The goal of this project has been to find ways to get rich quick, without taking morally reprehensible and soul-crushingly boring jobs with titles like “Junior Housing Analyst” or “Pricing Lead for Southeast Asia”. The following research represents hundreds of non-boring hours of online exploration and experimentation between June 2021 and May 2022, producing net profits upwards of $35,000.

This research began with an extensive literature and methods review. Our team quickly found guides on drop-shipping, bona fide import/export, fixing old cars, retail arbitrage, geo-arbitrage, gigging on upworthy, creating content… even learning to code. The authors personally experimented in our lab with most of these. The problem we found with all of these “passive income” hustlepreneur ideas is that they’re mostly just different jobs working for someone (or worse, a platform) in already-explored niches, and if they aren’t insanely competitive, they are only questionably more enjoyable than flipping bug burgers. What everyone wants is a good scheme, not a job.


Hustles can be divided canonically into two general groups; scams and schemes. Modern hustle scholarship is nearly unanimous that in practice there is significant overlap between these clades, but the distinction is still useful. Scams are hustles in which money is extracted from a real life sucker. Scams are plentiful, and some of them are really great at making money. From emailing grandmas about their Apple IDs to anything blockchain-related, that payday is coming from some schmuck’s family’s college fund, and that is not something our team has any interest in living with. When the legal liabilities are properly accounted for, it becomes clear that true scams need to have massive upside to be worthwhile, and most people don’t have the constitution for that in any case. Our team’s advice is to simply stay away from the whole category1.

The other, more virtuous clade of hustles targets systems, not random suckers. One popular and simple scheme is the self checkout underscan. The schemer simply scans every other item, or only pretends to scan the most expensive items, while really scanning cheap stuff. Just be careful, stay calm, and if anyone asks to see your receipt say oh dang sorry I thought it scanned let me pay for that now. Helps if you’re white but anyone can do it. Boom you’ve executed a scheme. Feels good right?

The whole reason the Whole Foods put in the self checkout is to pay fewer people, so there’s no way they’re gonna have some motherfucker stand there and make sure every item in your cart matches. They’re the ones who built this dumb system because a Bain consultant told them your shame and semi-concious Protestant values could be leveraged to lay off a quarter of their workforce.

These techniques and methods are effective, but hardly a breakthrough. Things get more interesting when a researcher carefully applies this mode of thought to other domains. Reaching a real breakthrough requires finding novel schemes with higher ceiling and better risk ratios.


This scheme is safe, almost entirely legal, and massively profitable. It is not unlimited, but it could probably put a child through college with hard work and sustained favorable conditions. It’s called gambling promotion arbitrage, and we will be as brief as possible.

Gambling companies are willing to spend a lot of money to acquire new users. User growth is a key metric for digital businesses, and c-suites are willing to spend a lot to keep their jobs. User growth is important because with the recent legalization of gambling in many states, whichever company can establish dominant market position will likely be the most profitable in the long term. Additionally, gambling is an addictive product, and even if many users make a small amount of money, the addicts will be very valuable to the gambling company (also known as a sportsbook) for a very long time. Therefore, sportsbooks are willing to spend a lot on getting new users.

The research team hypothesized that it would be possible to take advantage of “risk free bet” promotions. FanDuel, Barstool Sportsbook, UniBet, and BetMGM all have this type of promotional offer for new users at the time of writing. Normally, the way this offer works is that a new user can make a bet up to a certain dollar amount, and if that bet loses, they get something back. From the sportsbook’s perspective, this incentivizes relatively larger first forays into betting for new users, and increases the likelihood that a user becomes a habitual large-dollar user, either because they lose and want to win their money back, or because they win and feel the high. Getting new users to deposit relatively large amounts also makes it easier for users to lose track of how much they’re spending and treat the initial deposit as sunk costs. It’s a great deal for sportsbooks because a lot of the time new users lose their first bets, so even though the sportsbook isn’t making money yet, they also aren’t losing any. Another devious wrinkle is that many “risk free bet” promotions don’t even return your money in whole. Many give you a “free bet” if you first bet loses, which doesn’t return principal if it wins, effectively taking 50% of your “risk free” stake if it loses. Even the sportsbooks that give you back the full amount if your first bet loses won’t let you immediately withdraw that money; they require you to “play through” the full value of whatever they gave back, ensuring that they are usually making money even when they “return” your lost bet.

Our research team had the key insight that although “risk free bets” are a great deal, they have a lot of risk attached. You can lose all of the money you stake with just a little bit of bad luck. However, if you sign up for multiple sportsbooks at once, you can use two risk free bets to remove all risk.

An example will render this obvious. Suppose a researcher creates accounts and deposits 1000 dollars on sportsbooks X and Y, which both offer “1000-dollar risk free bet” promotions for new users. The researcher places a bet on sportsbook X that A > B, and a bet on sportsbook Y that B > A. When event A vs B occurs, the researcher is left with 2000 dollars in one sportsbook, let’s say X in this case, and a risk free bet in the other, Y. Once again, the researcher places bets. 1000 dollars on sportsbook X that C > D, and the risk free bet of 1000 dollars on sportsbook Y that D > C. Either way, after C vs D occurs, the researcher is returned 2000 dollars on one sportsbook and nothing on the other. Notice that we still have 1000 dollars in sportsbook X, bringing our total cash to 3000. The researcher withdraws the money, having completed the scheme, and pays rent in Philadelphia for two months.

The hypothetical above is inaccurate for a variety of reasons, but demonstrates the core concept. We will describe the practical difficulties and best practices of this procedure as determined by our group’s extensive research, but before proceeding must make something absolutely clear to the reader and potential colleague in this exciting field. Even the most dedicated, technologically advanced, brilliant gamblers in the world have a hard time making money sports betting. The best way to beat the house is to play a different game than them. Gambling ruins lives and you must not be a gambler. You are a ghost in the shell, a glitch in the matrix, and your profit and health depends on a low profile. This is probably not illegal, but it’s also probably possible to get sued over if you do it on a large enough scale and someone can prove it2.


With that being said, one of the primary concerns in promo arbitrage is the vig. A typical vig on spread betting markets would be notated as “-110”, meaning that in order to return 100 on a win, you would need to stake 110. This is roughly a 9% cut that the house takes off the top, so it is in the interest of the promotional arbitrageurs to find incongruities between sportsbooks to reduce losses to the vig. Paying full vigs reduces yield at each bet by 9%, so instead of finishing with 3000 our example is more likely to finish with around 2830 or so – still a great yield, but it adds up. If you can get -105 on each side, that’s quite good.

Another significant issue is that not all sportsbooks have the same promotional terms. Many have different sizes of “risk free bets”, and some of these promotions only return the winnings, not the staked cash. This effectively halves your expected value on these sportsbooks. Many of these promotions also have sneaky “play-through requirements” hidden in the fine print, which can totally destroy your margin, and many have limited ranges of odds which you can qualify for the risk free bet with, or use it with. These are sufficiently inconsistent state-to-state, and constantly changing, that the only reasonable option at first is using rules of thumb and doing a little math. Some good guideline are as follows:

  • If it’s your first time, keep it simple.
  • Chain your promos. You want to stake risk free bets against new money that will yield you risk free bets on another sportsbook.
  • You can always trade risk for higher possible payouts, and you have plenty of wiggle room before risking getting into the red. Generally as the spread of possible outcomes increases, average expected value also increases.
  • Read the actual full terms and conditions of any promotion you are using for the first time. You can skip most of the sections but really read the one that governs how the promotion works!
  • Do the math, then check it.
  • Check your bets twice.

Due to the extreme mathematical complexity of this emergent decision space, part of our research team has developed a strategic simulator, currently in closed beta. Unfortunately the code is totally un-commented and also really poorly optimized, so we won’t be publishing it, but it did confirm that novel strategies and local maxima are possible depending on the risk tolerances and goals of a given researcher. There is no one “best way”, but there are certainly “better” and “worse” ways.

One final and crucially important consideration is that each sportsbook will only allow each person with a social security number a single account. The sportsbooks also keep track of PayPal and bank accounts, so no two accounts can have the same payment info. This naturally limits the amount of value that a single person can extract. In our research group’s experimentation, we found that many people – when approached regarding a 50/50 split of roughly 2000 dollars contingent on a very small amount of logistical work and a somewhat larger amount of trust – were amenable to mutually beneficial collaboration. By building a whisper network of scheme-curious friends and acquaintances, and by buying a LOT of shitty burner phones to administer these collaborators’ accounts on, the social security number issue becomes totally workable. Call it consulting, or coaching.

This final factor does increase the logistical overhead and general exposure significantly, but our team finds that this profit split model is poetically in line with the ethos of this particular method. Only by helping those around us may we help ourselves, etc.


At the time this study was initiated, this particular ploy method was entirely novel. At the time of publishing, there exist multiple tools to help schemers take advantage of this loophole, but to our knowledge all are relatively costly and certainly less fun than figuring it out yourself, or with a couple friends. As knowledge of this scheme spreads and as the gambling market matures, the gambling companies will have a different cost/benefit analysis, and the schemers will have to find greener pastures elsewhere. It is therefore that our purpose in publishing this research is twofold: One, you should absolutely do this for yourself, or find someone to do it for you, even if you don’t want to mess with doing it for anyone else. Two, familiarizing the online public with the science and practice of schemes will create a ripple effect of critical thinking, collaboration, and community, which will improve many lives and maybe even lead to a better world.

With these twofold goals in mind, this research team issues a call to action and a request for collaboration. Need help getting started? What’s the new horizon? How else can we use systems of exploitation against their architects? Drop us a line at


1. There is one scam which deserves special attention due to its sheer aesthetic virtue. It’s called triangulation, and all it requires is an eBay account and a short list of stolen credit card numbers. Hypothetical instructions are as follows:

  • List a consumable, high margin product on eBay for SLIGHTLY less than it is listed for on other websites.
  • When an order comes in, order the consumable product directly from the manufacturer (or a third party retailer) to the customer’s address, using a stolen credit card.
  • Your customer gets their k-cups at a discount, you get the payment from the customer, eBay gets their little cut, and the k-cup manufacturer gets paid as normal.

There’s only an incomplete link between you and the order — your identity isn’t known to the manufacturer, eBay (most likely) has no way of knowing you used a stolen card, and no one has an incentive to try very hard to detect it. Obviously the person whose credit card was stolen is SOL, which is what makes this a scam.

It is the opinion of the researchers that if it could be guaranteed that the customer would be informed of the fraud in a timely manner and the credit card company was left with the bill, this particular scheme would be a true work of art. We have as of yet not found any reasonable solution to this problem, and given the high moral costs of error, have not explored in earnest.

2. Tax implications will not be fully discussed due to it being slightly boring, but it’s exceedingly unlikely that any individual participant in the scheme is audited. Getting audited, while definitely a pain, is relatively low-impact if you’ve done the rest of your taxes correctly. This means investing energy in avoiding it is fairly low-yield, especially given that figuring out how to properly pay tax on a small amount of winnings is itself somewhat of a pain.