Blackjack basic strategy explained: what it does and does not do
Quick answer
Basic strategy is the mathematically best play for each hand against the dealer's up card; on a standard eight-deck table it cuts the house edge from about 5.5% for a player copying the dealer to about 0.43%, but it never removes it.
- Edge with basic strategy
- About 0.43% (Wizard of Odds)
- Edge copying the dealer
- About 5.48% (Wizard of Odds)
- Does it remove the edge?
- No
- Source
- Gambling Commission
- Checked
What is basic strategy?
Basic strategy is the mathematically best play for each blackjack hand, given your cards and the dealer’s up card. It lowers the house edge but never removes it. For every combination, it picks the option, whether hit, stand, double or split, with the smallest expected loss or the largest expected gain over the long run.
It is not a system, a betting plan or a way to predict cards. It is simply the answer to a sum. Each decision has a long-run value that can be worked out exactly from the probabilities of the cards still to come, and basic strategy is the list of the best values. Our guides to how casino games work explain the house edge behind every game on this site, and blackjack is the one where those decisions change it most.
The rules and method behind the chart. The chart is for eight decks, dealer stands on soft 17, blackjack pays 3 to 2, double on any first two cards, double after splitting allowed, split up to four hands, no surrender, and the dealer checks for blackjack at the start. We worked it out ourselves with an exact calculation over every hand, using an infinite-deck model in which each card rank is equally likely on every draw. We then checked every decision against the expected-value tables the Wizard of Odds publishes for eight decks under those rules (checked 19 November 2026). Two decisions differ, both near-ties where the eight-deck tables favour doubling: ace and 2 against a 5, and ace and 4 against a 4. The chart below uses the eight-deck answer for both. Different rules shift a handful of decisions, and the notes under each table say where.
If you are new to the game, read how to play blackjack first. It covers card values, the order of play and what each option means.
Hard totals
A hard hand has no ace, or has an ace that must count as 1.
| Your hand | Against dealer 2 to 6 | Against dealer 7 to ace |
|---|---|---|
| 8 or less | Hit | Hit |
| 9 | Double against 3 to 6, hit against 2 | Hit |
| 10 | Double | Double against 7 to 9, hit against 10 and ace |
| 11 | Double | Double against 7 to 10, hit against ace |
| 12 | Stand against 4 to 6, hit against 2 and 3 | Hit |
| 13 to 16 | Stand | Hit |
| 17 or more | Stand | Stand |
Soft totals
A soft hand has an ace that still counts as 11.
| Your hand | What to do |
|---|---|
| Ace and 2 | Double against 5 and 6, otherwise hit |
| Ace and 3 | Double against 5 and 6, otherwise hit |
| Ace and 4 | Double against 4 to 6, otherwise hit |
| Ace and 5 | Double against 4 to 6, otherwise hit |
| Ace and 6 (soft 17) | Double against 3 to 6, otherwise hit |
| Ace and 7 (soft 18) | Stand against 2, 7 and 8; double against 3 to 6; hit against 9, 10 and ace |
| Ace and 8 or ace and 9 | Stand |
Where the chart says double but you already have three or more cards, hit instead, except on soft 18, where you stand.
Pairs
| Your pair | What to do |
|---|---|
| Aces | Always split |
| 8s | Always split |
| 10s | Never split; stand |
| 9s | Split against 2 to 6, 8 and 9; stand against 7, 10 and ace |
| 7s | Split against 2 to 7, otherwise hit |
| 6s | Split against 2 to 6, otherwise hit |
| 5s | Never split; play as a hard 10 |
| 4s | Split against 5 and 6, otherwise hit |
| 2s and 3s | Split against 2 to 7, otherwise hit |
Where the rules change the chart. On a no-hole-card table, where the dealer’s second card is dealt only after you have played, three decisions change. Hitting a hard 11 against a 10 instead of doubling is confirmed by the Wizard of Odds’ no-hole-card tables. Our own calculation adds two pair decisions for which we found no published table: hit a pair of 8s against a 10 or an ace, and hit a pair of aces against an ace. On those tables a dealer blackjack also takes any extra money you put in, so adding to a bet against a 10 or an ace costs more.
Using the chart
Read it in a fixed order each hand. If your first two cards are a pair, start with the pairs table. If not, and you hold an ace that can count as 11, use the soft table. Otherwise use the hard table. Find your hand in the left-hand column, then read across to the dealer’s up card.
After you split, play each new hand on its own with the same tables, as if it were a fresh two-card hand. After you hit, look your new total up again: a soft hand that takes a big card can become hard, and the right play changes with it. If the table offers surrender, which our base rules leave out, some of the weakest hands against a 10 or an ace are better given up for half the stake; the chart above does not cover that option.
Many players learn the hard totals first, because they come up most often, then the pairs and soft hands. Whatever order you learn it in, the chart only lowers the cost of playing.
A worked hand
You hold 6 and 5, a hard 11, and the dealer shows a 6. The chart says double. Here is why, in money terms, for each £1 of your original bet, from the Wizard of Odds’ eight-deck expected-value table:
| Option | Expected result per £1 |
|---|---|
| Stand on 11 | Lose about 15p |
| Hit, then play on as the chart says | Win about 34p |
| Double (stake £2, take one card) | Win about 68p |
Standing is clearly wrong, because 11 cannot bust and beats nothing on its own. Hitting is good. Doubling is better, because you put more money in at the moment the maths most favours you: the dealer, showing a 6, busts about 42% of the time, and your next card is often worth 10.
Most hands are nowhere near as clear. The closest call in the game is a hard 16 against a dealer’s 10. With 10 and 6, the published eight-deck figures show standing losing about 54.1p per £1 and hitting about 53.6p; hitting is better by half a penny. A hard 12 against a 4 is almost as tight. Either way the hand is a losing one, and the chart only picks the smaller loss.
Some hands look close but are not. With 12 against a dealer’s 7, standing loses about 48p per £1, while hitting loses about 21p. The dealer’s 7 busts far less often than a 4, 5 or 6, so waiting for the dealer to bust gives too much away.
Why the chart says what it says
The chart can look arbitrary until you see the numbers behind a few of its firmest rules. All figures are per £1 of your original bet, from the Wizard of Odds’ eight-deck expected-value tables for the rules above:
| Hand | Options compared | Expected result per £1 | Chart |
|---|---|---|---|
| Pair of 10s against a 6 | Stand on 20 / split | Win about 70p / win about 45p | Stand |
| Pair of 8s against a 10 | Stand / hit / split | Lose about 54p / lose about 54p / lose about 48p | Split |
| Pair of aces against a 7 | Hit soft 12 / split | Win about 16p / win about 47p | Split |
| Ace and 6 (soft 17) against a 3 | Stand / hit / double | Lose about 11p / win about 3p / win about 6p | Double |
| Pair of 9s against a 7 | Stand on 18 / split | Win about 40p / win about 36p | Stand |
The patterns are consistent. A 20 is already a strong hand, so splitting 10s swaps a likely win for two uncertain ones. Hard 16 is the worst total you can hold, so splitting 8s turns one bad hand into two that each start on 8, and loses less even when the dealer shows a 10. Two aces start two hands on 11. Soft 17 can never bust on one card, which makes adding to the bet against a weak dealer worth it. And a pair of 9s against a 7 is better left as 18, because the dealer’s most likely finish from a 7 is a total that 18 already beats.
The bet the chart always declines
When the dealer shows an ace, you will be offered insurance, and if you hold a blackjack you may be offered “even money” instead. Both are the same side bet on the dealer having a 10 underneath. In an infinite-deck model, our calculation, the hidden card is a 10 four times in thirteen, so insurance at 2 to 1 loses about 7.7p for every £1 staked on it. Taking even money pays you exactly £1 per £1 for certain, while declining it is worth about £1.04 per £1 on average. Basic strategy never takes either.
How the chart is worked out
Every entry comes from the same calculation. For each dealer up card, we first work out how the dealer’s hand finishes: the chance of each total from 17 to 21 and the chance of busting. For each player hand, the value of standing is then fixed by those chances. The value of hitting is the average, over every card that could come next, of the best value of the hand that results. Doubling and splitting are valued the same way with the stake changed. The option with the highest value goes in the chart. No guesswork or judgement enters at any point, which is why the chart is the same wherever it is printed for the same rules.
These are expected values: averages over a very large number of identical hands. Any single hand is won, lost or pushed, and the chart cannot tell you which.
How much does it reduce the house edge?
A lot, compared with playing by instinct. The Wizard of Odds publishes the house edge for three ways of playing under its standard rules: dealer stands on soft 17, double on any two cards, double after splitting and split up to four hands:
| How you play | House edge (Wizard of Odds) | Expected loss on £1,000 of bets |
|---|---|---|
| Basic strategy | 0.43% | About £4.30 |
| Never risk busting: never hit a hard 12 or more, otherwise basic strategy | 3.91% | About £39.10 |
| Copy the dealer: hit 16 or less, stand on 17 or more, never double or split | 5.48% | About £54.80 |
Basic strategy saves about £5 for every £100 staked compared with copying the dealer. The two homemade approaches fail for opposite reasons. Copying the dealer means you bust first and lose before the dealer plays. Refusing to bust means you stand on weak totals like 12 to 16 against a dealer’s 7 or higher, which are the hands where standing gives away the most.
Our own infinite-deck calculation, with only one split allowed, gives 0.57% for basic strategy and 5.67% for copying the dealer. The basic-strategy gap is explained by the extra splits and the shoe size; for copying the dealer the rules differ slightly, so we quote the published figure.
The figure moves with the table rules. Using the Wizard of Odds’ rule-variation figures, the edge with best play rises by 0.22 points, to about 0.65%, when the dealer hits soft 17, and by 1.39 points, to about 1.82%, when a blackjack pays 6 to 5 instead of 3 to 2. A strategy chart cannot make up for a worse payout, so the rules printed on the table matter before the first card is dealt. The full comparison is in our guide to how to play blackjack.
These figures are for eight decks. Tables with fewer decks shift them a little, and the rules on each table matter more than the shoe.
What strategy cannot do
Basic strategy cannot remove the house edge, and nothing in it turns blackjack into a way to make money. On a standard table, a player who makes every decision perfectly still loses about 43p for every £100 staked, on average, over the long run.
It also cannot:
- Control the short run. Results over an evening swing widely either way. A winning session does not mean the strategy is working, and a losing one does not mean it has failed.
- Make losses due to turn round. Each hand starts fresh. A run of losing hands does not make a winning one more likely, which is the trap explained in our guide to the gamblers fallacy.
- Be fixed with bet sizing. Raising your stake after a loss does not change the edge on each hand. It only increases how much is at risk. The same is true of every staking system at roulette, as our guide to roulette odds and the house edge shows with the numbers.
- Fix a bad table. If blackjack pays 6 to 5, or the dealer hits soft 17, the edge is higher whatever you do.
Online, the Gambling Commission requires random-number-driven games to be tested by an approved test house before release, and operators must monitor that each game performs as designed. So the maths above describes a fair game; it does not describe a beatable one.
Learning the chart can make blackjack cheaper to play. It cannot make it profitable, and the only sensible budget is money you can afford to lose. The NHS lists chasing losses, and needing to gamble with larger amounts to get the same feeling, among the signs that gambling is causing harm. If that sounds familiar, the National Gambling Helpline is free on 0808 8020 133, and the safer gambling page lists more support. You must be 18 or over to gamble. Our editorial policy explains how we produce and check calculations like these.
Sources
- How we ensure online gambling games are fair, Gambling Commission. Accessed 19 November 2026.
- Blackjack basics: house edge and bad strategies, Wizard of Odds. Accessed 19 November 2026.
- Blackjack rule variations, Wizard of Odds. Accessed 19 November 2026.
- Blackjack expected values: 8 decks, dealer stands on soft 17, Wizard of Odds. Accessed 19 November 2026.
- Blackjack expected values: 8 decks, stands on soft 17, no hole card, Wizard of Odds. Accessed 19 November 2026.
- Gambling addiction, NHS. Accessed 19 November 2026.