Perfect blackjack
basic strategy.
Every player hand against every dealer up card. Pick your rule set or your casino, hover any cell for the expected return, and memorize the chart. House edge drops from ~2.5% (typical play) to under 0.5% (perfect play) — the single biggest EV gain available in any casino on the floor.
What the cells mean
Hover any cell in the chart for the expected return per $1 bet · hard totals run high → low (18+ at top, 5–8 at bottom)
Pick your rule set or your casino
Click a casino to jump straight to the rule set its primary blackjack offering uses. Live-dealer tables and in-house Originals can run different rules — the chip tooltip shows every variant an operator runs.
6-deck · dealer stands on soft 17 · DAS
The default reference chart. What most regulated live-dealer tables run, including Evolution's Classic Speed Blackjack and the standard Pragmatic / Playtech multi-seat live tables on Stake, BC.Game, Cloudbet, and Bitcasino.
| Player | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | A |
|---|---|---|---|---|---|---|---|---|---|---|
| 18+ | ||||||||||
| 17 | ||||||||||
| 16 | ||||||||||
| 15 | ||||||||||
| 14 | ||||||||||
| 13 | ||||||||||
| 12 | ||||||||||
| 11 | ||||||||||
| 10 | ||||||||||
| 9 | ||||||||||
| 5–8 |
| Player | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | A |
|---|---|---|---|---|---|---|---|---|---|---|
| A,8+ | ||||||||||
| A,7 | ||||||||||
| A,6 | ||||||||||
| A,5 | ||||||||||
| A,4 | ||||||||||
| A,3 | ||||||||||
| A,2 |
| Pair | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | A |
|---|---|---|---|---|---|---|---|---|---|---|
| A,A | ||||||||||
| 10,10 | ||||||||||
| 9,9 | ||||||||||
| 8,8 | ||||||||||
| 7,7 | ||||||||||
| 6,6 | ||||||||||
| 5,5 | ||||||||||
| 4,4 | ||||||||||
| 3,3 | ||||||||||
| 2,2 |
Practice the chart
Charts memorize through repetition, not reading. Random hand, random dealer card. Pick the action; see the correct play and the reasoning. The trainer skews toward decision-heavy spots (hard 12-16, soft hands, pair 8s and 9s) where most money gets left on the table.
Practice the chart
The math behind the chart
Every cell in this chart is the action with the highest expected value (EV) for that exact situation, computed by simulating billions of hands. There is no judgment call. There is no intuition. The chart is a solved game.
The dealer's strategy is fixed and public: hit until 17, then stop. Knowing this, every player decision can be modeled as a probability distribution: what's the chance the dealer busts given their up card, vs. the chance you bust if you hit, vs. the chance the next card improves your hand.
Example: you have hard 16, dealer shows 10. If you hit, 62% of cards bust you. If you stand, you lose ~77% of the time anyway because the dealer reaches 17–21. Hitting is slightly better. Surrendering (giving up half the bet) is even better when available.
The 0.5% house edge that remains after perfect basic strategy is what you are inherently paying to participate. To get below it requires card counting, which crypto casinos with auto-shuffled or RNG-shoe blackjack make impossible.
Three patterns and you've got 80%
When the dealer shows 4, 5, or 6 the bust probability spikes (35–42%). You stand on hard 12+ and double on more hands than against any other dealer card.
When the dealer shows 7 through Ace, you have to make your hand because the dealer probably will. Hit hard totals up to 16; stand on 17+.
Always split aces and 8s. Never split 5s and 10s. Memorize these four; the other pairs follow the dealer-bust-card logic above.
When the table changes the math
Skip these. The math is bad.
Side bets are sold as "extra fun" but exist because the house edge on them is 5 to 15 times worse than the main game. A perfect-strategy main bet runs 0.5%. The numbers below are why you stop after the third one.
| Bet | House edge | What it pays on |
|---|---|---|
| Perfect Pairs | ~6.0% | First two cards are a pair (mixed / coloured / perfect). The "perfect pair" tier pays best but lands ~1.4% of hands. |
| 21+3 | ~3.2 to 7.8% | Player's two cards plus dealer up card form a 3-card poker hand. Edge depends on paytable; most crypto-casino tables sit at the higher end. |
| Bust It | ~6.9% | Dealer busts. Pays bigger when the bust is on more cards — variance is high, EV is bad. |
| Lucky Ladies | ~17% | Hand totals 20, with multipliers for matched 20s and Q-Q of hearts. The worst common side bet on the floor. |
| Insurance | ~7.4% | Half-bet that the dealer's hole card is a 10 when they show an Ace. Mathematically a separate bet on the hole card; declines unless you're counting. |
The single rule: don't take any side bet, ever. They exist to claw back the value basic strategy gives you. Take insurance only if you're counting cards (true count above +3) and the running edge has flipped.
The honest answer for crypto-casino blackjack
Hi-Lo card counting is the simplest and most popular system. You assign +1 to low cards (2-6), 0 to neutrals (7-9), and -1 to high cards (10-A). Add them as the shoe is dealt. When the running count is high relative to decks remaining, the undealt shoe is rich in 10s and aces — better for the player than the dealer.
Hi-Lo card values 2,3,4,5,6 → +1 7,8,9 → 0 10,J,Q,K,A → -1 True count = running count / decks remaining Bet bigger when true count > +1 House edge flips at roughly true count > +2
Done well, counting flips the 0.5% house edge into a ~0.5% player edge — small per hand, but real over volume. It is legal everywhere, but private casinos can ban you for it. The catch:
Counting does not work in any crypto-casino blackjack you can play right now.
RNG-shoe BJ (Stake Originals, BC Originals, similar) reshuffles the virtual deck every hand. Live dealer studios (Evolution, Pragmatic Live) use auto-shufflers that re-randomise after every shoe and often after every hand. The state that counting depends on, undealt cards being skewed, never builds up. The math is sound; the application is dead.
The only place counting still pays is in person, at a hand-shuffled shoe with deep penetration. Treat counting as cool theory, not a viable edge online.
