Regression to the min

In which I become a fortuneteller

Mar 27, 2026

Regression to the mean is a natural phenomenon that you can see in aggregations of large numbers. But what happens when we tinker with nature?

I remember when California started its Lotto—1985. Back then, gambling was controversial. In general, we still saw people who ran gambling establishments as close to—if not actual—criminals. But the people spoke: the Lotto was approved by voters. According to 2022 research, “In search of new markets for its products, Scientific Games Inc., the creator of instant, scratch off lottery tickets, persuaded voters in California and six other states that lotteries offered a solution for stressed state budgets.”1

If this sounds familiar, it is: We have a sociopolitical problem, and instead of solving it through the legislative process it gets “solved” by sending more of people’s hard-earned money into private industry. 1985’s Lotto was part of a wave of “solutions” that featured creating new money-making schemes for political entities that depended on private industry to suck the money out of the pockets of largely lower-income Americans. Americans hate taxes because they can see the money going to the government. But Lotto tickets gamify school funding and mask what’s really happening: continuing inequity.2

Need to pay rent? Place a bet!

Fast-forward to 2026 and if I wanted to, I could have put a prediction thingy at the top of this post. (Note: I am staying intentionally ignorant about what this is or how to do it!) Substack has contracted with Polymarket so that if, for example, I wanted my readers to know whether people predict that I’m right that Jesus Christ will return this year (the Times says Polymarket puts that event at 4% likelihood), I can place a little widget at the top of my article. Don’t ask me how the money angle of this transaction works. As I said, I’m staying intentionally ignorant.

Encouraging undereducated and addiction-prone Americans to engage in gambling more often and more easily was clearly a bad bet right from the beginning. There’s a reason that the mafia liked “bookmaking,” as they called it then: it sucks people in, gets them hooked, and makes them vulnerable. According to the same Times article, the rise in easy online gambling—not just Polymarket but also sports betting, which is incredibly popular with men—has led to a rise in bankruptcy and (I could have predicted this one on Polymarket!) a rise in spousal abuse on days when debts are called in.

The Babblery: Visit our modern Tower of Babble in monthly interviews — CLICK FOR THE PODCAST

Tip jar

The problems with a gambling-based economy are well-known

I couldn’t come up with a snappy header for this section, because let’s face it, we all knew what would happen when everyday Americans had easy access to—were even encouraged to take part in—gambling markets. Gambling is addictive, and as any addict can tell you, the first step toward conquering your addiction is changing your behaviors: no longer associating with fellow addicts and avoiding the places where you practiced your addiction.

When addiction moves online, the temptation goes everywhere with you. The addict can be attending a tense family gathering where they should be in the room hashing out a disagreement, and instead they retreat to the bathroom where they can vote on the Second Coming and receive their little dopamine rush. In the past, an addict had to go to a legal gambling enclave: reservation casinos (another unfortunate crack in our protective armor), the horse track, or if they were in a city, a “den of iniquity” that they’d learned about through the grapevine. Addicts could enlist professionals and loving family and friends to help them resist their addiction.

But now our Grifter-in-Chief and his cronies want all of us to retreat into our bathrooms and get sucked into our phones. All the better for them on voting day if a Polymarket prediction tells us it’s rigged and we should just stay home.

Public problems on private backs

State lottos were started to help fund education, a public good that should be a public responsibility. Tribal casinos were started to bolster the income of impoverished tribes. Neither solution was the right one.

Education is a public good that, despite the protestations of the right wing, benefits all of us. A well-educated public is healthier, more equipped to pursue employment in any sector that needs them, and better-prepared to face the uncertain future that we’re all facing. A well-educated public is more likely to be gainfully employed and less likely to gamble.3

Similarly, tribal poverty is a situation that our governments created and should have solved. I admit that I feel a little bit of empathy for how tribes are getting back at the descendants of the white people who did this to them, but unfortunately, they’re also weakening all of American society. You can see the effect in their immediate areas: although the standard of living goes up for tribal members, “The negative changes include about a 10 percent increase in auto thefts, larceny, violent crime, and bankruptcy in counties four years after a casino has opened, and an increase in bankruptcies within 50 miles of a new casino.”4 And I am personally uncomfortable with the divisions that arise when people who are registered as “native enough” get a bonus, while their cousins don’t.

Online life has unbalanced natural variations

“Regression to the mean” is a statistical phenomenon that happens in any aspect of natural behavior. You might see one outlier—a basketball player has one fabulous night,5 a mild-weathered coastal town has one hard freeze in 50 years, a mediocre jazz player knocks out one fabulous solo.6 But then when a larger set of numbers come in, that outlier doesn’t lead to an overall change. The basketball player’s career stats regress to the mean, the coastal weather is not getting colder in general, and that mediocre jazz player? Her solos regress back to what they usually are, hopefully improving incrementally though she doesn’t do the woodshedding she really needs to do to succeed.

But the online world unbalances natural variation to the point that, I believe, people’s instincts are as confused as bats flying in ultrasonic interference. Our political system’s longtime suspicion of unfettered gambling, largely keeping it illegal and otherwise confining it physically, was a natural instinct. Anyone could sense that gambling wasn’t a social good. But online behavior doesn’t follow any of the old rules. It would be an unusual teenager who would get so obsessed with reading teen magazines that they would lock themselves in their rooms, scrolling through the pages ceaselessly until they disintegrated. Yet teens now do that with their phones without the physical limitations of paper. It would be an unusual situation for a father and husband to be able to spend excessive time at betting establishments without his family noticing it, but now they can’t know what he’s doing on his phone till it’s too late.

Telling our fortune

When I was in college, a friend’s grandmother sent her birthday cash in an envelope (we used to do that sort of thing!) and commanded her to use it frivolously. The friend invited me on an adventure: we bought lotto tickets and then went to a fortuneteller. As I remember it, all of my predicted fortunes were bust (I was supposed to have twins), but I didn’t need the fortuneteller to predict my lotto winnings.

We lost, of course.

I will become a fortuneteller in this sentence because I know something about human nature and statistics: online access to gambling does not lead to a happy ending. It will suck more money out of the pockets of working people and into the bank accounts of the uber-wealthy. It will destabilize healthy social behavior and lead to fractured families and friendships. It may line the pockets of some Substack writers, but even after writing this, I still don’t know how I’d make money off Polymarket and I promise I never will.

Also, I voted against the Lotto—equal education for all!

Discussion about this post

Ready for more?