How to Win at Blackjack: 13 Tips and a Strategy Table
A 16 against a ten has a computed answer, and guessing it wrong is money you hand back on every shoe. Playing the solved table takes the house edge to about 0.5%. How to win at blackjack means shrinking that number before you are dealt a card, then not giving it back over the next 200 hands.
The rules screen sets your edge before the first card
Basic strategy puts blackjack at about 0.5%, against 2.70% on a single-zero wheel and 5.26% on a double-zero one, which is why the best roulette variants splits on the wheel and the rule before anything else, and against the 4% to 6% implied by a 94% to 96% RTP band on most slots. Blackjack is the cheapest seat in the building, but only if you play it properly.
Studios publish an RTP figure for slots and for a handful of live titles, which is how the highest RTP casino games get ranked at all. Blackjack mostly does not carry one. Evolution publishes no RTP for any of its blackjack titles, and its own Power Blackjack rules page is typical: eight decks, a full rule list, no percentage anywhere.
The single live blackjack figure a studio does publish is Pragmatic Play's Prive Lounge Blackjack, which the studio's own game page describes as increasing the RTP "up to 99.51%". That is a ceiling attached to particular rule options, not a rate the table returns by default.
So the honest comparison runs on rules, which is why the rule sheet is the only thing you can compare across live dealer tables. Four lines do almost all of the work, and the cost of each is a computed figure over a published model rather than a studio claim.
| Rule you can see before you sit | Effect on the house edge |
|---|---|
| Basic strategy, otherwise standard rules | about 0.5% in total |
| Blackjack pays 6:5 instead of 3:2 | about +1.36 points |
| Eight decks instead of one | +0.61 points |
| Dealer hits soft 17 instead of standing | +0.21 points |
| No surrender offered | +0.08 points |
The three lower figures come from a published analysis of blackjack rule effects.
1. Read four lines before you buy in
Open the rules panel and find four things: deck count, whether the dealer hits or stands on soft 17, whether blackjack pays 3:2 or 6:5, and whether you can double after a split. Those four lines separate the best blackjack games from the table that happens to load first. Everything else on that panel is decoration.
2. Walk away from any table paying 6:5
This is the most expensive rule a player can walk into, and the arithmetic is short. Computed over a six or eight deck shoe, a natural arrives about 4.75 times per 100 hands. At 3:2 a natural returns 1.5 units on a 1-unit bet; at 6:5 it returns 1.2. That is 0.3 units gone every time you turn one over, so 4.75 x 0.3 = 1.43 units per 100 units staked.
Strip out the naturals that push against a dealer natural and the figure settles nearer 1.36. Either way, one line on a rule sheet turns a 0.5% game into roughly a 1.9% game. Nothing else you do at that table moves the number nearly as far.
3. Take the table where the dealer stands on soft 17
A dealer hitting soft 17 instead of standing adds 0.21 points to the edge, and picking the other table costs you nothing. If two tables are otherwise identical, the one where the dealer stands is strictly better, and the rules panel says which is which before you buy in.
Play the solved table, all of it
Basic strategy is the computed best move for every combination of your hand and the dealer's upcard, and it has existed in published form since the 1950s. Playing it exactly is what produces the 0.5% figure.
The tables below cover a multi-deck game where the dealer stands on soft 17 and doubling after a split is allowed. Read your hand down the left, find the dealer's upcard, take the move.
4. Play every hand off the table, including the awkward ones
| Your hand | Dealer shows | Your move |
|---|---|---|
| Hard 17 or more | Anything | Stand |
| Hard 13 to 16 | 2 to 6 | Stand |
| Hard 13 to 16 | 7 to ace | Hit |
| Hard 12 | 4, 5, 6 | Stand |
| Hard 12 | 2, 3, 7 to ace | Hit |
| Hard 11 | 2 to 10 | Double, otherwise hit |
| Hard 10 | 2 to 9 | Double, otherwise hit |
| Hard 9 | 3 to 6 | Double, otherwise hit |
| Hard 8 or less | Anything | Hit |
| Soft 19 or more | Anything | Stand |
| Soft 18 | 3 to 6 | Double, otherwise stand |
| Soft 18 | 2, 7, 8 | Stand |
| Soft 18 | 9, 10, ace | Hit |
| Soft 17 | 3 to 6 | Double, otherwise hit |
| Soft 15 or 16 | 4, 5, 6 | Double, otherwise hit |
| Soft 13 or 14 | 5, 6 | Double, otherwise hit |
5. Take every double and split the table calls for
Doubling and splitting are the only two moves that put more money on the felt at the moment you are ahead of the dealer. A player who hits and stands correctly but flinches on the doubles and the splits is running a different, worse strategy than the one the 0.5% figure describes.
| Your pair | Dealer shows | Your move |
|---|---|---|
| Aces, eights | Anything | Split |
| Tens, fives | Anything | Never split |
| Nines | 2 to 6, 8, 9 | Split, otherwise stand |
| Sevens, threes, twos | 2 to 7 | Split, otherwise hit |
| Sixes | 2 to 6 | Split, otherwise hit |
| Fours | 5, 6 | Split, otherwise hit |
Four lines account for most of the resistance: hitting 12 against a 2, hitting 16 against a 7, hitting soft 18 against a 9, and splitting eights against a 10. Each one feels wrong at the table and each one is the computed answer.
6. Surrender where the table offers it
Late surrender returns half your stake and ends the hand. It is worth 0.08 points and it applies to exactly four spots: 16 against a 9, a 10 or an ace, and 15 against a 10. Where the button exists, use it on those four hands and nowhere else.
The three bets that cost the most
7. Never take insurance
Insurance asks you to bet that the dealer's hole card is ten-valued, and it pays 2:1. Computed over an eight-deck shoe with an ace showing, 128 of the 415 unseen cards are ten-valued, so the bet lands 30.8% of the time. One unit therefore returns 2 x 128/415 minus 287/415, which is minus 0.075. That is about a 7.5% edge against you, roughly fifteen times the edge on the hand you are already playing.
Insurance is offered on about one hand in thirteen, because that is how often the dealer's upcard is an ace. The offer carries no information about the hole card.
8. Refuse even money on your own natural
Even money is the same bet wearing a different button. Turn over a natural against a dealer ace and the table offers 1:1 immediately instead of the 3:2 you collect whenever the dealer misses. Taking it is insurance at the same 7.5% price, and it feels safe because it removes a push that was never going to cost you anything.
9. Skip the side bets
Perfect Pairs, 21+3 and the bust bets resolve on their own, independently of how you play your hand, and most studios publish no RTP for them at all. A side bet is a separate wager with its own price, closer to the bonus-round bets on the best game show games than to the hand in front of you.
Variants, multipliers and the online shoe
10. Find out what funds a multiplier before backing it
Any blackjack variant paying multipliers funds them from somewhere: a charge per hand on top of your stake, a shortened payout on a normal win, or a rule that costs you elsewhere in the shoe. The studio's own game page states which.
11. Do not carry the standard table into a modified shoe
Power Blackjack removes the 9s and 10s from every deck while keeping the picture cards, per Evolution's own rules. The intuitive read is that thinner ten density means fewer naturals, and it runs backwards. Computed over the published deck composition, a full eight-deck shoe holds 32 aces in 416 cards and deals a natural about 4.75% of the time. Strip out the 9s and 10s and you have 32 aces in 352 cards, and the rate rises to about 4.97%.
Ace density climbs faster than ten density falls. What does change is every hit and stand line above, because the cards you draw to a 12 or a 16 are different cards.
12. Stop trying to count an online shoe
Counting works because the edge moves with what is left. Edward Thorp's 1961 paper, hosted on a university mathematics course page, found that the crucial quantities are "almost linearly dependent on the proportion of tens in the deck", and that the cards becoming visible during play are what makes any winning strategy possible.
An online shoe that reshuffles after every hand has no proportion left to track. The count resets before it can mean anything, so the mechanism Thorp identified has nothing to work on. Live tables dealt from a physical shoe are the only place it survives at all, and eight decks blunt it heavily even there.
Money that never reaches the felt
13. Price an offer at a blackjack table before accepting it
Blackjack usually contributes to a wagering requirement at a reduced rate, which is why an offer that looks strong on the bonus ranking can be worth noticeably less on a blackjack table than the headline suggests. Read the contribution line and the turnover multiple together, work out the total you would have to stake, then price that total against the 0.5% you expect to give back per unit of it.
Key takeaways
- A 6:5 payout costs about 1.36 points on top of a 0.5% game, which is the one rule worth leaving a table over.
- Basic strategy only returns 0.5% if you play all of it, doubles and splits included.
- Insurance and even money are the same wager at about a 7.5% edge against you.
- Studios publish blackjack rules rather than RTP, so compare deck count, soft 17, the payout ratio and double after split.
- A shoe reshuffling after every hand cannot be counted, because nothing is left to track.
- Modified shoes change the strategy table, and removing the 9s and 10s makes a natural more frequent, not less.
Frequently asked questions
Can you actually win at blackjack?
You can win individual sessions, and playing basic strategy cuts the house edge to about 0.5%, but the edge stays negative. No sequence of decisions at a standard online table turns it positive. What changes is how much you expect to give back: at 0.5% and one unit a hand, 200 hands cost you one unit in expectation. At a table paying 6:5 the edge is about 1.9%, and those same 200 hands cost 3.8 units.
What is the house edge in blackjack?
About 0.5% at a standard multi-deck table when you play basic strategy on every hand. Rule changes move it, and a published analysis of blackjack rule effects puts eight decks instead of one at +0.61 points and a dealer hitting soft 17 at +0.21 points. Playing on instinct instead of the table pushes it several times higher. The 0.5% figure describes exact play.
Should you ever take insurance in blackjack?
No. Insurance pays 2:1 that the dealer's hole card is ten-valued, and computed over an eight-deck shoe with an ace showing, 128 of the 415 unseen cards qualify, so it lands 30.8% of the time. One unit returns 2 x 128/415 minus 287/415, which is minus 0.075, or about a 7.5% edge against you. Even money on your own natural is the same bet at the same price.
How much does 6:5 blackjack cost compared with 3:2?
About 1.36 percentage points, computed over the published shoe composition rather than taken from a studio figure. Naturals arrive about 4.75 times per 100 hands in a six or eight deck shoe, and 6:5 returns 1.2 units where 3:2 returns 1.5, so each one costs 0.3 units. Multiply out and you get roughly 1.43 units per 100 staked, settling nearer 1.36 once you remove the naturals that push against a dealer natural. That turns a 0.5% game into roughly a 1.9% game.
Can you count cards in online blackjack?
Not in a game that reshuffles after every hand, which covers most online blackjack. Counting works because the edge moves with the proportion of high cards left in the shoe, and a reshuffle removes that proportion before it can build. Live tables dealt from a physical shoe are the only place the mechanism survives at all, and eight decks blunt it heavily. Basic strategy pays on every hand and needs no tracking.
What is the highest RTP blackjack game?
No studio publishes enough figures for that question to have a clean answer, and Evolution publishes no RTP for any of its blackjack titles. The single live blackjack figure a studio does publish is Pragmatic Play's Prive Lounge Blackjack, described on the studio's own game page as increasing the RTP "up to 99.51%". That "up to" is doing real work, because it is a ceiling attached to particular rule options rather than a rate the table returns by default. Compare rule sheets instead.
Does removing the 9s and 10s in Power Blackjack hurt your odds?
Not in the way most players assume. Computed over the deck composition Evolution publishes, a standard eight-deck shoe holds 32 aces in 416 cards and deals a natural about 4.75% of the time, while removing the 9s and 10s leaves 32 aces in 352 cards and lifts that to about 4.97%. Ace density climbs faster than ten density falls. What does change is every hit and stand line, so the standard strategy table stops applying.
Should you split 8s against a dealer 10?
Yes. A hard 16 is the worst hand in blackjack, and two hands starting from 8 lose less than one hand starting from 16. It is the clearest example of a basic strategy line that feels wrong at the table and is still the computed answer. Always split aces and eights, and never split tens or fives.
How many decks should you look for?
Fewer. A published analysis of blackjack rule effects puts eight decks instead of one at +0.61 points on the house edge, which is more than the entire edge of a well-played standard game. Most online and live blackjack runs on six or eight decks, so the realistic choice is between eight and six rather than eight and one. Take six wherever both are on offer.
Are blackjack side bets worth taking?
No. Perfect Pairs, 21+3 and the bust bets resolve independently of how you play your hand, and most studios publish no RTP for them, so you cannot price the wager before placing it. They also add turnover, and the edge is charged on turnover. The main hand at about 0.5% is the cheapest bet on the table.