Blackjack variance is the statistical dispersion of outcomes from expected value. A player with a 0.5% advantage will win, over time. Over a session, they will often lose.
The work of Don Schlesinger (1994) showed that a blackjack player with a $10,000 bankroll and a 1% advantage faces a 5.5% risk of going broke within 300 hands. This is not a theoretical concern; this is actual data from counting gameplay.
The Dispersion
At a single-deck game with basic strategy, the standard deviation is approximately 1.1 units per hand. With an expectation of 0.01 units per hand, your confidence interval expands rapidly. After 100 hands, you expect to be ahead by 1 unit, but your standard deviation is 11 units. A 2-standard-deviation downswing leaves you at -21 units.
Don't confuse this with randomness. Randomness is the presence of variation. Variance is the amount of variation. Blackjack has high variance because each hand outcome is binary (win or lose) and the edge is small (0.5% to 1% depending on counting). A high-variance game with a small edge requires substantial bankroll to absorb downswings.
What This Means for Your Bankroll
Thorpe's original work on card counting suggested that a 1% advantage with 1.1 standard deviation per hand requires approximately 80-100 hand sizes of bankroll to weather a 2-standard-deviation downswing with acceptable risk (5% or lower).
If you are betting $50 per hand, you need $4,000 to $5,000 in bankroll. If you are betting $500 per hand, you need $40,000 to $50,000. Most card counters undercapitalize and bust out before their advantage has time to work.
The Normalization of Variance
Speaking from behavioral economics: people do not intuitively grasp variance. A player who loses $2,000 in a winning game reports feeling bad, period. The fact that the loss is a reasonable outcome given the variance does not soothe the feeling.
This is where mental accounting (Richard Thaler's concept) matters. If you frame the loss as "a possible outcome of a profitable game," you experience it differently than if you frame it as "I made a mistake and lost money."
Casinos benefit from the second framing. Players self-blame, tighten up, and play sub-optimally to recover losses. The casino's edge expands because variance in the player's mind becomes evidence of skill deficit.
Long-Term Perspective
Variance dampens over time. After 10,000 hands, a 1% advantage with 1.1 standard deviation produces a 99.7% confidence that you are ahead (within 3 standard deviations). The downswing you experience in 100 hands is noise in the 10,000-hand data.
The problem is that most blackjack players do not play 10,000 hands. They play 100, assess the results, adjust their strategy (incorrectly), and move on. Variance is the mechanism by which unlucky skilled players lose and lucky unskilled players win.
Key Takeaway
The key is understanding the system. Most people do not take the time. The ones who do are the ones who profit from the ones who do not.
Understanding the System
The lesson underneath all of this is the same: understand the system before you enter it. The system is designed. The odds are known. Your job is to know them too, and then decide whether the trade is worth making. Most people skip the first step and proceed directly to wondering why they lost money.
