The MIT Blackjack Team operated from approximately 1979 to 1995. They were not the only card-counting group operating in Las Vegas. They were the most successful because they understood economics better than other groups.
Card counting is straightforward. A low card (2-6) is bad for the player. A high card (7-Ace) is good for the player. You keep a count. When the count is positive (more high cards remain), you bet more. When the count is negative (more low cards remain), you bet less. The edge comes from having information that the casino does not.
But here is the economic problem: you cannot bet a billion dollars on a single hand. Card counting generates a small edge (about 1-1.5% of your wager). To make money, you need volume or large stake size. Volume requires many hands. Large stake requires large capital. Both require management.
The MIT team solved this by recruiting students with mathematical talent and pooling capital. A single counter betting $10 per hand on a 1% edge makes $10 per hand expected value. The variance is enormous. The counter might be down $5,000 after a week and up $5,000 after the next week. The counter needs a bankroll to survive the variance.
The team pooled bankrolls. They deployed multiple counters. They bet larger amounts. They survived the variance. They made money.
The Advantage and The Cost
The expected value of counting cards is positive, but small. The team's edge was approximately 1.5% per hand. A typical hand is $50 to $100 bet. Expected value per hand is $0.75 to $1.50. Playing 200 hands per day (4 hands per minute for 50 minutes), you expect $150 to $300 per day.
But variance is large. A losing streak of 50 hands ($2,500-$5,000 loss) happens regularly. The team needed bankroll to absorb this.
The solution was to pool capital. If you have 20 counters, you can bet $50 per hand at each table, generating $1.50 to $3.00 per hand expected value, $300 to $600 per day per table. With variance pooled across multiple tables and players, the team could sustain losses that would bankrupt a single counter.
This is capital allocation theory. The team had an edge in blackjack. The team did not have an infinite edge. The edge required capital to operate and time to manifest. The team allocated capital efficiently to maximize the risk-adjusted return.
Why The Team Failed
The team eventually stopped because casinos learned to identify counters. Counters have tells: they bet large when the count is high and small when the count is low. They play according to basic strategy. They do not deviate. A surveillance system watching bet patterns can identify a counter in an hour.
Casinos responded by banning counters. If the team deployed a counter, the counter was photographed, identified, and added to a database. After 1991, many casinos refused to deal to known counters. The team's advantage evaporated.
Some members of the team attempted to continue by using disguises, false identities, and traveling to smaller casinos. The economics no longer worked. The edge was real, but it was not realizable because the infrastructure to exploit it was not available.
The Lessons
The MIT team's success was not about blackjack. It was about capital allocation, risk management, and operational execution. These lessons apply to any advantage-seeking activity.
First: an edge that is small requires either large capital or large volume to be profitable.
Second: variance is a constraint. A strategy with a positive expected value can still bankrupt you if variance exceeds your bankroll.
Third: operational discipline matters. Counters who deviated from the strategy made worse decisions and lost the edge.
Fourth: the competitive environment determines whether an edge is exploitable. An edge that was real in 1980 was not real in 1990 because the industry adapted.
By the late 1990s, blackjack advantage play was a historical case study, not a viable strategy. The MIT team's success was the peak of card counting profitability. After them, the infrastructure adapted. The edge closed.
