Blackjack basic strategy: the complete chart and guide

Every hard total, soft total, and pair decision - the mathematically optimal play for standard S17 multi-deck blackjack, explained and referenced in full

Basic strategy is not a system, a hunch, or a "feel" for the table - it's the specific, mathematically optimal decision for every possible blackjack hand, derived from simulating billions of hands to determine which action (hit, stand, double down, split, or surrender) produces the best long-run expected outcome in each situation. This guide provides the complete reference chart for standard rules, explains the reasoning behind the most important and most commonly misunderstood decisions, and covers how specific rule variations (H17 vs S17, DAS availability, surrender options) shift the correct play.

What basic strategy actually does to the house edge

Without any strategy - playing purely on instinct or superstition - the average blackjack player faces a house edge of roughly 2% to 5%, depending on how far their intuitive decisions deviate from optimal play. Applying perfect basic strategy on every single hand reduces that house edge to approximately 0.5% under standard rules - among the best odds available in any casino game, comparable to the 99.5% RTP figure covered in this site's guide to RTP and house edge.

What this means in real money terms: Playing 100 hands at $10 per hand represents $1,000 in total wagers. At a 4% house edge (a realistic figure for an untrained player making regular mistakes), the expected loss is $40. At the 0.5% house edge basic strategy achieves, the expected loss on the same volume of play drops to $5 - a difference of $35 on just 100 hands, purely from making the mathematically correct decision each time rather than playing on instinct. At higher stakes, or over a longer session, this gap compounds significantly.

What basic strategy does not do: It doesn't eliminate the house edge, and it doesn't guarantee winning any individual session - you'll still lose slightly more than you win over a sufficiently long run, exactly as covered in this site's guide to RTP and house edge. What it guarantees is that you're never voluntarily giving the casino additional edge beyond what the game's rules already provide, on top of the genuine, unavoidable variance that affects any individual session regardless of strategy.

Reading your hand: hard totals, soft totals, and pairs

Before using any strategy chart, you need to correctly identify which of three hand categories you're holding, since each has its own distinct strategy table.

Hard hands: Any hand with no Ace, or with an Ace that must be counted as 1 (because counting it as 11 would bust the hand over 21). A 10-6 is a hard 16. A hand of Ace-6-8 is also a hard hand (the Ace can only count as 1, since counting it as 11 would make the total 25).

Soft hands: A hand containing an Ace counted as 11, where the total doesn't bust. Ace-6 is a "soft 17" - it can be played as either 7 (Ace as 1) or 17 (Ace as 11), giving you flexibility that hard hands don't have, since drawing another card can never bust a soft hand (worst case, the Ace simply converts to counting as 1).

Pairs: Two cards of identical value, which can be split into two separate hands, each receiving an additional card and playing out independently against the dealer's upcard.

Hard totals: the complete strategy

Why hard 12 is the trickiest hard total. Against a dealer showing 4, 5, or 6 - cards that give the dealer a meaningfully elevated bust probability - you stand on 12, accepting the risk of the dealer already beating you rather than risking busting yourself by hitting. Against a dealer 2 or 3, you hit - the dealer's bust probability isn't quite high enough on these upcards to justify standing on a weak 12, so you take the calculated risk of drawing.

Why hard 16 against a dealer 10 is "the worst hand in blackjack." You're a significant underdog whether you hit (high bust risk on a 10-value card landing) or stand (a dealer showing 10 is very likely to have a strong made hand). Basic strategy says hit anyway, because simulations of millions of hands confirm that hitting loses less money on average than standing, even though neither option is genuinely favourable - this is a hand where basic strategy minimises your losses rather than producing a winning expectation.

Soft totals: the complete strategy

Why you should never simply stand on soft 17. A soft 17 can never bust by hitting (the Ace's flexibility means the worst outcome is converting to a hard 7), which means hitting or doubling always carries genuine additional upside with zero downside risk of busting - standing on soft 17 leaves expected value on the table in every dealer-upcard scenario, which is why basic strategy never recommends it.

Why soft 18 is more nuanced than it looks. A total of 18 feels strong, but against a dealer showing 9, 10, or an Ace - upcards that suggest the dealer is likely to make a strong hand of their own, potentially 19 or 20 - soft 18 is actually an underdog, and hitting (converting your Ace to count as 1, giving you a genuine hard 8 to rebuild from) produces better long-run expected value than standing on a soft 18 that's likely to lose regardless.

Pairs: the complete strategy

Why you always split Aces. Two hands each starting with an Ace are dramatically stronger than a single hand of soft 12 - each split hand has a genuine, strong chance at drawing to 21 or close to it, whereas the combined soft-12 starting point is comparatively weak. Note that most casinos restrict you to receiving exactly one additional card per hand after splitting Aces, and prohibit re-splitting or doubling on the resulting hands.

Why you never split 10s, even though it "feels" tempting against a weak dealer upcard. A pair of 10s is a hard 20 - statistically one of the strongest possible starting hands in blackjack. Splitting trades a near-certain winner for two separate hands that each start from a single 10, meaningfully weaker starting positions individually, even against a dealer showing a genuinely vulnerable upcard like 5 or 6. Stand on 10s, always.

Why you split 8s even against a dealer 10, despite hard 16 being "the worst hand in blackjack." This is the pairing decision that surprises the most players. Since hard 16 is already a losing proposition regardless of how you play it (covered above), splitting 8s gives you two separate hands each starting from 8 - a meaningfully better starting position than a combined hard 16, even against a strong dealer upcard. Simulations consistently show splitting loses less money per dollar wagered than either hitting or standing on the combined 16.

Insurance: always decline (without card counting)

Insurance is a side bet offered when the dealer shows an Ace, paying 2:1 if the dealer has blackjack. It's structurally a bad bet for a basic-strategy player: the dealer has blackjack only around 30.8% of the time when showing an Ace, meaning the true break-even point for a 2:1 payout would require roughly a 33.3% probability - insurance pays less often than the odds it offers would require to break even, producing a clearly negative expected value.

The only exception: A skilled card counter tracking the ratio of high to low cards remaining in the shoe can, in specific genuinely ten-rich counts, identify moments where insurance becomes a positive-expectation side bet - but this requires active card counting to identify correctly, and is irrelevant to basic strategy play. For any player not actively counting cards, the correct decision is always to decline insurance, without exception.

How table rules change the correct strategy

Basic strategy charts assume specific standard rules - deviations from these rules shift several correct decisions, sometimes meaningfully.

H17 vs S17 (dealer hits or stands on soft 17): When the dealer is required to hit soft 17 rather than stand, the house edge increases by approximately 0.2% compared to S17 rules, and specific plays change - doubling on hard 11 against a dealer Ace becomes correct under H17, along with doubling on soft 19 against a dealer 6, neither of which are correct under standard S17 rules.

DAS (Double After Split) availability: When DAS isn't permitted, several marginal pair-splitting decisions change - splitting 4s against a dealer 5-6 becomes a straightforward hit instead of a split, and 6,6 against a dealer 2 similarly shifts from split to hit, since the doubling option that made those splits worthwhile is no longer available on the resulting hands.

Surrender availability: Where late surrender is offered (forfeiting half your stake to exit a clearly unfavourable hand rather than playing it out), basic strategy adds specific surrender recommendations - most commonly hard 16 against a dealer 9, 10, or Ace, and hard 15 against a dealer 10, situations where the expected loss from surrendering half your stake is genuinely smaller than the expected loss from playing the hand out.

Single-deck and double-deck games carry their own subtly adjusted strategy charts relative to standard 4-8 deck shoe games - most notably, doubling down on a hard 8 becomes correct in some single-deck situations, a play that's never correct in a standard multi-deck game.

How casino-style live dealer and RNG blackjack compare for strategy purposes

Basic strategy applies identically whether you're playing RNG (random number generator) blackjack or a live dealer table - the mathematically optimal decision for any given hand and dealer upcard doesn't change based on how the cards are dealt. What differs, as covered in detail in this site's guide to why card counting doesn't work at online casinos, is that RNG blackjack (dealt from a freshly randomised virtual shoe every hand) and most live dealer formats (using continuous shuffle technology) both structurally prevent the deck-depletion tracking that card counting requires - meaning basic strategy, not counting, is the complete edge-optimisation toolkit available in both formats.

Real-world data: what player mistakes actually cost, according to Nevada's own Gaming regulator

Beyond the theoretical house-edge figures covered above, there's genuine real-world regulatory data confirming how much basic strategy mistakes cost players in practice, worth citing directly rather than relying purely on simulated examples.

The Nevada Gaming Control Board's 2009 revenue data: Total wagering on blackjack tables in Nevada that year was $8.917 billion, with casinos winning $1.008 billion - an actual realised win rate of approximately 11.3%, dramatically higher than blackjack's theoretical 0.5% house edge under perfect basic strategy. According to gaming consultant Bill Zender's analysis of this gap, approximately 0.83 percentage points of that realised casino advantage is directly attributable to player mistakes - meaning real-world players, in aggregate, are giving up nearly double the game's built-in theoretical edge purely through imperfect strategy execution, before accounting for other factors like side bets and rule variations across different tables.

Why this real data point matters more than a simulated example. It confirms, using actual regulatory financial data rather than a hypothetical scenario, that the gap between basic-strategy blackjack and casually-played blackjack isn't a small theoretical nuance - it's a measurable, substantial difference in real casino win rates, observed across an entire state's blackjack volume over a full year.

If you play mixed H17/S17 games and only want to memorize one chart

This is a genuinely practical piece of guidance for anyone who plays at more than one casino, or who plays both online and at physical tables where the soft-17 rule sometimes differs.

The specific recommendation from professional blackjack analysis: If you're going to memorize only one basic strategy chart despite occasionally encountering both H17 and S17 tables, memorize the S17 (dealer stands on soft 17) version. The reasoning is precisely quantified: the cost of applying S17 strategy at an H17 table is meaningfully smaller than the reverse - applying H17 strategy at an S17 table costs approximately 2.3 times more in expected losses than the other way around. In other words, if you're only going to learn one chart, the S17 chart is the safer default to carry into an unfamiliar table.

Composition-dependent strategy: the next level beyond what this guide covers

Everything in this guide's charts is what's called total-dependent basic strategy - the correct play based purely on your hand's total value (a hard 16, a soft 18) and the dealer's upcard, regardless of which specific cards make up that total. This is standard basic strategy, and it's what the overwhelming majority of players should learn and use.

What composition-dependent strategy adds. In a very small number of specific situations, the exact cards making up your total (not just the total itself) can shift the mathematically optimal play by a tiny margin - for example, a hard 16 made from 10-6 is treated identically to a hard 16 made from 9-4-3 under standard total-dependent strategy, but in a small number of edge cases, composition-dependent strategy would treat these differently based on which specific cards have already been removed from the shoe.

Why this guide doesn't include composition-dependent deviations. The edge gained from composition-dependent play over standard total-dependent basic strategy is extremely small - a small fraction of a percentage point - and the number of additional situations to memorise is substantial relative to that tiny gain. For the overwhelming majority of players, mastering standard total-dependent basic strategy (everything covered in this guide) captures effectively all of the achievable edge improvement available without card counting. Composition-dependent refinements are genuinely a "next level" consideration primarily relevant to dedicated advantage players who have already fully mastered standard basic strategy and are looking for the last fractional improvements available.

What to do when the chart's recommended play isn't available

Basic strategy charts assume you can always double or split when the chart recommends it - but table-specific restrictions sometimes remove that option. Here's the correct fallback in each case.

If the chart says double, but you can't (because you already have three or more cards, or the specific table doesn't allow doubling on your current total): Hit instead - with one specific exception: a soft 18 that the chart recommends doubling should be played as a stand if doubling isn't available, not a hit, since standing preserves the hand's existing strength better than hitting would in that specific situation.

If the chart says split, but you can't (due to a table limit on re-splitting, most commonly relevant to a pair of Aces you've already split once): Treat the hand as its equivalent hard total instead - a pair of 8s you can't split is played as a hard 16; a pair of 10s you wouldn't have split anyway is simply a hard 20.

If the chart says surrender, but surrender isn't offered at your table: Hit instead, applying the standard hard-total strategy for that specific total against that specific dealer upcard.

How to actually learn basic strategy

Start with hard totals, then move to soft totals, then pairs - in that order of frequency. Hard totals come up far more often in actual play than soft totals or pairs, making them the highest-value category to memorise first for the fastest practical improvement in your results.

Most casinos, including online operators, permit reference charts at the table. Using a printed or on-screen basic strategy chart while playing is legal and widely permitted - it will slow your play pace, which is why serious, high-volume players eventually memorise the full chart, but there's no rule against reference use, and it's a perfectly reasonable way to play correctly while you're still learning.

Realistic learning timeline: Most players can absorb the most important, highest-frequency decisions (hard totals, the never-split-10s and always-split-Aces rules, never taking insurance) within an hour of focused study. Full memorisation of every hand across all three categories - hard totals, soft totals, and every pair combination - typically takes 5 to 10 hours of deliberate practice for most players to genuinely internalise to the point of confident, quick table-side recall.

Frequently asked questions

Does basic strategy guarantee i'll win at blackjack?

No. Basic strategy minimises the house edge to approximately 0.5% under standard rules, but doesn't eliminate it - you'll still lose slightly more than you win over a sufficiently long run. What it guarantees is that you're never giving the casino additional edge beyond the game's inherent mathematics through avoidable mistakes.

What's the difference between a hard hand and a soft hand?

A hard hand has no Ace, or an Ace that can only count as 1 without busting. A soft hand has an Ace counted as 11, giving it flexibility since the Ace can convert to counting as 1 if needed - meaning a soft hand can never bust by taking one additional card.

Should i ever take insurance?

No, never, unless you're actively counting cards and have identified a specific ten-rich count situation. For any player not counting cards, insurance carries a negative expected value regardless of your own hand strength, since the dealer only has blackjack roughly 30.8% of the time when showing an Ace - below the break-even threshold the 2:1 payout would require.

Why does basic strategy say to hit hard 16 against a dealer's 10, even though it feels risky?

Hard 16 against a dealer 10 is a losing proposition regardless of how you play it - simulations of millions of hands confirm hitting loses less money on average than standing, even though neither choice produces a favourable outcome. Basic strategy in this specific case is about minimising an unavoidable loss, not producing a win.

Does basic strategy change between online and live dealer blackjack?

No - the mathematically optimal decision for any hand against any dealer upcard is identical regardless of whether you're playing RNG or live dealer blackjack. What does differ is that card counting doesn't work in either modern format, since both use fresh randomisation or continuous shuffle technology that prevents the deck-depletion tracking counting requires.

This guide is for educational purposes. Basic strategy minimises but does not eliminate the house edge, and no strategy guarantees winning outcomes. Gambling problem? In the UK: BeGambleAware.org | 0808 8020 133. In the US: 1-800-GAMBLER. Information correct as of September 2026.