Blackjack Basic Strategy Trainer

Set your table's rules, see the chart they produce and the cells that move, then drill it hand after hand.

This prices the play, not the game. Basic strategy is the cheapest way to lose at blackjack, and on the default rules it still costs about 0.41% of every bet you place, on average, indefinitely.

Your table’s rules

Six decks and no surrender are fixed for now. Both change several plays rather than a handful, and neither is validated in our model yet, so the chart below does not pretend to cover them.

House edge under these rules

0.41%

Six decks, dealer stands on soft 17, double after split allowed, no surrender.

Dealer

Hard 16

The chart says

Hit

Outlined on the chart below.

What each table’s rules cost

All four combinations of the two rules above, ranked by house edge under perfect play. Your table is marked.

The dealer hitting soft 17 costs about 1.5x as much as losing double after splitting. Between two tables that differ on one rule each, that is the one to shop away from.

Basic strategy chart

Columns are the dealer’s upcard. Hard 17 and up always stands, hard 8 and below always hits, so neither has a row.

Hit
Stand
Double
Split

Hard totals no ace, or an ace counted as 1
Your hand	2	3	4	5	6	7	8	9	10	A
8	Hit	Hit	Hit	Hit	Hit	Hit	Hit	Hit	Hit	Hit
9	Hit	Double	Double	Double	Double	Hit	Hit	Hit	Hit	Hit
10	Double	Double	Double	Double	Double	Double	Double	Double	Hit	Hit
11	Double	Double	Double	Double	Double	Double	Double	Double	Double	Hit
12	Hit	Hit	Stand	Stand	Stand	Hit	Hit	Hit	Hit	Hit
13	Stand	Stand	Stand	Stand	Stand	Hit	Hit	Hit	Hit	Hit
14	Stand	Stand	Stand	Stand	Stand	Hit	Hit	Hit	Hit	Hit
15	Stand	Stand	Stand	Stand	Stand	Hit	Hit	Hit	Hit	Hit
16	Stand	Stand	Stand	Stand	Stand	Hit	Hit	Hit	Hit	Hit

Soft hands an ace counted as 11
Your hand	2	3	4	5	6	7	8	9	10	A
A,2	Hit	Hit	Hit	Double	Double	Hit	Hit	Hit	Hit	Hit
A,3	Hit	Hit	Hit	Double	Double	Hit	Hit	Hit	Hit	Hit
A,4	Hit	Hit	Double	Double	Double	Hit	Hit	Hit	Hit	Hit
A,5	Hit	Hit	Double	Double	Double	Hit	Hit	Hit	Hit	Hit
A,6	Hit	Double	Double	Double	Double	Hit	Hit	Hit	Hit	Hit
A,7	Stand	Double	Double	Double	Double	Stand	Stand	Hit	Hit	Hit
A,8	Stand	Stand	Stand	Stand	Stand	Stand	Stand	Stand	Stand	Stand
A,9	Stand	Stand	Stand	Stand	Stand	Stand	Stand	Stand	Stand	Stand

Pairs split or play the total
Your hand	2	3	4	5	6	7	8	9	10	A
2,2	Split	Split	Split	Split	Split	Split	Hit	Hit	Hit	Hit
3,3	Split	Split	Split	Split	Split	Split	Hit	Hit	Hit	Hit
4,4	Hit	Hit	Hit	Split	Split	Hit	Hit	Hit	Hit	Hit
5,5	Double	Double	Double	Double	Double	Double	Double	Double	Hit	Hit
6,6	Split	Split	Split	Split	Split	Hit	Hit	Hit	Hit	Hit
7,7	Split	Split	Split	Split	Split	Split	Hit	Hit	Hit	Hit
8,8	Split	Split	Split	Split	Split	Split	Split	Split	Split	Split
9,9	Split	Split	Split	Split	Split	Stand	Split	Split	Stand	Stand
10,10	Stand	Stand	Stand	Stand	Stand	Stand	Stand	Stand	Stand	Stand
A,A	Split	Split	Split	Split	Split	Split	Split	Split	Split	Split

This is the default chart, the six-deck game this site quotes everywhere else. Change either rule above and the cells that move are marked and listed here.

Strategy trainer, 20 hands

Best 0

Dealer

Your hand

You get a hand and the dealer’s upcard. Pick the move that matches the chart above, faster answers score more, and a clean streak multiplies. The drill grades against the rules you selected.

Playing this chart perfectly still loses about 0.41% of everything you stake, on average, and no rule set on this page turns that positive. Basic strategy makes the game cheaper, not profitable. Deviating on a feeling is what pushes most players back toward 2% or worse.

Written and reviewed by Alexander Korsager · last reviewed October 7, 2026 · corrections

The idea

Basic strategy is the mathematically optimal way to play every blackjack hand, the single decision that loses the least over millions of rounds. Followed on every hand it cuts the house edge on the default six-deck game set out below to about 0.41%, without any card counting. It does not remove the edge: no rule set on this page turns blackjack into a game you are expected to win.

The chart looks like a lot to memorize, but it is really a handful of patterns: always split aces and eights, never split tens or fives, stand on hard 17 and up, hit hard 8 and below, and double your strong 9, 10 and 11 against a weak dealer card. The full chart is on this page, and the trainer under it deals random hands and grades each choice against it so the patterns stick.
Two table rules are yours to set above the chart, because they genuinely vary from table to table: whether the dealer hits or stands on soft 17, and whether you may double after splitting. The chart re-derives for whichever combination you pick, and the cells that move are marked and listed, since a player who already knows basic strategy needs the difference rather than a second chart to memorize. The default is the common six-deck game: dealer stands on soft 17, double allowed on any two cards, double after split allowed, no surrender.
Charts are usually published for four to eight decks as a single block, and ours is very nearly that: every cell is identical at four, five, six and seven decks. At eight decks exactly one changes, and only just. An ace with a two against a dealer 5 becomes a hit rather than a double, by half a thousandth of a bet. Deck count does move the edge itself, which is why the figure above names a deck number: the same chart costs about 0.36% at four decks, 0.41% at six and 0.43% at eight.
Deck count and surrender are not adjustable yet, and that is deliberate rather than unfinished.The two rules above are checked against published figures for what each is worth, cell by cell and in the edge. The other two are not, and a control that generated advice nobody had validated would be worse than no control on a page a player acts on with money.

FAQ

Questions players ask

Does basic strategy guarantee I win?

No. It minimizes the house edge to about 0.41% on the six-deck game described above, but the casino still has an edge. Basic strategy makes you lose slower, not win, only card counting or advantage play can flip the edge.

What is the single most important rule?

Always split aces and eights, and never stand on a hand you should hit. The most expensive mistakes are standing on 16 or less against a dealer 7 or higher, and failing to double 11.

Why does the trainer sometimes want me to double a 9?

Hard 9 doubles against a dealer 3 through 6, the cards where the dealer is most likely to bust. You put more money in when the dealer is weak and you have a strong drawing hand.

How does the chart change if the dealer hits soft 17?

Three cells move, and nothing else does. Hard 11 against an ace becomes a double rather than a hit, A,7 against a 2 becomes a double rather than a stand, and A,8 against a 6 becomes a double rather than a stand. What that rule alone costs you, against the other three rule combinations this page offers, is in the rule-cost ladder above the chart. Set the dealer soft 17 control on this page to hits and the three cells are marked in the chart.

What if I cannot double after splitting?

Seven cells move, all of them pairs that stop being worth splitting: 2s against a 2 or 3, 3s against a 2 or 3, 4s against a 5 or 6, and 6s against a 2 all become a hit. Splitting is worth less when you cannot put extra money in on the good hands that follow. The rule-cost ladder above the chart shows what that costs against the other three rule combinations, and how it compares to the dealer soft 17 rule above.

Does this match what real online blackjack software returns?

Roughly, for the closest published comparisons, and not for all of them. Pragmatic Play publishes 99.6% RTP for American Blackjack and 99.62% for Multihand Blackjack (both reconfirmed 2026-09-29, unchanged since first read), within three hundredths of a point of the 99.59% this page computes for the default six-deck, dealer-stands-on-soft-17 game. Pragmatic Play does not state the exact rule set behind either figure, deck count, soft 17, doubling and splitting rules included, so a gap does not mean either number is wrong. It means that build is running different rules than the ones this chart assumes, and a game's own information screen, not this page, states the rules for the table in front of you. A second studio, Play'n GO, published three more blackjack titles here until it withdrew the pages those figures were read from; actfap.com/rtp no longer carries a Play'n GO comparison for that reason. The full sourced figures are at actfap.com/rtp.

Methodology

How this is calculated

Every cell on this chart is solved rather than transcribed.

For each row the engine computes the expected value of hitting, standing, doubling and splitting, averages each action over every two-card hand that makes the row's total, weights those hands by how often a six-deck shoe deals them, and prints whichever action pays most.

The shoe is finite: your two cards and the dealer's upcard come out before anything is evaluated, and every draw inside the recursion removes its own card, so the probabilities are exact without-replacement figures rather than the infinite-deck shortcut most published charts are derived under.

The house edge quoted with the chart is the cost of following this chart on every card, not the cost of perfect play, and it is expected loss per unit of the opening bet, averaged over every opening deal at the probability a full shoe produces it.

Common misreadings4

The edge is an average over a very large number of hands, not a budget. It says nothing about tonight, and a player following this chart exactly can lose for a long time without anything having gone wrong.
Playing this chart at a table with different rules is not playing this chart. Set the two controls to match the table before you commit any of it to memory, and read the blackjack payout off the felt while you are there.
A correct decision that loses was still the correct decision. Each cell is the action that loses least across every hand of that type, which is a claim about thousands of hands and not about the one in front of you.
Nothing here counts cards. Following the chart perfectly reduces what the game costs you and does not turn it into a game you are expected to win.

Assumptions5

The rule set is the one selected above, plus three the chart's header does not print: the dealer checks for blackjack before you act, splitting can leave you with at most four hands, and split aces get one card each and cannot be split again.
Blackjack pays 3 to 2. That is the assumption most worth checking against the table in front of you, because it is worth more than every decision on this chart put together. At 6 to 5 the chart does not change by a single cell, since the payout lands on a hand that is over before you read one, but the edge goes from about 0.41% to about 1.77%.
The chart is total-dependent, which is the convention every printed chart uses. Hard 16 gets one answer rather than separate answers for 10+6 and 9+7.
Every hand is dealt from a full shoe. The chart does not move as cards come out of it, which is the whole difference between basic strategy and counting.
The edge assumes you play the chart on every card, not just the first two. A cell that says double cannot be one on a hand of three cards or more, and the convention used there is the standard one: hit, except on soft 18, which stands.

Limits5

Deck count and surrender are not adjustable. The chart shown is correct at four, five, six and seven decks. At eight exactly one cell moves, an ace with a two against a dealer 5, and by less than a thousandth of a bet.
The dealer's final-total distribution is computed once per opening hand and then held fixed while you draw, so the cards you take are not removed from the dealer's shoe. This is the standard simplification. It is worth roughly 0.0001 on an expected value and cannot flip a cell that is not already a coin flip.
Deep resplits are priced with a slight over-credit. Each hand created by a split tracks its own remaining splits rather than sharing one budget across the whole tree, so a path needing three or more further cards of the same rank can be counted as available for one more split than the four-hand cap actually allows. It is worth well under a thousandth of a bet.
Insurance, even money and surrender are not modeled and have no row on the chart.
The 20-hand drill is a curriculum, not a table. It deals a pair on 28% of hands where a six-deck shoe deals one on about 15% of the hands you actually decide, a soft hand on 24% against about 10%, and it shows each dealer upcard one time in ten where a real table shows a ten-value card about three times in ten. It also never deals the totals with only one answer. Your drill score measures how well you know the chart, not how you would do at a table.

Evidence

The dealer model is validated first, against figures computed by someone else: it reproduces all sixty published cells of the dealer's outcome distribution by upcard (Wizard of Odds, Dealer Odds in Blackjack under U.S. Rules) to their printed precision.

The edge is then bracketed by two sourced numbers, a published 0.334% for six decks with late surrender and a published 0.08% for what surrender itself is worth to a player, which puts this rule set near 0.414% against the 0.41% computed here.

The two rules you can change are checked against their own separately published costs, 0.22 and 0.14 percentage points. Every numeric claim on this page about deck count, about the 6 to 5 payout and about what the drill deals is asserted in engine.test.ts and deal.test.ts rather than written down as prose.

A passing test means the code matches the model and this page matches the code, not that the table you sit at follows these rules.

Gambling can cause serious harm. The National Problem Gambling Helpline is free, confidential and open 24/7, call or text 1-800-MY-RESET (1-800-697-3738), or chat at ncpgambling.org/chat.

Read up

Go deeper

Related tools