Blackjack strategy — basic strategy, card counting, honestly reviewed

A blackjack strategy that genuinely cuts the house edge is rare among casino games — and blackjack is the rare game where the player actually has one. Unlike most games on the casino floor, your decisions change the math. Played correctly, blackjack’s house edge drops to roughly 0.5%, lower than almost anything else on the floor. But “correctly” is doing a lot of work in that sentence: it means following basic strategy on every hand, ignoring hunches, and knowing which of the popular “systems” actually move the needle and which just rearrange your variance. We’ll be honest throughout: no strategy makes blackjack a positive-expectation game for the average online player — but basic strategy gets you closer to break-even than any other casino bet.

Basic strategy: the only thing that actually cuts the edge

Basic strategy is the mathematically optimal play for every possible hand, derived from the two facts you can always see: your total and the dealer’s upcard. It isn’t a betting system and it isn’t a trick — it’s the answer to a solved problem. Computers have simulated every blackjack decision billions of times, and for each combination of your cards versus the dealer’s upcard there is exactly one play that loses the least money over the long run. Learn that play, make it every time, and you hold the house edge near 0.5% in a typical multi-deck game where the dealer stands on 17.

The four decisions you control

Every hand comes down to a choice among a small set of actions. Get these right and you’ve captured almost all of basic strategy’s value:

• Hit — take another card. Correct when your total is low enough that busting is unlikely and the dealer’s upcard is strong (7 through Ace).
• Stand — take no more cards. Correct on stiff hands when the dealer shows a weak upcard (2 through 6) and is likely to bust.
• Double down — double your bet for exactly one more card. The aggressive, profitable move on totals of 10 and 11, and on some soft hands, when the dealer is weak.
• Split — when dealt a pair, separate it into two hands with a second matching bet. Always split Aces and 8s; never split 10s or 5s.

Reading the dealer’s upcard

The single most important input is the dealer’s upcard, because it predicts how often the dealer busts. When the dealer shows 2 through 6, they are far more likely to break — so you play conservatively, standing on stiff totals (12–16) and letting them take the risk. When the dealer shows 7 through Ace, they will probably make a strong hand, so you must improve yours: hit your stiffs even though busting hurts, because standing pat loses more often. This single asymmetry explains most of the chart.

Use a basic-strategy chart — and why we don’t paste one here

The complete set of decisions is conventionally printed as a colour-coded grid: your hand down the side, the dealer’s upcard across the top, one cell per situation. It’s the single most useful reference in blackjack, and reputable charts are freely available; we won’t reproduce a copyrighted chart image here, but find one for the exact rules you play (number of decks, whether the dealer hits or stands on soft 17) and keep it open while you learn. Most online tables let you consult it freely — there’s no clock pressure as there can be at a live table. Run hands on the free table until the correct play is automatic before you risk real money.

More blackjack variants

Play at casino →

Single Deck Blackjack

★ 4.9/5

👍11👎0

RTP: 99.80%

Play at casino →

Double Exposure

★ 4.5/5

👍5👎1

RTP: 99.30%

Play at casino →

Blackjack Switch

★ 4.5/5

👍6👎0

RTP: 99.40%

Play at casino →

Spanish 21

★ 4.4/5

👍4👎0

RTP: 99.40%

Insurance and surrender: the side decisions

Two extra options round out the basic-strategy picture. Insurance is offered when the dealer shows an Ace: a side bet that the dealer has blackjack, paying 2:1. Basic strategy’s verdict is blunt — never take insurance. It’s a sucker bet with its own house edge, unrelated to your hand. Surrender, where offered, lets you fold a bad hand and reclaim half your stake; it’s correct on a hard 16 against a dealer 9, 10, or Ace, and on hard 15 against a 10. Used in exactly those spots it saves money; used emotionally it bleeds it.

Why blackjack pays 3:2 — and the 6:5 trap

A natural blackjack (an Ace plus a ten-value card on your first two cards) should pay 3:2 — bet $10, collect $15. This single rule is worth a large chunk of the game’s favourable math. In recent years many tables, especially single-deck and “party pit” games, quietly pay 6:5 instead — that same $10 bet returns only $12. The switch roughly triples the house edge, from about 0.5% to over 1.5%, no matter how perfectly you play. The lesson is simple and absolute: check the payout before you sit down, and never play a 6:5 table when a 3:2 one is available.

Card counting: real, hard, and useless online

Basic strategy gets you to break-even territory; card counting is the only widely known method that can push the edge to the player — and it works only under conditions that online play removes entirely. It’s worth understanding precisely so you know why it can’t help you on an RNG table.

Hi-Lo: running count to true count

The most common system is Hi-Lo. You assign a value to each card as it appears: low cards (2–6) count +1, neutral cards (7–9) count 0, and tens and Aces count −1. Summing these as the shoe is dealt gives the running count. A high positive count means many low cards are gone, so the remaining shoe is rich in tens and Aces — which favours the player, because it makes dealer busts and player blackjacks more likely. To act on it you convert to the true count by dividing the running count by the number of decks still unplayed. The true count is what drives decisions.

Bet spread and the actual edge

Counting doesn’t change how you play most hands — you still lean on basic strategy. What it changes is how much you bet: small when the true count is low or negative, large when it’s high, a pattern called a bet spread. Done well against a favourable live shoe, this can flip the house edge into a thin player advantage of around 0.5% to 1% — real, but small, requiring a large bankroll and long hours to overcome variance. It is legal, but casinos may bar players they suspect of it, so skilled counters disguise their spread.

Why it fails on every online RNG table

Here is the part that matters for most readers: card counting does nothing against online blackjack. A standard online table uses a random number generator that reshuffles a full shoe before every single hand. The count never carries from one hand to the next, because there is no persistent shoe to deplete — the deck is effectively new every time. No amount of tracking helps when the past tells you nothing about the future. Counting requires a physical shoe dealt deep before a shuffle: that exists only at a live table (and live-dealer studio games are watched and shuffled specifically to defeat it). Online, basic strategy is the whole game.

Betting progressions: why Martingale and friends don’t beat the edge

If you’ve read about blackjack “strategy” elsewhere, you’ve probably met the betting systems — Martingale, Paroli, the rest. They are seductive, they feel like a plan, and they cannot beat the house edge. Not “rarely”; not ever. This deserves a clear, skeptical look, because the marketing around these systems is relentless.

Martingale: double-after-loss, and the wall it hits

Martingale says: after every loss, double your next bet, so the first win recovers everything plus one unit. On paper it never loses. In reality two walls stop it. First, the table limit: a run of losses pushes your required bet past the maximum, and the chain breaks. Second, your bankroll: starting at $10, eight losses in a row require a ninth bet of $2,560 to stay in the system — and losing streaks that long are not rare across a session. When the chain breaks, one bad run erases dozens of small wins. The expected loss per dollar wagered is unchanged from flat betting; all Martingale does is trade a high chance of a small win for a small chance of a catastrophic one.

Paroli, D’Alembert and the rest: same math, gentler ride

Paroli (reverse Martingale) doubles after wins instead of losses, so you risk house money on streaks and cap your downside. D’Alembert nudges the bet up one unit after a loss and down one after a win. Both are calmer than Martingale and easier on the bankroll — but neither changes a thing about expected value. The house edge is a property of the rules and the cards, applied to every wager independently. No sequence of bet sizes can alter the average of a series of independent negative-expectation bets. What progressions genuinely do is redistribute variance — reshaping the spread of outcomes, making wins more frequent but smaller or losses rarer but larger. That can suit your taste for risk; it is not an edge.

The gambler’s fallacy underneath all of them

Every progression leans, quietly, on the idea that a loss makes the next win “due.” It does not. Each hand is dealt from a freshly shuffled shoe online, and even at a live table the cards have no memory of your bet size. After five losses your next hand is governed by exactly the same probabilities as your first. Chasing losses with bigger bets doesn’t correct anything — it just enlarges your exposure to the same edge. Discipline beats any progression: set a bankroll, flat-bet it, play perfect basic strategy, and stop at your target or your stop-loss.

How to choose your approach

You don’t pick a betting “system” in blackjack the way gamblers chase one at other casino games. You pick rules, then you play them correctly. Here’s the order of priority that actually protects your money.

Step 1: find the best rules before you find a seat

Rules matter more than anything you do at the table. Prefer 3:2 blackjack payouts (never 6:5), dealer stands on soft 17 over hits, fewer decks, and tables that allow double after split and late surrender. A good rule set can shave the house edge to around 0.3%; a bad one (6:5, dealer hits soft 17) can push it past 2%. Our rules guide breaks down how each rule shifts the math, and how we rate explains the audit behind every table review.

Step 2: learn basic strategy cold

There is no shortcut here and no substitute. Drill the chart for your specific rule set until the right play is reflex. Start on the free table where you can keep a chart open and pay nothing for mistakes. This single step is worth more than every betting system combined.

Step 3: set a bankroll and bet flat

Decide your session bankroll and your unit size in advance, and keep the bet the same unless you are genuinely counting at a live shoe. Flat betting maximises your time at the table for a given risk and removes the emotional pull of chasing. Set a win target and a stop-loss, and honour both.

Step 4: skip insurance, skip side bets

Insurance and the colourful side bets (Perfect Pairs, 21+3, and the like) all carry house edges far above the main game — often 5% or more. They’re entertainment, not strategy. If you came to lower the edge, leave them alone.

Verdict: where the real edge lives

Blackjack rewards the player who does the boring thing. Basic strategy is the whole edge for an online player — master it and you hold the house to around 0.5%, better than any other bet in the casino. Card counting is real but irrelevant online, where every hand is a fresh shuffle; it belongs to disciplined live-shoe players with big bankrolls and longer patience than most. Betting progressions are a story we tell ourselves — Martingale, Paroli and the rest redistribute variance without touching the edge, and the math behind that is not in dispute.

So start simple. Pick a 3:2 table with player-friendly rules, drill the chart on the free table until it’s automatic, set a bankroll you can lose without flinching, and flat-bet your way through perfect basic strategy. That’s not a magic system. It’s the closest thing to a fair fight the casino offers — and the only blackjack “strategy” the math actually backs.