The MIT Blackjack Team was not one team, and it was not a heist. It was a series of card-counting partnerships run out of Cambridge, Massachusetts between 1979 and 1993 (formal businesses with outside investors, salaried players, mandatory training and audited hours) that beat casino blackjack legally, at scale, for over a decade. This page is the documented record: what is proven, what is claimed, and what is invented. Every number carries its source, and every number we could not verify says so.
The story in 30 seconds. Card counting is legal. It’s arithmetic, and no law prohibits thinking. So teams split the thinking across roles the pit could never connect: a Spotter counts while betting the minimum, a back-spotter carries the signal, and a Big Player arrives “drunk and lucky” only when the deck is hot. The best-documented operation (Strategic Investments, June 1992) was a Massachusetts limited partnership: ~$1M raised, ~80 players on the roster, pay by hours × expected value, profits split 50/50 with investors. It hit its win goal and dissolved on schedule. What ended the era wasn’t muscle. It was a private detective agency with a photograph.
The record, in numbers
| The number | What it is |
|---|---|
| $89,000 | the first formal bank, opened 1 August 1980 with ten players |
| 10 weeks | time for that first stake to more than double |
| ~$1,000,000 | raised by Strategic Investments LP, June 1992 |
| ~80 | players on the SI roster; ~30 in play simultaneously worldwide |
| 50/50 | profit split between players and investors at win goal |
| 31 Dec 1993 | SI pays out and dissolves, on schedule |
| Sept 2005 | Griffin Investigations files Chapter 11 after losing to two card counters |
Is card counting actually illegal? No, and that’s the whole story
Card counting breaks no law anywhere blackjack is played. A shoe that runs rich in tens and aces favours the player; noticing that is paying attention, not cheating. What casinos can do is refuse your business, and they watch for the one tell every lone counter gives off: minimum bets that surge when the deck turns good.
The teams’ real invention was never the math. It was making the math invisible by splitting it across people who never appear to know each other.
Counting is not the only legal way to beat a blackjack table. In 2010 a former gaming regulator did it without counting a single card, by negotiating the rules before he sat down.
How the MIT Blackjack Team actually worked
Four roles, defined the way the advantage-play literature defines them, real technical terms, coined a decade before MIT scaled them:
The Spotter
Sits at the table for hours betting the minimum, counting every card out of the shoe. His bets never rise. The moment they track the count, surveillance has him. His only output is a signal.
The back-spotter (our film calls him the Caller)
Watches from the floor without playing. When a spotter marks a hot deck, this is who brings in the money. The documented term is back-spotter or Controller.
The Big Player
Arrives loud, expensive, apparently half-drunk, and bets big from his first hand, but only at tables that are already hot. To the eye in the sky his wins look like luck, because his bets never track a rising count at any one table.
The Gorilla: the decoy who thinks nothing
The role the movie never showed: a big player who does not count at all. He performs the drunk high-roller and bets exactly as signalled. The most conspicuous man in the room, doing the least thinking, which is precisely why he was safe.
The signal layer: teams conveyed the running count by code word, and real teams rotated their signals regularly, a discipline the film skipped. The specific code-word lists that circulate (“tree” = +1, “stool” = +3) come from film-promotion material, not team documents; we flag them as such.
The $1-million company behind the “students”
This is the part that survives every fact-check, and it is better than the fiction: the operation was a business, and the business design (not the counting) is why it outlasted every other team.
Capital-managed like a fund. Banks ran to a pre-agreed win goal; on hitting it, the bank was “broken” and paid out. Strategic Investments even dissolved on schedule, 31 December 1993, a planned wind-down, not a collapse.
Pay was decoupled from luck. Players were paid on hours played × simulated win rate, expenses covered, the same pay whether the table ran hot or cold. Aponte: “overall compensation was shared 50/50 between players and investors.” That one design choice removed every incentive to chase losses.
Bet sizing followed the bankroll, Kelly-style. Kaplan: “You are always betting in proportion to your capital. If you lose money, you lower your bet.”
Entry was gated. Formal management, mandatory training, logged play, a checkout exam. Aponte: “No one ever passed the checkouts on the first try.”
Returns? By Kaplan’s own account, investors saw returns he has described in different tellings as ~100% and as over 250% annualised. The two figures conflict; both are his, neither is audited. We print the conflict rather than pick the flattering one.
And the mentor? There was no professor recruiting in a lecture hall. Bill Kaplan was a Harvard MBA who had already run a winning Vegas team in 1977 before backing the MIT players, as an investor, on condition they run it as a business. The film’s professor is a composite of Kaplan, John Chang and J.P. Massar.
Where the method really came from, a timeline
| Year | Event |
|---|---|
| 1962 | Edward Thorp publishes Beat the Dealer (the proof that blackjack is beatable |
| ~1971 | Al Francesco invents the Big Player team system |
| 1976 | Ken Uston’s The Big Player tells the world; casinos start hunting the pattern |
| 1977 | Kaplan’s pre-MIT Vegas team) by his account, ~35x in nine months |
| 1980 | First formal MIT bank: $89,000, ten players, 1 August |
| 1984 | ~35 players, ~$350,000 capitalisation |
| 1992 | Strategic Investments LP raises ~$1M; roster grows toward ~80 |
| 1993 | SI dissolves on schedule, 31 December; successor banks play on |
| 1995 | Biggest documented trip: ~$500,000, Super Bowl XXIX weekend (successor team) |
| 2005 | Griffin loses a Nevada jury verdict to counters Grosjean & Russo (June); files Chapter 11 (September) |
| 2008 | The film 21 releases; MIT’s own newspaper fact-checks the book behind it |
MIT’s innovation was management, not method. The system was a decade old when they scaled it. What they added was the fund structure, the training gate, and the EV payroll.
What the movie “21” got wrong
| Hollywood said | The record says |
|---|---|
| A charismatic professor recruits six students | No professor existed (a Harvard MBA investor and ex-player managers; the mentor is a composite of three real people |
| Six kids and a duffel bag | ~80 on the SI roster, ~30 playing simultaneously, 22 separate partnerships in the 1980s alone |
| A heist run on nerve | A registered limited partnership with a win goal, a payout, and a scheduled dissolution |
| Players get rich off their own hot hands | Paid hours × simulated EV) identical pay whether the table ran hot or cold |
| Vegas excess: drinking, strip clubs | Forbidden, grounds for dismissal. Chang: “You’d be considered such a losing sucker” |
| The counter is the table’s hero | The best-disguised player counted nothing, the Gorilla performed while others thought |
| Cash in hollow crutches and casts | Chips in carry-on bags; inventory left in Vegas. Jeff Ma never heard of the crutches outside the book |
| Vegas justice = a back-room beating | Never happened to anyone. A detective agency and a database ended it (administratively |
| A white leading cast | The real leads were Asian-American: Jeff Ma ’94, Mike Aponte ’95, John Chang ’85), and the book noted their appearance was operational cover |
| “Based on a true story” | Credited “inspired by true events”, the studio’s own hedge. The author to MIT’s paper: “I took literary license to make it readable” |
That casting row deserves its own sentence: Mezrich said in 2005 the studio’s plan was that “most of the film’s actors would be White, with perhaps an Asian female”, and the irony is that the real team’s diversity wasn’t incidental. They were not who pit bosses expected to see betting a thousand a hand.
How was the MIT Blackjack Team caught? A photograph.
Nobody was dragged into a basement. Casinos handed the problem to Griffin Investigations, founded 1967, publisher of the “Griffin Book,” the industry blacklist sold by subscription to roughly half the major casinos in America. Griffin worked it like an intelligence case: the same faces, different cities, winning too consistently. A Griffin investigator bought the names and photographs identifying team members; among the material, by the team’s own account, an MIT senior-yearbook photo sat in the file beside two surveillance stills. Names attached to faces, faces to a network, and every subscribing casino could stop a player at the door before the first chip hit felt.
The team wasn’t outplayed. It was indexed.
The coda the movie could never write: Griffin itself went down to card counters. In June 2005 a Nevada jury found for players James Grosjean and Michael Russo in a defamation suit over fabricated Griffin Book entries; that September, Griffin Investigations filed for Chapter 11. The blacklist that ended the golden age of team play was ended, in court, by two of the players it listed.
Coming next: the Griffin file deserves its own story, the database that ended advantage play’s golden age, and the two counters who bankrupted it. Follow @bigwinstories. That film is in production.
How much did the MIT team really win? The honest answer
We don’t know, and neither does anyone quoting $50 million or $100 million. The operation ran as private partnerships across two decades and never filed public results. Every aggregate figure in circulation traces to affiliate-site guesswork with no methodology. We publish what’s defensible and grade the rest:
The BigWinStories Confidence Ledger
| Figure | Grade | Why |
|---|---|---|
| $89,000 first bank, Aug 1980 | DOCUMENTED | Consistent across every account, no competing figure |
| ~$1M raised by Strategic Investments, 1992 | DOCUMENTED | Massachusetts partnership record |
| ~$500,000 team trip, Super Bowl XXIX 1995 | ATTRIBUTED (Aponte | First person, on the record; the team’s best trip, not his personal win |
| $200,000 Aponte’s personal best trip | ATTRIBUTED) Aponte | His own figure; worst loss ~$60,000, largest single bet $15,000 |
| “$400,000+ in one Vegas weekend” | ATTRIBUTED, secondary | Single-sourced; possibly the same event as the Super Bowl weekend. We refuse to count it twice |
| >$80/hour average player earnings, 1980s | ATTRIBUTED, team accounting | Self-reported, unaudited. Kaplan’s “$500/hour” is a ceiling, not an average |
| ~100% vs >250% annualised investor returns | ATTRIBUTED, Kaplan, conflicting | Both figures are his, from different interviews. Unaudited. We print the conflict |
| Kaplan left with >$10M | ATTRIBUTED, Inc. (2008) | Single outlet, sourced to Kaplan |
| “$50–100M lifetime team winnings” | REJECTED | No credible source, no methodology. Anyone printing this as fact is guessing |
| $300,000/year per player | REJECTED | Realistic average was closer to ~$25,000/year |
Grades: DOCUMENTED = stated as fact · ATTRIBUTED = someone’s own account, printed with its name · REJECTED = no defensible source. Full method: How We Verify.
The same grading runs across every category on the BigWinStories leaderboard: the biggest verified wins in slots, poker, sports betting and advantage play, plus the famous figures that do not survive a source check.
FAQ
Is card counting illegal?
No. It is legal everywhere blackjack is played. Casinos are private businesses and may refuse or eject suspected counters, but no law is broken by thinking.
Did the MIT Blackjack Team really exist?
Yes, as a series of teams and partnerships from 1979 into the 1990s. The best-documented, Strategic Investments, was a Massachusetts limited partnership formed June 1992 with roughly $1 million from outside investors and a salaried roster.
Who was the real leader of the MIT Blackjack Team?
No single professor. Bill Kaplan (Harvard MBA) organised and capitalised the best-known operation; John Chang and J.P. Massar managed play. The film’s mentor is a composite of all three.
Is the movie “21” accurate?
The core idea (teams beating blackjack with signals) is real. The professor, the back-room beating, the smuggling props, the aggregate winnings and most of the cast profile are invention. The book’s own author conceded literary license; co-manager John Chang called its stories “so exaggerated they’re basically untrue.”
How was the MIT Blackjack Team caught?
Griffin Investigations connected faces across casinos and matched them to identifying photographs, the team’s own accounts describe an MIT yearbook photo in the Griffin file. Once the database existed, the teams couldn’t get a seat.
How much did the MIT Blackjack Team win in total?
No defensible total exists. The partnerships never published results; circulating figures ($50M–$100M+) have no methodology. Documented anchors: roughly $1M raised in 1992, a ~$500,000 best weekend in 1995, and investor returns claimed (not audited) by Kaplan.
Does card counting still work today?
The math still works; the environment mostly doesn’t. Continuous shufflers, 6:5 payouts, aggressive back-offs and shared surveillance databases have squeezed team play to the margins.
Sources
- The Tech (MIT), 8 Apr 2008, “Blackjack Team Book Embellished”
- The Tech (MIT), 1 Apr 2008, “’21’ Discriminatory Casting Unjustified”
- Mike Aponte interview, BlackjackInfo
- Wizard of Odds: “21: Truth and Fiction”
- Las Vegas Sun, 13 Sep 2005, Griffin files Chapter 11
- Inc., Aug 2008, “Luck Is for Losers”
- Boston Magazine, June 2019, the Kaplan profile
- Wikipedia: MIT Blackjack Team · Griffin Investigations
- Ken Uston, The Big Player (1976) · Million Dollar Blackjack (1981)
Cite this page: BigWinStories, “The MIT Blackjack Team: The True Story the Movie Buried,” bigwinstories.com/mit-blackjack-team-true-story (2026). Every figure carries its grade; quote the grade with the number.
Watch the 60-second film this story comes from (The Signal on YouTube) then follow @bigwinstories: the next film is the photograph that ended it all.
Were you there? Team member, surveillance, pit boss, write to stories@bigwinstories.com (amount, date, venue, proof). Verified accounts enter the record; credit or anonymity, your call.
Gambling should be entertainment. If you’re concerned about your gambling, visit BeGambleAware.org.
Leave a Reply