Blackjack strategy: the complete basic strategy chart and the maths behind it

Basic strategy is the mathematically best decision for every hand you can hold against every card the dealer can show. Learn it and a good table costs about 0.4% per hand. Ignore it and the same table costs five times as much.

By CasinoTrust Editorial TeamUpdated 13 min readFact-checked against the Ranking Protocol

Unlike every roulette system, blackjack strategy is not a belief. It is the output of a full calculation of every possible hand against every possible dealer card, and it has been verified by computer simulation billions of times over. This guide from the blackjack academy gives the complete chart for the most common online rule set, explains the logic behind its lines, quantifies what it saves, then covers insurance, card counting and the side bets that undo all of it.

Basic strategy chart: hard totals

A hard total has no ace counted as 11. Key: H = hit, S = stand, D = double (hit if not allowed), Rs = surrender if offered, otherwise stand, Rh = surrender if offered, otherwise hit.

Hard totals vs dealer up-card (multi-deck, S17, DAS)
Your hand2345678910A
5–8HHHHHHHHHH
9HDDDDHHHHH
10DDDDDDDDHH
11DDDDDDDDDH
12HHSSSHHHHH
13SSSSSHHHHH
14SSSSSHHHHH
15SSSSSHHHRhH
16SSSSSHHRhRhRh
17+SSSSSSSSSS

Basic strategy chart: soft totals

A soft total counts an ace as 11. Because it cannot bust on one card, you double far more often. Ds = double if allowed, otherwise stand.

Soft totals vs dealer up-card (multi-deck, S17, DAS)
Your hand2345678910A
A,2 (soft 13)HHHDDHHHHH
A,3 (soft 14)HHHDDHHHHH
A,4 (soft 15)HHDDDHHHHH
A,5 (soft 16)HHDDDHHHHH
A,6 (soft 17)HDDDDHHHHH
A,7 (soft 18)SDsDsDsDsSSHHH
A,8 (soft 19)SSSSSSSSSS
A,9 (soft 20)SSSSSSSSSS

Basic strategy chart: pairs

P = split. Where the chart says not to split, play the hand as its hard total (5,5 = hard 10; 10,10 = hard 20). The 2,2 / 3,3 / 4,4 / 6,6 splits against small cards depend on DAS being allowed; without DAS, hit 4,4 always, hit 2,2 and 3,3 against a 2 and 3, and hit 6,6 against a 2.

Pairs vs dealer up-card (multi-deck, S17, DAS)
Your hand2345678910A
A,APPPPPPPPPP
10,10SSSSSSSSSS
9,9PPPPPSPPSS
8,8PPPPPPPPPP
7,7PPPPPPHHHH
6,6PPPPPHHHHH
5,5DDDDDDDDHH
4,4HHHPPHHHHH
3,3PPPPPPHHHH
2,2PPPPPPHHHH

Adjustments for other rule sets

  • Dealer hits soft 17 (H17): double 11 against an ace; double A,8 (soft 19) against a 6; double A,7 against a 2; surrender 15 and 17 against an ace (17 only if surrender is allowed — otherwise stand). Everything else is unchanged.
  • European no-hole-card (ENHC): the dealer may still have blackjack after you have doubled or split and you lose the extra bets. Do not double 11 against a 10; do not split A,A or 8,8 against an ace; do not split 8,8 against a 10. Play those as hit.
  • No DAS: see the pairs note above. Double only on 10–11: hit the hands shown as D on 9 and all soft totals.
  • Single and double deck: a few more doubles and splits become correct (for example double 9 against 2 and 11 against an ace in single deck). Use a deck-specific chart; the multi-deck chart costs only about 0.03% extra if you use it there.

Why basic strategy works

Every line in the chart is the answer to the same question: for this hand against this up-card, which action has the highest expected value? The logic falls into a handful of patterns that make the chart far easier to remember than 300 individual cells.

The reasoning behind the main patterns
PatternWhy
Stand on 12–16 against 2–6The dealer busts 35–43% of the time from these up-cards. Your stiff hand wins more often by waiting than by risking a bust yourself (hitting 16 busts 62% of the time).
Hit 12 against 2 and 3Only four ranks (10, J, Q, K) bust a 12, and the dealer busts least often from 2 and 3 among the low cards. Hitting is marginally better.
Hit 12–16 against 7–AThe dealer finishes with 17+ roughly 74–88% of the time from these cards. Standing on 15 loses about 57% net; hitting loses slightly less. Both are bad; hitting is less bad.
Double 10 and 11 against small cardsYou are 31–38% to make 20 or 21 with one card and the dealer is likely to bust or make a weak hand. Doubling roughly doubles a positive expectation.
Double soft hands against 4–6You cannot bust, the dealer is weak, and you gain a second bet in a spot where you are already the favourite.
Always split aces and eightsA,A: two chances at 21 instead of a soft 12. 8,8: hard 16 is the worst hand in the game; two hands starting from 8 lose less, even against a 10.
Never split tens or fivesHard 20 wins about 70% of the time and pushes another 10%; breaking it up is throwing money away. 5,5 is hard 10, a doubling hand.
Surrender 16 vs 9/10/A and 15 vs 10When your expected loss from playing exceeds 50% of the bet, giving up half is the cheaper option. Hard 16 against a 10 loses about 54% net when hit.

What it costs to ignore

House edge at a 6-deck, S17, DAS, 3:2 table by playing style
How you playApproximate house edgeExpected loss per hour at €10/hand, 60 hands
Perfect basic strategy with late surrender0.33%€2.00
Perfect basic strategy, no surrender0.40%€2.40
Basic strategy with common errors (no soft doubles, no surrender, never hit 12–16 vs 7+)1.5–2%€9–€12
"Never bust" (stand on all 12+)≈4%€24
Mimic the dealer (hit to 16, never double or split)≈5.5%€33
Same table paying 6:5, perfect basic strategy≈1.8%€11

The typical self-taught player sits somewhere around 2%, five times the cost of the same table played correctly. Ten minutes a day on the basic strategy trainer for a week closes that gap permanently. For any hand you are unsure about, the blackjack chance calculator shows the expected value of each option.

Insurance: the maths

Insurance is a side bet that the dealer's hole card is a ten-value card, offered when the up-card is an ace. It pays 2 to 1 and you may bet up to half your original stake. Whether it is a good bet depends on one number only: the share of ten-value cards left in the shoe.

Insurance in a 6-deck shoe (312 cards, 96 tens), dealer shows an ace
Your two cardsTens remaining / cards remainingP(dealer has ten)Fair payoutExpected value of a €5 insurance bet
No tens (e.g. 7-9)96 / 30931.1%2.22 to 1−€0.34 (−6.8%)
One ten (e.g. 10-6)95 / 30930.7%2.25 to 1−€0.39 (−7.8%)
Two tens (hard 20)94 / 30930.4%2.29 to 1−€0.44 (−8.7%)
Blackjack (A-10), taking even money95 / 30930.7%2.25 to 1About 3.9% of the main bet given away

Insurance would break even only if more than one card in three remaining were a ten (33.3%). A fresh shoe has 30.8%, and the cards you hold make it slightly worse. Only a card counter who knows the shoe is ten-rich (Hi-Lo true count of roughly +3 or higher) should ever take it. Even money on your own blackjack is mathematically identical: you are declining an expected 1.5 × 69.3% = 1.039 units to accept a guaranteed 1.0 unit, a 3.9% giveaway.

Card counting: how it works and why it barely works online

Basic strategy assumes a freshly shuffled shoe. Card counting tracks how the remaining shoe differs from that. A shoe rich in tens and aces favours the player — more blackjacks paid at 3:2, more successful doubles, more dealer busts from stiff hands — while a shoe rich in small cards favours the dealer. Counters bet more when the shoe is rich and the minimum when it is poor. It is legal everywhere, and casinos are free to refuse your action.

The Hi-Lo system

Hi-Lo card values
CardsCount valueWhy
2, 3, 4, 5, 6+1Small cards leaving the shoe make it richer in high cards
7, 8, 90Neutral
10, J, Q, K, A−1High cards leaving the shoe make it poorer
  1. Keep the running count

    Start at 0 after the shuffle. Add or subtract the value of every card you see, including other players' and the dealer's. A full deck sums to zero (balanced count).

  2. Convert to a true count

    Divide the running count by the number of decks still to be dealt. Running count +8 with four decks left is a true count of +2; the same +8 with one deck left is +8. The true count is what measures your edge.

  3. Size your bet

    Each true-count point is worth about 0.5% to the player. Starting from −0.4% at a good table, you are roughly break-even at +1 and have about +0.5% at +2, +1% at +3. A typical spread bets 1 unit at true count ≤ 1, then 2, 4, 8 units at +2, +3, +4 and above (a 1–8 spread in 6 decks; 1–12 is common in 8).

  4. Adjust a few plays

    The most valuable deviations (the "Illustrious 18") include taking insurance at +3, standing on 16 vs 10 at 0 or higher, standing on 15 vs 10 at +4, and doubling 10 vs 10 at +4. Everything else stays basic strategy.

Done well in a casino with 75%+ penetration and a 1–12 spread, Hi-Lo earns around 0.5–1% of the average bet over the long run — with a standard deviation per hand of more than one bet, which is why counters need bankrolls of 500–1,000 units to survive the swings.

Why counting fails online

Card counting conditions: land-based vs online
ConditionGood land-based gameLive dealer onlineRNG online
Penetration (share of shoe dealt)75–80%≈50% (cut card at 4 of 8 decks)0% — reshuffled every hand
Continuous shuffle machineSome tablesCommon on many studio tablesn/a
Hands per hour60–10050–70200+
Maximum bet spread tolerated1–12 or more before heatLimited by table max; 1–10 typicalIrrelevant
Counter's realistic edge+0.5% to +1%≈0% to +0.2% at bestNone
CountermeasuresBack-off, flat-bet request, banShallow cut, CSM, account restrictionNone needed

With eight decks dealt to the halfway point, high true counts almost never arrive before the shuffle; the times you would bet big are cut away. Combined with a slow game and the same house right to close your account, live dealer counting is a hobby, not an edge. Anyone selling a course on "online card counting" is selling a course.

Side bets to avoid

Nearly every online table offers optional side bets. They are separate wagers, they do not change the main game's edge, and they are priced like slot machines. Exact figures depend on the paytable and the number of decks; the ranges below cover common versions.

Common blackjack side bets and house edges
Side betWhat it pays onTypical house edge
InsuranceDealer hole card is a ten≈7%
Perfect PairsYour first two cards are a pair (mixed, coloured, perfect)4–11%
21+3Your two cards plus the dealer up-card make a flush, straight, three of a kind or straight flush3–7%
Lucky LadiesYour first two cards total 20 (bonuses for queens of hearts)17–25%
Bust It / Buster BlackjackDealer busts, paid by number of cards6–10%
Royal MatchSuited first two cards4–7%
Lucky LuckyYour two cards plus dealer up-card total 19–212.5–7%

A player betting €10 on the main hand at 0.4% and €5 on 21+3 at 5% has an expected loss of €0.04 on the game and €0.25 on the side bet. The side bet is six times the cost for half the stake. If you want a long-shot in the session, it is still cheaper than most, but do not confuse it with blackjack.

Bankroll for blackjack

Blackjack's edge is small and its variance is large. Standard deviation is about 1.15 bets per hand once doubles and splits are included, so over 100 hands at €10 your expected result is −€4 with a one-standard-deviation range of roughly ±€115. Losing streaks of eight or more hands happen about once every 300 hands. A session bankroll of 50 units and a stop-loss you actually obey are the minimum; the bankroll management guide covers unit sizing, session limits and how the maths changes when you add a bet spread.

Drill the chart until it's automatic

The free CasinoTrust blackjack trainer deals random hands against random up-cards and corrects every decision against this chart. Most players are close to error-free within a week.

Open the strategy trainer

Frequently asked questions

What is basic strategy in blackjack?

The mathematically optimal decision — hit, stand, double, split or surrender — for every player hand against every dealer up-card, assuming a freshly shuffled shoe. It is derived by calculating the expected value of each option and has been confirmed by simulation. Following it at a 6-deck S17 DAS 3:2 table reduces the house edge to about 0.4%.

Does basic strategy guarantee you win?

No. It minimises the house edge; it does not reverse it. At a good table you still lose about €0.40 per €100 wagered in the long run, with large swings in either direction over any single session. It is the difference between losing 0.4% and losing 2–5%, which over a year of play is substantial.

Should you hit 16 against a dealer 10?

Yes, unless surrender is available, in which case surrender. Hitting hard 16 busts 62% of the time, but standing wins only when the dealer busts, about 21% from a 10. Hitting loses slightly less on average. If your 16 is made of three or more cards, standing becomes marginally correct in some shoe compositions, but hit is right for the two-card hand.

Is card counting illegal?

No. Counting is using your memory; no law in any major jurisdiction prohibits it. Casinos are private businesses and may ask counters to leave, restrict their bets or ban them, and online operators can close accounts under their terms. Using a device or app to count is a criminal offence in many places.

Can you use basic strategy in live dealer blackjack?

Yes, and you should; the chart applies exactly. Check the rules panel first: most live tables use eight decks, and some hit soft 17 or pay 6:5 on certain tables. The H17 adjustments are listed on this page. You can also keep a chart open in another window — nothing prevents it online.

What is the worst hand in blackjack?

Hard 16 against a dealer 10. Every option loses more than half the bet on average: standing loses about 54%, hitting about 54% as well, with hitting marginally ahead. It is the reason late surrender exists and why 8,8 — a hard 16 — is always split, even against a 10.