• #### Worldwide Contacts

If you don't find your country in the list, see our worldwide contacts in:

# Bringing down the house

### Former professional card counter Jeff Ma's new book tells how he used statistics to beat the game of blackjack

We all have a defining moment in our lives – a moment that is a leap from hesitation to action, a single decision that sets us on a completely new course. For Jeff Ma – inspiration for the bestselling novel Bringing Down the House and the movie 21 – it was when he was a 22-year-old professional card counter, using math and statistics to beat the game of blackjack. He had perfected straightforward formulas that told him how much money to bet on each hand, and if he followed these equations correctly, he won. Until the night he lost \$100,000 in less than an hour.

Doubting his faith in statistics, he left the blackjack table to review the math and replay the hands. Confirming the odds were in his favor, he was faced with a decision to quit or carry on. With his faith confirmed, he returned to the table – winning back his initial \$100,000 and netting an extra \$70,000 for the weekend. His belief in statistics had got him out of the hole and he became a believer in the religion of statistics.

The following is an abbreviated excerpt from Ma's new book, The House Advantage, which provides insight into Ma's world of statistics and how they can be applied in business.

Basic strategy is a set of rules for optimal play in blackjack. Displayed in matrix format, it tells the player, based on his cards and the dealer's up card, exactly what action should be taken. It changes slightly based on the rules at the table, but as long as you are familiar with these rules and have them memorized, basic strategy reduces a casino's edge to less than 1 percent. It was developed in 1957 by four army technicians who used approximations of mathematical algorithms and then ran calculations on a desktop calculator to determine probabilities of all possible combinations of hands.

As part of the MIT blackjack team, I used math and statistics to beat the casinos legally. We were some of the most successful card counters in the world, and we believed in what we were doing because it always worked.

It may sound dramatic, but for us belief in the power of analytics and statistics was not unlike believing in God.

You need to have a long-term perspective and the corresponding financial commitment to make analytics pay off.

Forecasting team performance
Sports betting shares one very important similarity with blackjack: It is a game that its players may beat using analytics and statistics. Bob Stoll studied statistics at the University of California, Berkeley, and got his first taste of sports gambling during his sophomore year.

As Stoll continued the pursuit of his statistics degree, he saw parallels between what he was learning in the lecture hall and his newfound sports gambling hobby.

For a long time people have known that teams behave in strange ways throughout a season – like letdowns after a big win or suddenly playing well and bouncing back after a big loss – but until Stoll's work, no one had really found a way to predict it reliably.

And the proof is in his results. The proof is in a career that has yielded, on average, a 56 percent record against the spread – a record good enough to give him a business with thousands of subscribers and more than a million dollars in annual revenue.

From card counting to options trading
My first job was at an options trading firm in Chicago called O'Connor and Associates. It didn't matter to O'Connor that I had never taken a finance or business class at MIT. I learned later that this was by design, as O'Connor did not hire MBAs and preferred to work with a clean slate. So with no formal education in finance, every interview I had with the firm centered on a small line at the bottom of my resume. In the "Other Interests" section I had listed card counting.

The card-counting system was based on hard math that wouldn't change unless the rules of blackjack changed. Every bet we made had a theoretical edge that would be realized over time. There was very little risk as long as we had enough money to withstand the impact of variance. There was very little gambling to what we were doing in the casino. The trading system was another story. When I really broke it down, it simply fell short of the certainty of blackjack. But blackjack had me believing in an extremely high standard of analytics – a standard that was not met during my time at O'Connor.

The commandments of analytics
The first commandment is the importance of understanding variance, and it's a very difficult concept for people to become comfortable with. In 2007, Dr. Bob had a run where he went 5–32 (5 wins and 32 losses) and received hate mail from his clients. "The problem," he said, "is that people just don't understand variance." Yet Bob's clients began canceling their subscriptions and deciding not to follow him ... [but] win he did, to the tune of 38-7 over the next three weeks. Learning to cope with variance is an important lesson regardless of whether you bet on blackjack, or on sports, or never plan to set foot in a casino.

The second commandment of our statistical religion is the importance of a long-term perspective and the commitment to invest in it. Dr. Bob's bad run over 37 games destroyed both the confidence and bankroll of his non-believing clients. He had a terrible success rate of 14 percent in those 37 games, but his record in nearly 20 years and over 10,000 games was a solid 56 percent.

Long runs of bad luck in Vegas constantly tested our nerves and our belief in statistics. We had to maintain a long-term perspective and, more important, we had to have a big enough bankroll to ride out the inevitable downward swings. So the lesson in business, as in blackjack, is that you need to have a long-term perspective and the corresponding financial commitment to make analytics pay off.