Start with the session, not the hand
The question sounds like it wants a single number, but the useful answer is a decision sequence. You are not really betting a fraction of your total bankroll on one hand; you are deciding how much of a session budget to put at risk across a run of hands, and how large each individual wager should be relative to the swings that run will produce. Those are two different calculations, and confusing them is how players end up betting more per hand than their session can survive.
Blackjack is a negative-expectation game. The house edge sits on every wager, and no sizing pattern removes it. What bet sizing controls is how quickly you burn through a session budget when results run against you, and how wide the swings are while you play. A larger bet does not raise your chance of winning; it raises the speed at which the built-in edge acts on your money.
What a bankroll is and what it is not
In blackjack, a bankroll is the money you have set aside for play and are prepared to lose in full. It is not an investment pool, and it is not money needed for rent, food, or debt. The only bankroll that can be sized rationally is discretionary money whose loss changes nothing about your life. If the amount you are considering betting would hurt to lose, the answer to the question is that the bet is too large before any arithmetic begins.
Once the bankroll is genuinely discretionary, the next step is to split it into session amounts. A session is one continuous period of play. Dividing the bankroll into sessions does not change the house edge, but it limits the damage any single bad run can do and gives you a natural stopping point. How many sessions you divide it into is a personal choice, not a rule; the point is that the session amount, not the whole bankroll, is what a single table session can realistically consume.
Unit sizing: the smallest decision that matters
Most blackjack players think in units. A unit is your standard bet for one hand. It is typically expressed as a small fraction of the session budget, because blackjack outcomes are volatile enough that a run of losses is ordinary rather than unusual. The exact fraction is not something this article can hand you, because it depends on the rule set, the table minimum, and how many hands you expect to play. What you can do is work backwards from the table minimum and the session amount.
Suppose a session budget is 100 units. A single unit is then 1% of the session, and a run of twenty lost hands costs twenty units, leaving the session largely intact. Now suppose the same session is only ten units. The same twenty-hand losing run is impossible to survive, because the session is gone twice over. The arithmetic is hypothetical, but the relationship is real: the smaller the session is in units, the more likely a normal downswing is to end it.
The table minimum sets a floor. If the casino's minimum bet is higher than the unit size you calculated, you cannot play at that table on that budget without betting more per hand than your session plan allows. The correct move is to find a lower-minimum table or wait until the bankroll supports the minimum comfortably. Betting the minimum when it is already too large for your session is how a budget fails in the first twenty minutes.
Why bet size does not change the house edge
Blackjack has one of the lower house edges among casino table games when played with a published basic strategy that matches the specific rule set in use. But the edge applies to every unit wagered, whether that unit is small or large. Doubling your bet doubles the expected loss on that hand. It also doubles the variance, which is the short-term swing around that expectation.
This is where most sizing mistakes come from. A player loses a few hands, raises the bet to recover, and now the same edge is eating a larger amount. The losing run does not become less likely because the bet is bigger. The progression systems built on doubling after losses do not remove the edge; they redistribute outcomes so that most sessions show a small win and a few sessions show a catastrophic loss. The average result is still negative, and the variance is far more dangerous to a fixed bankroll.
What genuinely varies under your control
Several things about blackjack are under your control, and they interact with bet sizing more than most players realise.
- Rule selection. The house edge on blackjack changes with the rule set: number of decks, whether the dealer stands or hits on soft 17, doubling and splitting permissions, and surrender options. A table with worse rules has a higher edge, which means a given bet size loses more in expectation. Choosing a better rule set is a real, measurable way to reduce expected loss without touching bet size.
- Game speed. Blackjack hands per hour vary with the number of players and the dealer. A full table plays fewer hands per hour than a heads-up game. The more hands you play, the more times the edge is applied to your unit. A larger unit at a slow table can cost less per hour than a smaller unit at a fast one.
- Session length. The longer you sit, the more hands you play, and the more likely your actual result is to drift toward the expected loss. Short sessions are dominated by variance; long sessions are dominated by the edge. Bet sizing interacts with this: a player who bets large and plays long is maximising exposure to the edge.
- Stopping rules. Deciding in advance to stop at a loss limit or a time limit does not change the house edge, but it caps the damage a bad session can do. A stopping rule is not a winning strategy; it is a budget protection mechanism.
Comparing the common approaches
Players use several frameworks for sizing bets, and they have different failure modes. The table below compares them on what they do to your session risk.
| Approach | What it does | Main risk | What it does not do |
|---|---|---|---|
| Flat betting | Same unit every hand | Slow bleed if the session is long | Does not remove the house edge |
| Small positive progression | Raise bet after wins | Bigger bets on a run that can reverse | Does not make wins more likely |
| Martingale-style doubling | Double after losses | Table limits and bankroll exhaustion | Does not recover losses; redistributes them |
| Session-budget unit sizing | Unit is a small fraction of session | Discipline required to stick to it | Does not guarantee session survival |
Flat betting and session-budget unit sizing are the only approaches that keep the bet size connected to the money you brought. The progression systems are the ones that most often end a session early, because they increase the bet precisely when the bankroll is shrinking.
Setting a number you can follow
Start with the session amount you are prepared to lose. Divide it into enough units that a normal losing run does not end the session. A useful test is to imagine the worst run you have personally seen at a blackjack table and check whether your session in units survives it. If it does not, the unit is too large or the session is too small.
Then check the table minimum. If the minimum is more than your unit, the table is too expensive for this session. Walk away or wait. The casino will still be there when the bankroll matches the table.
Finally, write down the unit and the session loss limit before you sit down. The act of writing it does not change the mathematics, but it changes the psychology. A number decided in advance is easier to honour than a number improvised mid-session, especially after a losing run.
Where the missing numbers live
This article deliberately does not give you a universal percentage, because no honest one exists. The house edge on your specific table is printed nowhere on the felt, but it can be derived from the rule set: deck count, dealer stand rule, double and split permissions, and surrender. Published basic-strategy charts exist for common rule combinations, and they show the edge for each. The table minimum is posted at the table or in the game lobby. The session amount is yours alone.
If you want a number to start from, work it backwards: decide the session loss you can absorb, divide by the number of hands you expect to play, and compare that per-hand figure to the table minimum. The result is a unit size that fits your session, not a promise about the outcome.
FAQ
Is there a bankroll percentage that guarantees I will not go broke?
No. Blackjack is a negative-expectation game, and any positive percentage of a bankroll wagered repeatedly will eventually encounter a losing run that exhausts it. Sizing bets as a small fraction of a session budget reduces the chance of a single bad run ending the session, but it cannot remove the house edge or the possibility of loss.
Should I raise my bet after a win because I am playing with house money?
The idea of house money is a mental accounting trick. Once a win is credited to your balance, it is your money, and the next hand's edge applies to it exactly as it would to money you brought from home. Raising the bet after a win increases the amount exposed to the edge on subsequent hands and does not make those hands more likely to win.
How do I know if a blackjack table is too expensive for my bankroll?
Compare the table minimum to the unit size you calculated from your session budget. If the minimum bet is larger than the unit you can afford to lose repeatedly without ending the session, the table is too expensive. The remedy is a lower-minimum table or a larger session budget, not a bigger bet on the same table.
If gambling has stopped being a choice for you or someone close to you, free and confidential support is available. Speak with a trained advisor at any time; the service is independent of any operator, and calls are not recorded against your name. The helpline printed in the footer below is the fastest way to reach a person who can help.
Play responsibly
Gambling should be treated as paid entertainment, never as a way to earn income or recover losses. 1-800-MY-RESET (1-800-697-3738)
18+ or 21+ depending on where you are; follow the minimum age that applies to you.
This article is general information about how these mechanics work. It is not legal advice and not a recommendation to gamble or to use any particular operator. Availability and legality differ by jurisdiction — check the rules that apply where you are.