Blackjack Bankroll Guide
How much you actually need to play blackjack without going broke — and the part most guides skip: the edge is tiny, the swings are enormous, and if you're not beating the game, no bankroll on earth will keep you out of the red forever. Real numbers, sourced math, no wishful thinking.
The five truths
If you only read this section, you'll have the 90% that actually matters.
The edge is tiny — and it's against you
Blackjack has one of the lowest house edges on the casino floor, which is exactly why it gets a reputation as "beatable." Play textbook basic strategy in a good six-deck game and the house edge is around 0.5% — the Wizard of Odds puts it near 0.41% under the most favorable common rules. That sounds like almost nothing. But "almost nothing" is not "nothing," and it's pointed the wrong way: it's the casino's edge, not yours.
Here's the part that surprises people. Even playing perfectly, you lose more hands than you win — roughly 42% of hands win, 49% lose, and the rest push. The reason the edge is still small is that blackjack pays 3:2 on a natural, and lets you double and split to put more money out when you're ahead. Strip out the 3:2 (that's what a 6:5 table does) or let the dealer hit soft 17, and the edge climbs toward and past 1%. Play on hunches instead of the chart and it balloons:
| How you play | House edge |
|---|---|
| 6 decks, dealer stands on soft 17, good rules | ~0.4–0.5% |
| 6 decks, dealer hits soft 17 | ~0.6% |
| 6:5 blackjack (avoid these tables) | ~1.4%+ |
| "Never bust" — never hit a hard 12 or more | 3.9% |
| "Mimic the dealer" — hit 16 or less, stand on 17+ | 5.5% |
| Card counting with a correct bet spread | player +0.5% to +1.5% |
So the very best edge a skilled card counter ever gets — under good rules, deep penetration, a real bet spread — tops out around 1 to 1.5% in their favor. To put that in perspective: that's a thinner advantage than reaching into a jar of 100 marbles where 52 are red and 48 are blue and betting you pull a red one. That's the whole game. A tiny, fragile edge, riding on top of a mountain of variance.
The variance is brutal
This is the number that humbles everyone. The standard deviation of a single blackjack round is about 1.15 betting units (Stanford Wong's Professional Blackjack and the Wizard of Odds both land right around there). Compare that to the edge — well under one unit per hundred hands — and you can see the problem: the noise is roughly a hundred times louder than the signal on any given hand.
Make it concrete. Bet $1 a hand and play 100 hands of basic strategy. Your expected result is about minus 50 cents. But the standard deviation of those 100 hands is about $11.40. So a perfectly ordinary night swings you eleven dollars up or down around a fifty-cent expectation — and two standard deviations, which happens all the time, is over twenty dollars either way. The expected loss is a rounding error next to the swing.
That gap between a slow edge and fast, violent swings is the entire reason bankroll matters. The bankroll isn't there to make you win — the edge does that, eventually. The bankroll is there to keep you in your seat long enough for "eventually" to arrive.
Two completely different situations
Before any dollar figure means anything, you have to know which player you are — because the entire logic of a bankroll changes depending on whether your edge is positive or negative.
How much a counter actually needs
If you've crossed into positive-edge territory, the question becomes: how many betting units do you need so the swings don't wipe you out before your edge shows up? That probability is your risk of ruin, and counters pick a level they can stomach. Here are figures straight from Blackjack Apprenticeship's own bankroll math, assuming perfect play, standard rules, and a 1-to-12 bet spread:
| Bankroll (betting units) | Risk of ruin | What it means |
|---|---|---|
| 200 units | ~40% | 4 of 10 counters lose the whole bankroll |
| 400 units | ~20% | 1 of 5 still go broke |
| What pros target | ≤1% | Requires substantially more than 400 units |
Read that top row again: even with a real edge and 200 units behind you, the most-cited practitioner source in the game says 4 out of 10 players in that spot go broke. To get risk of ruin down to the 1% a professional wants, you need a far deeper bankroll. In dollars, Blackjack Apprenticeship pegs a $1,000 bankroll at roughly a 45% risk of ruin and only about $10/hour of expected value — which is why serious counters generally run tens of thousands of dollars, not a few hundred, before the stakes are worth the hours.
The hard truth: no bankroll beats a negative edge
This is the most important paragraph on the page, so we'll say it plainly. If your expected value is negative — every player who isn't counting and spreading correctly — then no bankroll, of any size, will keep you from eventually going broke. This isn't an opinion; it's a centuries-old result mathematicians call gambler's ruin: when the game is tilted against you, the probability of losing your entire bankroll climbs toward 100% the longer you keep playing.
A bigger bankroll changes the journey, not the destination. It buys you more hands, more hours, more near-misses — but the house edge is a small, relentless tax on every bet, and over enough bets it collects with certainty. Doubling your bankroll roughly doubles how long you last; it does not change where you end up. That's why a basic-strategy player who wants the most table time for their money should bet small relative to their bankroll — not to win, which isn't on the menu, but to stretch the entertainment as far as it goes.
And it gets worse the further you drift from correct play. A "never bust" or "mimic the dealer" style hands the house a 4–10% edge instead of 0.5% — so ruin that would've taken a basic-strategy player a long evening arrives in a fraction of the hands. Playing wrong doesn't just lose faster; it makes any bankroll irrelevant.
Counting alone isn't enough — you have to spread
Here's the trap that catches new counters: they learn to keep a perfect count, sit down, bet the same chip every hand, and wonder why they're still losing. The count by itself doesn't make money. It only tells you when the remaining cards favor you. If you bet the same amount in good counts and bad counts, your average bet is still being made into a negative game — your edge stays underwater.
The edge appears only when you bet more when the count is high (you're favored) and less, or nothing, when it's low (the house is favored). Almost all of a counter's profit comes from those bigger bets in the high counts. A flat-betting counter is just a basic-strategy player who happens to know the number — still a long-run loser. The skill is two halves working together: keeping the count accurately, and turning that count into a bet that's correctly sized for your bankroll.
That's exactly what our trainers drill: the Card Counting Trainer for keeping and converting the count, and the Deviations Trainer for the count-based playing decisions. Pair them with the bet discipline below.
Bet sizing — the Kelly criterion
Once you have an edge, the question is how big to bet. Bet too small and you barely beat inflation; bet too big and one bad shoe can cripple you even though you're "winning." The Kelly criterion is the formula that finds the bet fraction that maximizes long-run bankroll growth for a given edge. The catch is that full Kelly is volatile — it carries roughly a 13.5% risk of ruin, and a real chance of cutting your bankroll in half before it doubles.
So almost no one bets full Kelly. The standard practice is a fraction of it:
| Betting level | Approx. risk of ruin | Trade-off |
|---|---|---|
| Full Kelly | ~13.5% | Fastest growth, gut-churning swings |
| Half Kelly | ~5% | Most of the growth, far smoother — the common choice |
| Quarter Kelly | ~1% | Conservative, what most pros actually use |
The practical version: your maximum bet should be a small slice of your bankroll, and you resize as your bankroll moves. Tools like CVCX (by QFIT) will compute the optimal spread, expected value, and risk of ruin for your exact game and bankroll — well worth it before you risk real money.
Protect the roll
Whether you're a counter or a recreational player, a few habits do more to keep you solvent than anything you'll do at the table:
Wall it off. Keep your blackjack bankroll physically and mentally separate from life money. The moment rent and gambling money share an account, you've lost the firewall that protects both. Resize down without ego. If a counter loses a third of their bankroll, their risk of ruin has jumped — the disciplined response is to drop your unit size, not chase it back at the old stakes. Track every session — date, game, hours, win or loss. Most people remember wins and forget losses; honest records fix that. And pick your game: good rules, deep penetration, and a 3:2 payout do more for your edge and your survival than any decision you'll make on a single hand. Walk past the 6:5 tables.
Sources & further reading
The numbers on this page are drawn from the sources serious players actually trust:
Wizard of Odds — Michael Shackleford's house-edge calculator, variance appendix, and risk-of-ruin tables are the standard reference for the underlying math. Blackjack Apprenticeship — Colin Jones's write-ups on the math behind advantage play and recommended bankroll, plus the CVCX tooling, are the practitioner's go-to. Stanford Wong, Professional Blackjack — the benchmark variance and covariance figures. Don Schlesinger, Blackjack Attack — the definitive treatment of risk of ruin, N0, and bankroll theory. And the Kelly criterion for the bet-sizing foundation.
Frequently asked questions
The questions players ask most about blackjack bankrolls.
How much bankroll do I need to count cards in blackjack?
Hundreds of betting units. Blackjack Apprenticeship's own figures, for perfect play with a 1-to-12 spread, put 200 units at roughly a 40% risk of ruin and 400 units at about 20%. Professionals target a risk of ruin of 1% or less, which needs substantially more. In dollars, a serious counter usually needs tens of thousands set aside — and it should be money you can afford to lose, not your savings.
Can a big enough bankroll make me a winner at blackjack?
No. If your expected value is negative — which it is for everyone not counting and spreading their bets correctly — a bigger bankroll only makes the loss take longer. This is gambler's ruin: against a negative edge, the chance of eventually losing your whole bankroll approaches 100% the more you play. A bankroll changes the journey, not the destination. The only thing that turns blackjack into a long-run win is a real positive edge.
What is the house edge with perfect basic strategy?
About 0.5% in a typical good six-deck game — roughly 0.41% under the best common rules per the Wizard of Odds, and around 0.6% or more if the dealer hits soft 17. A 6:5 game pushes it past 1.4%. Even playing perfectly you win only about 42% of hands and lose about 49%; the 3:2 payout, doubling, and splitting close the gap. Basic strategy makes the game close, but it stays in the house's favor.
Why is blackjack variance so high if the edge is small?
The standard deviation of a single round is about 1.15 betting units. Over 100 hands at $1 a hand, your expected result is about minus 50 cents, but the standard deviation is about $11.40 — so a normal swing dwarfs the expected result by twenty times or more. The edge is small and slow; the swings are large and fast. That gap is exactly why a big bankroll is necessary to survive long enough for an edge to show up.
Do I have to spread my bets, or is counting enough?
You have to spread. The count only tells you when the cards favor you. Bet the same amount every hand and you make the same bet in good counts and bad ones, so your average edge stays negative. The profit comes almost entirely from the bigger bets you place when the count is high. A flat-betting counter is just a basic-strategy player who knows the count — still a long-run loser.
What is risk of ruin?
It's the probability you lose your entire bankroll before your edge plays out. Counters target a level they can live with — pros usually want 1% or less, while many part-timers accept around 5%. A bigger bankroll relative to your bets lowers it, but for a player with a real edge it never hits zero, and for a player with a negative edge it climbs toward 100% over time.
Is counting cards actually worth it for most people?
For most people, no. The edge is thin — under 1.5% even in good conditions — and capturing it takes a large bankroll, long hours, real discipline, and a tolerance for downswings that can last weeks or months. If that's not you, the honest move is to play perfect basic strategy for fun with money you've accepted you may lose, and treat every session as a night out, not an investment.
Keep sharpening
- Card Counting Trainer → Build the count this whole guide depends on.
- Deviations Trainer → The count-based plays that grow your edge.
- Basic Strategy Trainer → Get the baseline automatic first — it's non-negotiable.
- Card Counting 101 → How the whole thing works, start to finish.
- Counting Systems Compared → Pick the system that fits how you'll play.
- Play the Long Game → Bankroll mindset, limits, and free helplines.
- Poker Bankroll Guide → The same brutal-variance math, for the felt.