More than four hundred years of card play, mathematical breakthrough, casino warfare, and unlikely fortunes — traced from a French aristocrat's drawing room to a browser window. Here's how a game called Vingt-et-Un became the game called blackjack.
Blackjack has one of the strangest origin stories of any casino game. It started as parlor entertainment for European aristocrats, crossed the Atlantic in the suitcases of French colonists, took its name from a promotional gimmick in early-twentieth-century Nevada, was almost killed off by a mathematician with an IBM computer, and survived by reinventing itself half a dozen times. The game you play today is the latest version of a story that has been running for more than four hundred years.
What follows is the through-line — the people, the dates, the moments where the game changed. Where the historical record is clear, we say so. Where it's contested or fuzzy, we say that too. There are gaps. That's what four centuries of card-playing does to a record.
The earliest documented ancestor of modern blackjack is Vingt-et-Un — literally "twenty-one" in French — played in the gambling houses of Paris in the early seventeenth century. By the mid-1700s it was well established as a parlor game across the French aristocracy.
But "earliest documented" doesn't mean "earliest existing." Cards that count toward a target total were a common game family across early modern Europe. Italians played Sette e Mezzo (Seven and a Half) with a 40-card deck and a target of 7.5. The Spanish played Trente-un (Thirty-one). The French had Quinze (Fifteen). Most card historians believe these games cross-pollinated each other for decades before Vingt-et-Un emerged as the version that stuck.
Adding to the murk: Miguel de Cervantes — yes, the Don Quixote Cervantes — wrote a short story called Rinconete y Cortadillo around 1601-1602, in which two card sharps play a game called Veintiuna ("twenty-one" in Spanish) with rules that sound suspiciously like modern blackjack: target of 21, Ace counts as 1 or 11, players take cards until they decide to stop. Whether this Spanish version preceded the French Vingt-et-Un, descended from it, or developed in parallel is unknowable from existing records.
What we can say with confidence: by the late 1700s, a card game called Vingt-et-Un, with rules recognizable to any modern player, was being played across France. From there, it traveled.
The Atlantic crossing came in stages. French aristocrats fleeing the Revolution brought card games with them to Louisiana in the 1790s. Vingt-et-Un took root in New Orleans, where French was still widely spoken and gambling was a normal feature of social life.
From New Orleans, the game spread up the Mississippi on the riverboats — those floating gambling halls that were a defining feature of nineteenth-century American culture. It moved west with the gold rush, settling into the saloons of San Francisco, Denver, Tombstone, Dodge City. By the time of the Civil War, the game was popularly known in English as "twenty-one," with rules essentially identical to Vingt-et-Un.
The American frontier version had distinct edges. Saloons were not regulated, dealers were not always honest, and the games often featured side bets, bonuses, and rule variations that drifted from establishment to establishment. The standardization that modern casino-goers take for granted didn't exist yet. A "twenty-one" game in San Francisco might play very differently from one in Kansas City.
Here is the moment when the game got its modern name — and the story is, like much of card history, partly legend.
The most-cited account: sometime around 1910-1915, a casino somewhere in the western United States (sources variously say Nevada, Indiana, Indianapolis, Evansville — historians don't agree) offered a promotional bonus payout. If a player's first two cards were the Ace of Spades plus either of the black Jacks (Jack of Spades or Jack of Clubs), the payout was 10:1 instead of the standard even money.
Players started calling the bonus hand a "blackjack." The promotion died — bonus payouts are expensive and casinos quickly figured that out — but the name lingered. By the 1920s, "blackjack" had become a common nickname for any natural 21 (Ace + 10-value), and by the time Nevada formally legalized gambling in 1931, the name was effectively the standard. The original promotional structure was long gone; the word it created is still here.
This is a perfect example of how casino history actually works: a marketing gimmick gives a thing a name, the gimmick dies, the name survives.
By the mid-twentieth century, blackjack had become one of the most popular table games in American casinos. It was assumed to be a game of luck with mild skill — a coin-flip with embellishments. That assumption survived intact until four soldiers stationed at Aberdeen Proving Ground in Maryland decided to actually do the math.
Their names: Roger Baldwin, Wilbert Cantey, Herbert Maisel, James McDermott. All four were US Army personnel with mathematics training. Using only the desk calculators available to them (no electronic computers — these were mechanical adding machines), they spent three years calculating optimal play for every starting hand against every dealer upcard.
Their paper, "The Optimum Strategy in Blackjack," appeared in the Journal of the American Statistical Association in September 1956. It was the first published computation of basic blackjack strategy. They followed it with a book, Playing Blackjack to Win, in 1957.
The book sold poorly. Casinos dismissed it as a curiosity. Players who heard about it usually thought the math sounded too complicated to bother with. The Four Horsemen had quietly done one of the most important pieces of work in casino game theory, and almost no one noticed.
That was about to change.
Edward Thorp was a young mathematics professor at MIT in the late 1950s. He read the Baldwin paper and saw something the Four Horsemen had missed: if basic strategy is computable, then card-counting strategies should be too. Because cards are dealt without replacement, the composition of the remaining deck changes as cards come out — and so should the player's decisions.
Thorp had access to an IBM 704, one of the most powerful computers of the era (about the computing power of a 1995 calculator, but science fiction at the time). He spent months running simulations that would have taken the Four Horsemen lifetimes on their mechanical adders. By 1961, he had a counting system that could theoretically give the player a small edge over the house.
Then he went to test it.
Thorp's trips to Reno and Las Vegas in 1961-1962 were partly tourist visits, partly cold experiments. He bankrolled himself, sat down at tables, and played. He won. Not millions — the trips netted around $11,000 across several sessions — but the principle was proven. The math worked. Blackjack could be beaten.
He wrote it up. Beat the Dealer was published in 1962 and updated in 1966. It became one of the best-selling gambling books of all time and is still in print. For the first time, the general public was told — in clear, accessible language — that casino blackjack had been mathematically conquered.
Casinos panicked. The first response was to change the rules: more decks, no doubling on certain hands, higher minimums for some table types, restrictions on side bets. Some of these changes stuck; some were rolled back after losing players got annoyed. The arms race had begun.
The decades after Beat the Dealer were a kind of mathematical Cold War between casino floor staff and an emerging community of counters. Each side iterated.
In 1963, electrical engineer Harvey Dubner published a simplified counting system — what we now call Hi-Lo — at the Fall Joint Computer Conference. Where Thorp's original system involved tracking specific card values, Hi-Lo collapsed everything into three buckets: +1 for low cards (2-6), zero for middle cards (7-9), -1 for high cards (10-A). The math became trivially easy at speed. Hi-Lo remains the most widely-used counting system more than sixty years later.
Lawrence Revere, a longtime professional gambler, published Playing Blackjack as a Business in 1969. Revere coined some of the visual conventions still used in blackjack instruction — colored strategy charts, simplified tables — and brought counting to a broader audience.
Stanford Wong's Professional Blackjack, published in 1975 and updated repeatedly thereafter, advanced the field by adding the concept of "wonging" (later named after him): not playing at all unless the count was favorable. Cantonese-by-birth, Wong became the most-quoted blackjack writer of the 1970s and 1980s.
Ken Uston, a former vice president at the Pacific Stock Exchange turned full-time blackjack professional, took counting in a different direction. In The Big Player (1977) and later Million Dollar Blackjack (1981), Uston pioneered the team approach: small bettors counting cards and signaling a "big player" who came in only when the count was high. When Atlantic City casinos refused to seat him in the late 1970s, Uston sued. The 1982 New Jersey Supreme Court decision in Uston v. Resorts International Hotel established that counting cards was a legal mental skill — not grounds for criminal exclusion. (To this day, Atlantic City is the only US jurisdiction where casinos cannot exclude players for skilled play alone.)
By the 1980s, counting was no longer a fringe pursuit. There were dozens of books, training tapes, computer programs, and seminars. The community had names, conventions, in-jokes. The casinos responded with more aggressive countermeasures: continuous shuffle machines (1990s onward), 6:5 payouts (became widespread in the early 2000s), face-recognition surveillance, shared blacklists.
The most famous chapter of advantage play started, fittingly, in a Cambridge dorm. In 1979, an MIT undergraduate named JP Massar and a Harvard alumnus named Bill Kaplan started running blackjack expeditions to Atlantic City. Kaplan had previous experience: he'd taken a year off from Harvard Business School to test counting systems in Las Vegas (with $1,000 of his savings; he came back with around $35,000).
The team they built was structured like a small hedge fund. There were investors, treasurers, signed contracts, performance reviews. Players were trained in counting, signaling, and "tells" — the body language casino staff used to identify advantage players. The team operated under various incarnations from 1979 through the early 2000s, eventually known collectively as Strategic Investments.
The total amount the team took out of casinos is contested. Estimates range from "several million" (conservative accounting) to "tens of millions" (the more dramatic versions). What's not contested: at their peak in the early 1990s, the team had a couple dozen active players, infrastructure for training, and a reputation that made them recognizable on sight to pit staff in Las Vegas and Atlantic City.
The story went public with Ben Mezrich's Bringing Down the House in 2002 — a book that's a tremendously good read and known by participants to be heavily fictionalized. The 2008 film 21, starring Kevin Spacey, drew from it. Both moved the MIT Blackjack Team from obscure subculture into popular mythology. Today they're probably the most recognizable name in the entire history of advantage play.
For a more accurate version of what happened, Bill Kaplan and others have given interviews and written articles distinguishing fact from Hollywood. The honest version is more interesting than the dramatic one. It involved spreadsheets.
The last three decades have shaped blackjack as much as any earlier period. Three forces in particular.
Online gambling. The first real-money online casino, InterCasino, opened in 1996, operating from Antigua under a license created by the Antigua and Barbuda Free Trade & Processing Zone Act of 1994. Microgaming, a software vendor, supplied the early platforms. By the early 2000s, online blackjack was commonplace; by the 2010s, live-dealer streaming let players play against an actual human dealer in a studio while watching real cards being dealt over video. The rules of online play often mirror physical-table rules, but the casino can adjust deck composition, shuffle frequency, and side bets in ways that aren't physically possible at a brick-and-mortar table. (For the math of how this affects house edge, see our house-edge page.)
Continuous shuffle machines and 6:5 payouts. Both technologies hit popular Vegas tables in the late 1990s and 2000s. CSMs reshuffle the played cards back into the deck continuously, eliminating any opportunity for card counting. The 6:5 payout — paying only 6:5 on a natural blackjack instead of the traditional 3:2 — quadrupled the house edge on otherwise-identical tables. Together, they reflected the casinos' long answer to Beat the Dealer: design the game so the math no longer matters.
Don Johnson. In 2011, a former professional horse-racing handicapper named Don Johnson won approximately $15 million from three Atlantic City casinos — Tropicana, Borgata, and Caesars — over a six-month run. The headline news was that he didn't use counting. He used basic strategy and aggressive rule negotiation: he was a recognized "whale" (high-volume player), and he'd negotiated favorable rules and a 20% loss rebate before sitting down. The casinos had assumed they'd win anyway. They were wrong. Documented by Mark Bowden in The Atlantic in 2012 ("The Man Who Broke Atlantic City"), Johnson's run is the most-cited modern example that the house can be beaten — even by a player using only the rules they're given.
Walk into a Las Vegas casino in 2026 and you'll find a strange mixed picture. Many of the lower-minimum tables pay 6:5 — these are tourist traps designed to capture players who don't know the rule difference. The high-limit rooms still pay 3:2 with reasonable rule sets, often with single or double-deck games that would have made Edward Thorp's heart leap.
Continuous shuffle machines are common but not universal. Counting is still possible at hand-shuffled tables; it's just harder to find them and the pit attention is more sophisticated than it was in Ken Uston's day. The total number of active card counters worldwide is probably small — under a thousand serious professionals at most — but they exist, and they still win.
Online, the picture is different. Licensed RNG blackjack offers a clean, audited version of the game with no counting possible (the deck is reshuffled every hand). Live-dealer online tables sometimes allow counting in principle, but the deck penetration is so shallow that the edge is minimal even for skilled counters. The accessibility, though, is unprecedented: anyone with a phone can play the same standard-rules blackjack that was being played at a Reno table in 1961 when Edward Thorp was running his experiment.
And then there's free-play. Sites like this one — every game free, no real money, no payouts — exist as a third stream that didn't exist before about 2010. They're used by players learning the game, players practicing strategy, and players who simply enjoy the cards without wanting to lose money. We're a small part of a long, strange story. Four hundred years from Vingt-et-Un in a Parisian drawing room to here.
If you want to try the version that's evolved from all of that, the free tables are open. If you want to dig into the math that started the modern era, our basic strategy guide is the same chart Thorp helped compute on the IBM 704 in 1961, refined by sixty more years of simulation. The game is older than your country. It will probably outlast it too.
No single inventor exists. The game evolved over centuries from European card games — Vingt-et-Un in France, Veintiuna in Spain, similar games elsewhere. Cervantes references it around 1601, by which time it was already an established game. The modern American version emerged from this tradition through French colonists in Louisiana in the 1800s.
From a promotional bonus offered by US casinos around 1910-1915: a 10:1 payout when a player's first two cards were the Ace of Spades plus either black Jack (Jack of Spades or Clubs). The bonus was discontinued, but the name stuck. By the 1920s, "blackjack" referred to any natural 21.
Yes. In trips to Reno and Las Vegas in 1961-1962, Thorp won approximately $11,000 using the counting system he later published in Beat the Dealer. His winnings weren't huge by professional gambling standards — but they were enough to prove the system worked. The book that followed transformed the game.
Contested. Conservative estimates from team participants suggest "several million dollars" over the team's twenty-plus years of operation. The popular figure of "tens of millions" comes mostly from Mezrich's dramatized account in Bringing Down the House. The truth is closer to the low end.
Real-money online blackjack first appeared in 1996, when InterCasino launched under an Antigua license. Microgaming and other software vendors supplied the platforms. Live-dealer online blackjack — where you watch a real dealer over video — became common in the early 2010s.
At hand-shuffled tables, yes — the math hasn't changed. What has changed is the casino response: aggressive monitoring, continuous shuffle machines, 6:5 payout tables, and shared blacklists make counting harder to do profitably. Probably a few hundred to a few thousand serious counters operate worldwide. Our counting guide explains the modern landscape.
For mathematical history, Edward Thorp's own Beat the Dealer (still in print, 1966 revision) is foundational. For the modern advantage-play era, Stanford Wong's Professional Blackjack (1975, updated) covers most of the technical ground. For social and casino history, Don Schlesinger's Blackjack Attack includes useful historical chapters. Bringing Down the House is entertaining but heavily fictionalized — read it as a novel, not as documentation.