It was in many ways a secret life. The whole thing was completely secret for years and years and years. It was something where you couldn't tell anyone. We were totally paranoid about security and the casinos could throw you out at any time - so it was all hush hush. From the beginning in 1979, I never talked about it publicly until after the movie came out in 2008. Among those who knew - my family and friends who became investors - we called it "Camp Cambridge."In 1979 I lived with four roommates in the kind of crummy housing you live in after college when you have your first job but haven't gotten paid yet. It was a Cambridge, Massachusetts "triple decker" with a missing front door and a list of over 100 health and safety code violations. It was just over half a mile from MIT. One day a kind of rough-looking guy with a backpack who wasn't too clean showed up at our apartment door looking for Leo Gordon. I'd never heard of any Leo Gordon, and I tried to get rid of him. Eventually my roommate Geoff came home and he knew Leo, it turned out Leo’s name was on our lease. The guy at the door, Dave, used to live in our apartment years ago, and because Geoff knew Leo and Geoff thought Dave knew Leo, Dave got to sleep on the floor in the living room. Dave was coming from Las Vegas to meet some MIT people about playing blackjack for money. I was convinced the whole thing was a fraud. You can go into a casino and win money? Consistently? With a "secret system"?Yeah, right.Dave had a whole story of playing blackjack "professionally" in Las Vegas on a "blackjack team" with Ken Uston - whom I'd never heard of. Dave talked a lot about blackjack and of anarchy and anarchists and creating an anarchist commune - it was very all entertaining in a "you aren't staying long, are you?" sort of way, and I looked forward to his departure. A couple of days later he introduced me to JP, who was 'the real deal' from a mathematical point of view. JP had a master's in mathematics from MIT, he had a high paying job at a tech company in Kendall Square, an elegant recursive computer model (which I understood), he had the numbers, he had all the math, and he had a copy of Richard Epstein's The Theory of Gambling and Statistical Knowledge (which became like the bible to us). In a few days he convinced me that there was a real, mathematical basis to this outrageous story, that there was a real, statistical likelihood of winning money, and that it was not as easy as it sounded. He gave the basic strategy and the sheet of numbers and I started memorizing and practicing. First, a few of the key points of blackjack. You want your cards to add up to as close to 21 as possible without going over 21. The dealer has the same goal. If you are closer to 21 than the dealer, you win. The big difference is that the dealer must follow a set rule - if the dealer's cards total less than 17, the dealer must take an additional card. When the dealer has 12 or more, getting a "10" card means the dealer exceeds 21 and loses, so the remaining players win. So when the deck has more ‘ten cards’ (10, Jack, Queen, King) this favors the player. When the deck has more small cards (2,3,4,5,6) it favors the house - 12 or 13 or 14 or 15 plus 6 is still OK for the dealer.The basic idea of card counting is to know whether there are more small cards left or more big cards. So you keep a running total of small cards vs large cards - each small card you see is plus one and each 10 card you see is minus one, and the count is running total. For example, the running count might be plus 6 which means 6 more small cards have been dealt than large cards. If you are playing against a six deck shoe, and half the shoe (three decks) has been dealt, that +6 is divided by the three remaining decks. This gives you the average count per deck, which is +6/3 = +2, also called the "true count". The larger the count, the greater your advantage, so you more bet. The smaller the count the less you bet. Negative count bets table minimum. Wikipedia has a more detailed explanation – card countingAs a result of a lawsuit, there was a famous experiment in Atlantic City where people could openly count cards and the casinos could not throw them out [Unsurprisingly, this experiment only lasted a couple of weeks.] It was an opportunity too good to resist, so we went, we five. I had not yet ‘qualified’ so I didn’t play that first bank. And although some of the guys had jobs, I was the only one with money. The starting bank was just over $5,200 total, I put in about five thousand.Going to Atlantic City was chaotic. We had a $135 per week "hotel suite.” Technically it was a suite, it had 4 or 5 beds in a big L shaped room. It was a run-down four story building, you had to walk up four flights of stairs. Atlantic City was a real dump at that time: mostly buildings that hadn't been maintained in decades. Everything outside the three casinos was very cheap. Every blackjack team in the world was there for the experiment, including all the people who had written books, all the people who were banned from playing everywhere, and the Czechs who kept multiple racks of purple five hundred dollar chips in front of each player at all times. At that time the casinos were so crowded that the $25 tables were the only ones at which you could - sometimes - get a seat. At the beginning, our bank was too small to sit down and play the $25 tables at an acceptable level of risk (according to the Kelly Criterion). When there were empty seats at a table one of us would stand a few feet from the table and "discreetly" keep the count. If the shoe went to +2 [true count] or better you would signal - put your hand on the back of your neck - and one of our roaming players would sit down and bet $25 (or more, depending on the count). At the end of the week we were slightly up, and most of us had to go back to work, Dave stayed and doubled our bank in a few more days. Suddenly we had a business, and came up with a business model. We threw money in the bank, we played until we doubled the bank, and then we split up the bank, dividing half the profit among the players and half among the investors with a few expenses off the top. We started recruiting and training players – mostly MIT students. We were all broke, every dollar we made playing went back into the bank, and every weekend we piled into one car and drove for five hours from Cambridge to Atlantic City, stayed in the team apartment and tried to live on the $5 a day per diem the team paid. Five dollars paid for one meal (usually a calzone) unless you paid to ride the jitney bus to and from the casinos. There could be 8 to 12 people sleeping in that apartment on Friday and Saturday night.As we won and our bank got bigger, everything got easier, especially when we could finally just sit down at the start of a new shoe and bet $25. That was a landmark day.By trial and error, we figured you could play really well for an hour—that was your estimated attention span in the pressure cooker environment of the casinos. You would go to Resorts (for example), walk around until you found a table that looked good, maybe wait for a positive count, then sit down and play. We were all too young and tried to play inconspicuously, but we were 20-25 year old kids with tens of thousands of dollars, and it stood out. After you had played for about an hour, you cashed out and went to the men's room where you sat in a stall and counted your money and filled in your player sheet, then you walked to the next casino (there were only two or three at the beginning) and did it again. You'd get to Atlantic City Friday night at 10 or 11 pm, play about an hour at each casino then go to sleep until the shift changed. That was the rule - one hour of play during a given shift at a given casino - you didn't want to stay anywhere long enough to be memorable - so you would go to the casino once during each of the three daily shifts. Sunday night at 8 or 9 or 10 or 11 pm we'd drive back to Boston ( or New York). We had check-ins where you'd meet the banker at certain times— 11 pm or 2 am - usually in a video gaming arcade on the boardwalk. You'd play Galaxian (a video game) and when the banker arrived you'd report your cash position, and the banker might adjust our betting using the Kelly Criterion. After a year or so we started to have "big players” as a disguise. A counter would sit at a table and flat bet $25 every hand. The big player would be ‘a real adult’ – someone over 50 - someone's uncle for example. The big player had been taught basic strategy and how to read our signals on betting. They barely had to pay attention - they could chat with the pit boss, flirt with the cocktail waitress. This dramatically reduced the rate of us getting barred. We kept on doubling our bank and eventually accumulated over $100k. From the beginning, we had trouble running the team. One of the founders had a job on Wall Street at Citibank and he put together an entire NY contingent. However, some of the NY guys played a different counting system - one team played a system called Hi Opt 1, one played a system called Hi Low - we were all doing different stuff. It wasn't going smoothly, there were fights and arguments. I don't remember how we (or JP) found Bill Kaplan, but once Bill arrived and we hired him to manage the team he got us all onto the same page. He wrote a business play, he raised serious capital ($500,000, in cash), we had simple rules and simple procedures. Remember, this was an all cash businesses.With Bill in charge, some of the NY guys stayed, some split and formed their own team.Playing in a casino was a real grind. We all had "real jobs" or were students so we were fitting blackjack in on the side. It wasn't glamorous and it wasn't interesting—just follow the formulas, DON'T make mistakes. And you could play perfectly and lose $3/4/5,000 in a night—that was hard to watch. Just like when flipping a coin it is 50% chance it will be heads, but its not that rare to throw 3,4,5 heads in a row. In the long run it all evens out, but in the short run its called ‘variance’ and it happens in blackjack. You could play perfectly and lose hand after hand after hand for no apparent reason except a bad run of the cards. There was also a certain depressing aspect to it. The casinos have no windows, full of smoke, full of people drinking. You'd quickly see that most people in the casinos had no idea how to play and were just giving money to the casinos. People too drunk to sit on a stool betting $25 or $50 a hand - they might be sitting next to you. Since it was all statistical the luck and gambling elements seemed minimized it quickly became dull. But high stress – don’t make mistakes, don’t let anyone know you are counting, don’t act suspicious, don’t let the casino see how much you are winning, don’t get pick pocketed, don’t draw attention. And we were always counting cash—you'd count your cash 7,8,9 times a day. Bills smaller than $100 were “lettuce” – annoying to have. You'd be walking around slummy Atlantic City with $15,000 or $20,000 in cash spread out through your pockets. At 2 am you’d be hanging out playing Galaxians in the video arcade waiting for the banker to arrive for a check-in, and worrying about getting your pocket picked or worrying about getting mugged. At night in the apartment between shifts, people would be counting cash – there could be over $100,000 on the table, and people would be sleeping or discussing (or arguing). Sometimes the mathematical discussions would be in base 13 [a deck of cards has 4 suits of 13 cards each] — there were no personal computers back then so people did real math, or at least talked in math. [The real computer models were back in Boston / Cambridge.]During the week in Boston there would be practice sessions. A practice session lasted a couple of hours of play. You'd practice counting 6 decks - the test was 6 decks in under 2 minutes, and then you'd play a couple of shoes. If you didn't play perfectly, you couldn't "work" that weekend. Friday was 'visit the safety deposit box' day.Some people tried playing in Vegas, playing in Monte Carlo. That meant learning different rules, different numbers. Monte Carlo was easier to play, but the dealing was much slower - fewer hands per hour. Vegas was tough.Back then you really had to have a computer model to have the correct basic strategy for a given rule set, and to calculate the correct true count at which you would altering play for each hand combination. We were lucky, we had computers. [ Ed Thorpe, who "invented" (or quantified) card counting** in 1962 did most of the math by hand] On top of that we had to figure out how to set up and run a team. How to recruit, train and qualify counters and "Big Players" and manage an all cash business and prevent skimming, theft, and loss. Today it's all been worked out, and it's all on the net. Bill had experience and common sense – and some times it must have been like herding cats for him.It was very lucrative. No one talked about counting cards at all unless they couldn’t play anymore. People like Ken Uston only wrote books when they had been banned so many times and were too famous / too well known to play. We would have preferred that Uston and Revere et al. would have not written their books - we didn't want any publicity.At first the casinos were new, most of the employees were new, and if you could count cards, there was more money to be made playing than working for the casinos. I remember when the first 'counter catchers' started working for the casinos - we watched the people in the pit, and tried to figure out who the counter catcher was and what they looked like. You really needed a Big Player, we were all still way too young for the amounts of money we played. The casinos hated us. We believed that the press around card counting and the knowledge that the game could be beaten made blackjack the most popular game. We could tell that 99% of the players had no clue how to play, so it seemed like a big win for the casinos. However the casinos knew we were taking money off their bottom line. So in addition to trying to ban or bar counters, the casinos experimented with rule changes and game changes to take our edge. So at the same time, JP was always experimenting, modifying and running his model. The wildest thing he came up with – but didn’t invent – was ‘shuffle tracking’ – each casino has a precise system for shuffling a shoe of cards. JP came up with a system to track cards through the shuffle – we called it the 'non-random shuffle' – we played it for a while, but that was a bitch, it was hard to do well. There were certainly difficulties of running an "all cash business": what do you do when you think someone is skimming? (it happened). Once there was one picked pocket. Personalities and personality conflicts - really smart people who wanted to play their own method, or do it their own way. One of "our founders" liked to play based on the first and second derivative of the true count, rather than the true count itself, since there was no statistical basis for this and he wouldn’t stop doing it, we had to ask him to leave. Eventually every casino had multiple employees who could count (and catch counters). Anyone who knows how to count can spot every other counter at the table. Casinos have modified the rules of the game, and increased the number of decks in the shoe and adopted shuffling machines and so on and so on. And the counter’s edge became smaller and smaller, and it became harder and harder to play without being barred. They know everything about using Big Players and shills and signaling. I played professionally for the first few wild years. It was lucrative, but wasn’t fun, and it became harder and harder to commute to Atlantic City for sleepless weekends grinding it out at the tables. Finally I got a job in Rochester, NY, and that was just too far away and I stopped playing. But the team went on, grew to 80 players. The edge became smaller, it became harder, until they finally stopped playing. And then one day in some bar at a tech convention, Jeff Ma told some guy the story, and that guy was Ben Mezrich who wrote Bringing Down the House, and eventually they made the movie “21”*.And now you know the rest of the story.*FWIW, the book is a highly fictionalized version of Jeff Ma’s story.** see comment to this answer (below) by Edward A. Curley Sr. as to who really "invented" card counting. Thorpe built on the work of others, and had the advantage as he was the first with access to a modern computer.