How Card Counting Works in Blackjack (and Why It's Harder Than You Think)

Card counting works. i need to lead with that because everyone wants to know. yes. the math is real. you can gain an edge. but the edge is thin. the variance is brutal. and the casino will notice.

i spent eight years doing this full-time. left Las Vegas with $140k and a backache. stayed in rental cars and casino hotels. the romance of it lasted about three weeks.

The Math is Simple

High cards (10 through Ace) are bad for the dealer. The dealer hits on 16 and stands on 17. Low cards help the dealer hit stiff hands without busting. A true count above 1 means the remaining deck is rich in 10s and Aces. You bet more. A true count below 1 means the remaining deck is rich in 2 through 6. You bet minimum.

Hi-Lo count is the standard. Cards 2 through 6 are plus one. Cards 7 through 9 are zero. Cards 10 through Ace are minus one. You track the running count. Divide by estimated remaining decks. That is your true count. Adjust bet size and strategy based on true count. Hit on soft 17 when the count is high. Stand on 16 when the count is low. The basic adjustments are in every card counting book.

The edge you can gain is about 1% at a true count of plus 2 or higher. That means betting $100 at a plus 2 true count gives you an expected value of $1. Over 100 hands that is $100 in edge. It sounds good. It is not good enough.

Variance at blackjack is enormous. You can have a true count of plus 3, bet $500, and lose five hands in a row. That is not a sign you are doing something wrong. That is just the nature of the game. I have been up $8,000 in a night and down $12,000 by morning. Both nights had solid true counts and correct bet sizing.

The Execution Problem

You need to run numbers at speed. Converting running count to true count in your head while maintaining table presence is hard. Most people overestimate how well they can do this under pressure.

You also need cover play. You cannot always bet $25 when the count is negative and $500 when the count is positive. The pit bosses watch for that exact pattern. So you make some -EV bets to blend in. You hit on stiff hands when you should stand. You take insurance even though it is bad math. You lose money intentionally to look normal.

I lost roughly $3,000 per month in cover play over eight years. That is $288,000 in deliberate losses to avoid getting backed off. The edge you win from counting is smaller than the vig you pay just to stay in the game.

Then there is the actual problem: the casino fights back. They use cut cards that eliminate the last 40 percent of the shoe. Shoes with six decks instead of eight. Multiple deck games only because counting is harder with more cards in play. Constant dealer changes to prevent observation. Some casinos now have eye-in-the-sky AI that flags bet-sizing patterns in real-time.

The Backoff

You will get backed off. That is not an if. It is a when. The casino may be polite about it. "We appreciate you playing with us but we would prefer if you took your action elsewhere." Or they may be rude. I was escorted out of a Red Rock property by security for betting $10 to $500 based on count. No warning.

Once you are backed off from a casino, you are listed in a database. Other casinos see it. Your window for play narrows. I eventually reached a point where I could only play in card rooms that did not surveil aggressively, which meant lower table limits and worse penetration.

Penetration is the percentage of the shoe dealt before reshuffling. When I started, good penetration was 75 percent. By the time I quit, casinos were reshuffling at 60 percent to kill the edge. The edge compressed as the industry adapted.

The Real Path

If you want to try it: start with Team play. Multiple people spotting different tables. The spotter tracks count. The big player sits out most hands and enters only at high counts. This requires bankroll and coordination. Most teams blow up after four months because one member gets emotional.

If you are solo: accept that you will lose money covering. Accept that you will get backed off. Plan for 150 hours of play before you see statistical edge. Budget $15k in losses during the variance phase. That is realistic.

The job became harder because the industry got smarter. The edge compressed. The variance stayed the same. After eight years I could not justify the grind. The days of casual counters making five figures a year are over. if you are doing this now, you are a hobbyist with a mathematical edge. Not a professional. The math still works. The life does not.