America spent eight years deleting every inch of distance between a young man and a bet. The bill is arriving on schedule.
Somewhere in America, during a recent championship season, a college athlete's phone became a hostile crowd. More than 1,400 abusive messages arrived in two weeks. The NCAA counted them. It withheld the athlete's name, which tells you something about what was in them.
The count comes from the NCAA's first online harassment study, conducted with Signify Group's Threat Matrix across the 2023-24 championships. Monitoring covered the accounts of more than 3,000 college athletes, along with coaches, officials, and teams. An algorithm flagged over 72,000 messages; human reviewers confirmed more than 5,000 as abusive, discriminatory, or threatening. Of the confirmed abuse, at least 12 percent traced back to angry bettors, and 743 messages referenced betting or match-fixing outright: fury over lost wagers, accusations of throwing games. Nearly three quarters of the betting-related messages landed during March Madness, which is another way of saying they landed while the sender watched, phone in hand, money on the line.
Notice what the athlete never had to do to receive any of this: place a bet, download an app, agree to anything. The blast radius found them anyway.
This is a story about how that radius got so wide, so fast. Call it a design story. A product that put an active betting account in the pocket of nearly half of American men under fifty took less than eight years to do it, and the machinery behind that speed was engineered, tested, and aimed. The engineering is the part almost nobody writes about, and the young men inside it will get no sermon here. Targets deserve reporters.
What losing money used to cost
For most of American history, losing your paycheck on a point spread required effort.
The legal version demanded a plane ticket to Nevada. The illegal version demanded a guy, and finding the guy, and looking the guy in the eye. Either way, the transaction came wrapped in friction: a door, a drive, daylight, witnesses, cash counted out where your hands could feel it leaving. Friction never stopped a determined gambler. It slowed the casual one, and it gave the struggling one a hundred small exits between the impulse and the wager.
The old system carried one more brake, and nobody thought of it as a consumer protection at the time: another human being. The bet used to require saying it out loud. To the guy, to the teller at the window, to the friend who covered the drive. Somebody heard the number, and hearing it made it real, and making it real is precisely what a man three losses deep does not want. The confession was built into the transaction. The app's founding achievement was removing the witness.
On May 14, 2018, the Supreme Court decided Murphy v. NCAA and struck down the federal law that had walled sports betting inside Nevada. The Court did not legalize anything. It handed the question to the states, and the states answered like a starting gun had fired.
Eight years later, 38 states and Washington, D.C. offer legal sports betting, and roughly 30 of them let you do it from your phone, according to the American Gaming Association's own tracking. Americans legally wagered $166.94 billion on sports in 2025. The sportsbooks kept $16.89 billion of it. Both figures come from the industry's trade association, which publishes them with pride.
Every barrier the old world imposed is gone. The door became a login, the drive a thumbprint. Cash turned into a number that never touches your hands. Even daylight became optional, because the app works at 2 a.m. exactly as well as it works at noon. What used to be a trip is now a reflex, and the entire journey from impulse to wager takes less time than reading this sentence.
The sales pitch to statehouses was revenue: legalize it, regulate it, tax it, and watch the money that once flowed to offshore books fill the school budget instead. The revenue came in smaller than promised. When the Manhattan Institute ran the numbers on that promise in June 2026, it titled the report "A Bad Bet," and spent much of it documenting what arrived instead of the windfall. We will get to what arrived.
The industry calls all this deleted distance convenience, and convenience is the truthful word for most of it. A man who wants twenty dollars on his team on a Saturday is not a public health problem, and the app serves him fine. The trouble lives one layer down, in what the app was built to notice about him, and what it was built to do with what it notices.
The proclivity machine
To understand what the app actually sells, you have to look at one piece of its menu: the parlay.
A straight bet is a single wager on a single outcome. The Cowboys cover or they don't. Sportsbooks keep a thin slice of straight bets. Data compiled from Nevada's regulated books between 1992 and 2023 by the University of Nevada, Las Vegas puts the average hold on standard sports wagers around 5 percent.
A parlay chains several outcomes together. Every leg has to hit or the whole ticket dies, and the payouts look like lottery numbers, which is the point. On parlays, that same three decades of Nevada data shows an average hold of 31 percent.
Read those two numbers again. The house keeps roughly six times more of every parlay dollar than of every straight-bet dollar. New Jersey's regulator reported a 19.9 percent parlay hold for January 2024 alone. If a sportsbook were a restaurant, the parlay would be the dish with the highest margin on the menu, and the menu would be redesigned nightly to put it in front of the customers who order it most.
The parlay also happens to be the bet that feels smartest to the person placing it. A five-dollar stake promises a four-figure payout. Each leg looks reasonable on its own, and a young fan who watches forty games a season believes, with some justification, that he knows which legs are reasonable. Stacked together, the legs multiply against him, and the math that does the multiplying never appears on the screen. What appears on the screen is the payout, in a large font, next to a button. The same-game parlay, the industry's proudest recent invention, compresses the whole seduction into a single evening: your quarterback, your over, your anytime touchdown scorer, all riding together on the game you were going to watch anyway.
Knowledge of the sport, the one asset the young bettor genuinely owns, becomes the hook that sets deepest. He is not wrong that he knows football. He is wrong about what that knowledge is worth against a correlated pricing engine, and the app has no incentive to correct him.
That is not a metaphor. It is the stated strategy, in the company's own filings. On its August 2022 earnings call, DraftKings' CEO told investors: "we're seeing great success driving parlay mix in our sportsbook product," reporting that parlay bet mix had jumped 1,700 basis points in a single quarter, and adding, "I really love how we are merchandising our same game parlays." The company's February 2024 shareholder letter states that its "foundational customer and state unit economics are exceptional," that "structural sportsbook hold percentage increased as we drove parlay mix," and that it deploys "surgical promotional tactics on a customer-specific level."
Sit with the vocabulary. Drove. Surgical. Customer-specific. The bettor did not drift toward the highest-margin, lowest-odds product on the menu; the menu drove him there, and the letter says so as a boast. In most industries, aiming individually tailored promotions at the customers most drawn to your worst-odds product would trigger a conversation about duty of care. In this one, it triggers a push notification, timed to the game you already told the app you care about.
None of this is illegal. All of it is working precisely as designed. The question a country ought to ask about a machine like this is not whether the people inside it should have known better. It is who tuned the machine, and to find what.
The bill
Ask how the experiment is going and the data answers from four different directions. The directions agree.
Start with diagnoses. Epic Research, which studies medical records across major American health systems, reported in June 2026 that among adult patients in states that legalized sports betting, diagnosed gambling disorder rose from 3.0 per 100,000 patients in early 2018 to 4.8 per 100,000 in early 2026, an increase of about 61 percent. Here is the finding that should end the shrug: over the same eight years, in states that never legalized, the rate fell 29 percent. Two lines on a chart, starting from the same place. One bends up where the apps switched on. The other sinks.
Epic frames this as an association, not proof of cause, and honest writing keeps that frame. But the divergence is the kind that keeps epidemiologists up at night.
The phones agree with the charts. The National Council on Problem Gambling reports that call traffic to 1-800-GAMBLER has surged more than 150 percent since 2018, and its helpline network logged a 33 percent increase in calls, chats, and texts between 2023 and 2024 alone. In 2021, the year mobile betting reached its first wave of big states, the council recorded calls up 43 percent, texts up nearly 60 percent, and chat volume up 84 percent in a single year. Traffic counts are not diagnoses. They are people reaching for a rope, and the rope keeps getting busier.
The money agrees with the phones. A widely cited working paper by researchers at UCLA and USC, not yet peer reviewed and worth flagging as such, tracked household finances across states as online betting arrived and found bankruptcy likelihood rising 25 to 30 percent within three to four years of legalization, alongside roughly $8 billion in additional annual debt collections. Even the cautious end of the research runs the same direction: the Manhattan Institute, a think tank not known for hostility to markets, published its own analysis in June 2026 and, working from Ohio's state gambling survey, offered what it called a conservative estimate: legalization increased problem gambling by half a percentage point, from 2.8 to 3.3 percent. In Ohio alone, by the report's own math, that is 46,523 additional people. The same report cites research finding a 79 percent increase in calls to a problem-gambling hotline after legalization, and estimating that the share of people spending more than 1 percent of their income on gambling rose more than fivefold.
And the revenue that was supposed to justify the trade? That was the Manhattan Institute report's actual subject, the reason for its title. The tax windfall pitched to statehouses came in below the promises, the report found, while the social costs listed above kept compounding off the books. The states sold the guardrail to fund the school budget, and the check was smaller than the brochure.
Diagnoses, hotlines, bankruptcies, disappointing tax receipts, and a right-leaning think tank's floor-level guess. Five instruments, one melody.
Who is holding the ticket
Now ask who the machine was aimed at, and let the surveys answer.
The NCAA polled 3,527 Americans aged 18 to 22 in 2023 and found that 58 percent had engaged in at least one sports betting activity. Among students living on campus, it was 67 percent. More than a third of the bettors had wagered on their own school's teams, and 35 percent had placed bets through a student bookmaker, which is a phrase that should not exist and does.
Sit with that last one. A student bookmaker is a nineteen-year-old running a book out of a dorm, in an era when the legal product is one age verification away. The demand is deep enough to sustain both markets at once.
The pattern holds as the men get older. A February 2025 Siena College Research Institute poll found that 48 percent of American men aged 18 to 49 hold an active online sportsbook account, and that a majority of bettors place wagers at least weekly. Pew Research reported in October 2025 that 31 percent of adults under 30 had bet on sports in the past year, including 36 percent of the men, and that one in five young men had wagered online. The National Council on Problem Gambling's national survey found that two thirds of all adult sports bettors are under 45, and that among Americans under 35, nearly half bet on sports.
It has become fashionable to read those numbers as a verdict on young men: reckless, dopamine-chasing, allergic to delayed gratification. That reading flatters everyone who profits from the simpler explanation. Young men are where the proclivity lives, and the machine was aimed, in the company's own word, surgically. A product engineered to convert impulse into wagers found the demographic with the most impulse and the deepest sports knowledge, the heaviest phone habit, the thinnest scar tissue. Marketing did the rest. A bullseye behaves like a bullseye.
And the ones who never gamble still get hit. Return to the athletes. In the NCAA's survey of nearly 6,800 Division I athletes, 7 percent of men's sports athletes reported receiving negative or threatening messages from fans who bet on their games. Among men's basketball players, the figure was 36 percent. The association now runs a campaign called Draw the Line to teach twenty-year-olds how to absorb the rage of strangers who lost money on their performance. The people whose free throws move the money have become customer-service targets for the money's bad nights, without ever opening the app.
The mood is turning
Something is shifting, and you can watch it move through the polling.
Pew's October 2025 headline finding was not the participation number. It was the turn: Americans increasingly say legal sports betting is a bad thing for society and for sports. The country that said yes at the ballot box and the statehouse is starting to feel the draft through the door it opened.
Washington has noticed. A federal proposal called the SAFE BET Act would ban sportsbook advertising during live events, require deposit limits, and bar operators from using artificial intelligence to track individual bettors' behavior. It is a proposal, not a law, and proposals in this space have a way of dying quietly in committee while the handle grows another ten billion. But read that third provision again, because it is a confession wearing legislative clothing. You do not write a law banning AI tracking of bettors unless the tracking exists, works, and worries you. The bill's authors looked at the machine and named the part the ads never mention: the product studies its user, and the study has a purpose.
The conversation, in other words, has moved. It used to be about whether the product is fun. Now it is about what the product knows, and what it does with the knowledge at 2 a.m.
The honest version of this argument owes the other side its due. Prohibition failed for alcohol and would fail here. Tens of millions of Americans bet without harm, and the old illegal market offered neither deposit limits nor a hotline on the screen, and answered to no regulator at all. Bringing the wager into the light was a defensible bet in its own right.
But the light came with a lab attached. What the country legalized in 2018 was not the neighborhood bookie with better paperwork. It was a machine that learns each user's proclivities and prices its menu against them, running on the supercomputer in the pocket of every nineteen-year-old in America, with promotional artillery no bookie ever dreamed of. We regulated it like the bookie. We got the machine.
The phone, again
Go back to that athlete, wherever they are. Fourteen hundred messages in two weeks. Each one written by a person who, a decade ago, would have crumpled a losing ticket in a parking lot in Nevada, or eaten the loss quietly because admitting the bet meant admitting the guy. The distance between that person's anger and its target used to be measured in miles and shame. Now it is measured in taps.
Friction was the guardrail, and nobody voted to remove it, not really. It left the way most guardrails leave now: one convenient update at a time, each approved by someone reasonable, until a college kid checks their phone after a missed free throw and finds a crowd waiting that no arena could hold.
The house always had an edge. Now it has an address for every customer, a dossier of proclivities, and a seat in the room where you sleep. The wager stopped being the product somewhere along the way. The product is the bettor himself: his impulses mapped, his menu priced against him, his phone glowing at 2 a.m. with an offer built from everything it knows.
The parlay is just the packaging.
Reyes Castillo writes about the subjects most writers walk around - addiction, money, shame, the things people actually live with. If your publication or brand has a story that's hard to tell well, he takes commissions: rcastj3@gmail.com.
Sources
1. NCAA / Signify Group Threat Matrix online harassment study, 2023-24 championships (accounts of 3,000+ athletes, ~500 coaches, 200 officials, 165 teams monitored; 72,000+ messages algorithm-flagged; 5,000+ confirmed abusive; at least 12% tied to bettors; 743 messages referencing betting or match-fixing; 73% during March Madness; one unidentified athlete receiving 1,400+ harassing messages in two weeks). Coverage verified: espn.com/college-sports/story/_/id/41665899/ (Oct 8, 2024); sports.yahoo.com/ncaa-pilot-study-finds-widespread-225513061.html
2. Supreme Court of the United States, Murphy v. National Collegiate Athletic Association, No. 16-476, decided May 14, 2018. supremecourt.gov/opinions/17pdf/16-476_dbfi.pdf
3. American Gaming Association, State of the States 2026 (2025 handle $166.94B; revenue $16.89B) and State of Play map (38 states + DC legal). americangaming.org/resources/state-of-the-states-2026/ ; americangaming.org/research/state-of-play-map/ ; mobile-state count: legalsportsreport.com/sports-betting/states/
4. University of Nevada, Las Vegas, Center for Gaming Research, Nevada sports betting hold data 1992-2023 (straight ~5%, parlay ~31%), as compiled at howgamblingworks.substack.com/p/understanding-parlays-americas-favorite ; New Jersey January 2024 parlay hold 19.9%: establishtherun.com/the-math-behind-parlay-betting/
5. DraftKings Inc.: Q4 2023 Shareholder Letter, February 2024 ("foundational customer and state unit economics are exceptional"; "Structural sportsbook hold percentage increased as we drove parlay mix"; "surgical promotional tactics on a customer-specific level"). ir.aboutdraftkings.com/files/doc_financials/2023/q4/DKNG-Q4-2023-Shareholder-Letter.pdf ; Q2 2022 earnings call, August 5, 2022 ("great success driving parlay mix"; parlay mix +1,700 basis points; "I really love how we are merchandising our same game parlays"). Transcript: fool.com/earnings/call-transcripts/2022/08/05/draftkings-inc-dkng-q2-2022-earnings-call-transcri/
6. Epic Research, "Gambling Disorder Diagnoses Have Risen More Than 60% in States That Legalized Sports Betting," June 26, 2026 (3.0 → 4.8 per 100,000, Q1 2018-Q1 2026; non-legal states fell 29%). epicresearch.org/articles/gambling-disorder-diagnoses-have-risen-more-than-60-in-states-that-legalized-sports-betting/
7. National Council on Problem Gambling: 1-800-GAMBLER traffic up 150.71% since 2018 (ncpgambling.org/news/ncpg-statement-on-national-access-to-1-800-gambler/); helpline volume +33% 2023-2024 (NCPG Helpline Symposium report, 2025).
8. Hollenbeck, B., Larsen, P., & Proserpio, D., "The Financial Consequences of Legalized Sports Gambling," working paper, 2024 (not yet peer reviewed): bankruptcy likelihood +25-30% within 3-4 years of online legalization; ~$8B additional annual debt collections. anderson.ucla.edu/sites/default/files/document/2025-05/Hollenbeck_The_Financial_Consequences_of_Legalized_Sports_Gambling.pdf
9. Lehman, C. F., "A Bad Bet: Why Sports-Gambling Tax Revenue Disappoints," Manhattan Institute, June 18, 2026. Verified verbatim: "conservative estimate that sports-gambling legalization increased problem gambling by just half a percentage point (from 2.8% to 3.3%), which amounts to an additional 46,523 people with gambling problems" (baseline: Ohio's 2022 state gambling survey); report cites McCarthy, Taylor, and Wilbur finding a 79% increase in problem-gambling hotline calls and a more-than-fivefold rise (0.17% to 0.95%) in the share spending over 1% of income on gambling. manhattan.institute/article/a-bad-bet-why-sports-gambling-tax-revenue-disappoints
10. NCAA, Sports Betting Activities Survey, April 2023 (n=3,527, ages 18-22: 58% bet; 67% on-campus; 41% own school; 35% student bookmaker). ncaa.org/media-center-ncaa-releases-sports-wagering-survey-data/
11. Siena College Research Institute / St. Bonaventure University national poll, February 18, 2025 (48% of men 18-49 hold active online sportsbook accounts; 54% of bettors bet at least weekly). sri.siena.edu/2025/02/18/
12. Pew Research Center, "Americans increasingly see legal sports betting as a bad thing for society and sports," October 2, 2025 (31% of under-30s bet in past year; 36% of under-30 men). pewresearch.org/short-reads/2025/10/02/
13. National Council on Problem Gambling, NGAGE 2.0 national survey (67% of adult sports bettors under 45; 46% of under-35s bet on sports). ncpgambling.org/wp-content/uploads/2025/01/NCPG_NGAGE-2.0-2021_National_Detailed_Report.pdf
14. SAFE BET Act (proposed federal legislation: live-event ad ban, deposit limits, AI-tracking ban), as described in the Siena/St. Bonaventure release, February 2025.
© Reyes Castillo.