool increases until someone wins, at which point it resets. So a ticket for an individual drawing can be +EV if no one has won in a while, with the excess money coming from players in previous games rather than the lottery administrator.
That would be true if each combination could only be bought once.
However, as you cannot know in advance how many other players will play, or what numbers they will select, you cannot properly calculate an EV. Further, lottery players rarely play randomly. A winning combination made up exclusively of numbers less than 31 has a higher chance that someone else will have the same numbers (as many people use birthdays to pick the numbers they will play).
The recent Powerball lottery prize was for $1 billion but there are fewer than 300,000,000 possible tickets. But like somebody else mentioned, there's always the possibility of multiple winners.
For the recent billion dollar prize, a single person won.
The prizes get huge because when there's no winner, the prize rolls over to the next drawing.
On a related note, Hungary's state-owned gambling company sold soccer bets in 1998 that could guarantee 25% returns. Unlike the lottery tickets in OP's story, the Hungarian arbitrage required buying only a couple of tickets and hence needed much less capital. Alas, there was no math-savvy enterpreneur to game the system, so the Hungarian gambling company made record profits. [1]
for all of his mathematical , logistical prowess, he should have focused his attention on the bigger casino: the stock market. He would have found far easier pickings.
Stefan Mendel is a jew born during the Holocaust. He described how the economic state and ethnic persecution motivated him to find ways to raise the funds for getting his family out of Romania. He defines himself as Israeli-Australian that was born in Romania not as a "Romanian" as he is described in this article. A distinction that I believe is very important to him.
> Stefan Mendel is a jew born during the Holocaust.
He was born earlier than the Holocaust (especially the holocaust in Romania). Not to dismiss your point, but it's a disctinction very important to the people born duing the Holocaust.
> He described how the economic state and ethnic persecution motivated him to find ways to raise the funds for getting his family out of Romania.
As a matter of fact, Jewish people were the only people able to leave communist Romania when the borders were closed. Ana Pauker, the Romanian foreign minister (herself jewish) brokered a deal with Israel to allow 4000 jews/month to emigrate to Israel. Several other rounds of exit visas given by the Romanian government to jewish people emigrating to Israel (with oil equipement and hard cash received in return) led to basically an exodus of most jews in Romania. As an anecdote, my hometown was 80% jewish when my parents were born and two years ago, the last jewish person passed away at a littler over 100 years of age. I also have jewish ancenstry.. To post some numbers, there were ~400k jews in Romania at the end of WW2 and less than ~10k when the Iron Curtain fell. A large part of (what I consider to be) Romanian culture was lost during the cummunist quest of rewriting history and creating a 'pure' nation.
Just to put things into perspective. The communist rule in Romania was brutal and criminal even after the Red Army retreated and was under 100% Romanian leadership, and was especially brutal against ethnic minorities. Jewish people actually targeted during the anti-semitic Stalin purge which echoed in Romania as well, as local aparatchicks used this to purge the jews out of government (Anna Pauker is also a victim of that - not to defend her, she was also commie trash but a victim of antisemitism nontheless).
In fact Mandel was not able to calculate the odds of multiple winners. With publicity about an outsized jackpot, the general public is also expected to buy tickets in larger numbers, making it hard to estimate the expected winnings.
For that reason, I would not consider this to be gaming, or cheating the lottery. There was no guaranteed profit, and a moderate chance of large losses. So he took a big gamble, and was lucky to be the only winner in the end.
An interesting question is whether lotteries should publish the number of tickets sold when it implies expected winnings for every single ticket sold (as was the case in Mandel's lottery). On the one hand that should sell more tickets, but on the other hand it will deter some people from buying once the numbers stop being published...