History
Blackjack evolved from European games like vingt‑et‑un. The modern mathematical approach began in the 1960s with the publication of Beat the Dealer, which popularized basic strategy and card counting. Since then, casinos and players have engaged in a rules‑and‑strategy arms race: more decks and hit‑soft‑17 rules versus better strategy charts, bet‑spreads, and game selection by players.
Rules (Standard)
- Goal: beat the dealer without busting (exceeding 21).
- Blackjack (A + 10‑value) beats any non‑blackjack 21.
- Actions: hit, stand, double (one card), split pairs, surrender (where offered).
- Dealer: draws to 16 and stands on 17 (some games have H17 – dealer hits soft 17).
- Insurance: side bet vs dealer blackjack (generally negative EV for players).
Basic Strategy (Quick Keys)
- Never take insurance / even money in standard rules.
- Always split A,A and 8,8; never split 5,5 or 10,10.
- Double 11 vs 2–10; double 10 vs 2–9; double soft A,6 vs 3–6 (S17 tables).
- Hit hard 12 vs dealer 2–3; stand vs 4–6; otherwise follow chart.
- Use late surrender on hard 16 vs 9–A (where allowed), except 8,8 is split.
For perfect play, use a printed/basic‑strategy chart matched to your exact rules (decks, S17/H17, DAS, surrender).
Basic Strategy Highlights (Selected)
| Situation | Recommended Action | Notes |
|---|---|---|
| Hard 11 vs 2–10 | Double | Hit vs Ace in some rule sets; check chart. |
| Hard 10 vs 2–9 | Double | Otherwise Hit. |
| Soft A,6 vs 3–6 (S17) | Double | Otherwise Hit. |
| Pairs A,A or 8,8 | Split | Never split 5,5 or 10,10. |
| Hard 12 vs 2–3 | Hit | Stand vs 4–6. |
Exact choices depend on decks and table rules (S17/H17, DAS, surrender). Use a chart tailored to your game.
Odds & House Edge (by Rule)
Effects below are the change in player expected return against our base game (3:2, S17, DAS, no surrender), each computed by re-running the engine with one rule changed. The two deck rows are quoted from Wizard of Odds against an eight-deck base, because the engine models an infinite deck and has no deck count. Positive = better for player (lower house edge). Negative = worse.
| Rule change | Effect on player return | Notes |
|---|---|---|
| Blackjack pays 6:5 (vs 3:2) | −1.35% | Major penalty; avoid these tables. |
| Dealer hits soft 17 (H17) instead of S17 | −0.22% | Common hidden edge increase. |
| Late surrender (LS) allowed | +0.09% | Small but real improvement. |
| Single deck (vs 8 decks, quoted) | +0.48% | Fewer decks help players (all else equal). |
| Double deck (vs 8 decks, quoted) | +0.19% | Moderate improvement. |
| European no‑hole‑card (ENHC) | −0.11% | Dealer doesn’t peek; doubles/splits risk a dealer blackjack. |
| DAS not allowed (vs DAS) | −0.14% | Loss of flexibility after splits hurts. |
Benchmark, computed here rather than quoted: under an infinite-deck model (independent draws), 3:2, S17, DAS, no surrender, our engine returns 0.5117% against perfect basic strategy, and that is the figure in the dataset. Published analyses of Atlantic City rules land nearer 0.43%. Most of the gap is two things: Atlantic City deals late surrender, which our base game leaves out (worth 0.09 points here), and an eight-deck shoe, which the infinite-deck model prices slightly worse for the player. Where our number and someone else's differ, this page states ours and says why.
Bankroll & Variance
- Expect swings: even with a 0.5% house edge, short‑term variance dominates.
- Set session stop‑loss / stop‑win points; avoid chasing losses.
- Flat bet for simplicity; if you ramp bets, do so modestly and within budget.
Advanced Topics (Overview)
- Index Deviations: small strategy changes by true count (e.g., stand 16 vs 10 at high counts).
- Side Counts & Penetration: deeper shoes + accurate counts improve EV but require skill and discipline.
- Game Selection: picking 3:2, S17, DAS, LS tables typically beats any minor counting edge at bad 6:5 games.
Famous Figures
- Edward O. Thorp, father of card counting and author of Beat the Dealer.
- Ken Uston; popularized team play and legal battles on player rights.
- MIT Blackjack Team; organized advantage play; subject of books/films.
- Don Schlesinger: risk of ruin, Illustrious 18, bankroll theory.
FAQ
What’s the best quick filter for a good table?
Look for 3:2 payout on the felt, S17, DAS, and surrender; avoid 6:5 tables.
Is insurance ever good?
Generally no for non‑counters; it has negative EV in standard rules.
How much does H17 hurt?
It’s roughly a −0.22% hit to player return compared with S17.
On counting. The Hi-Lo values every guide asks you to memorise are derived rather than chosen: removing each rank and repricing the game produces them, and it shows the five to be worth more than any other card, then the four and the ace.
Corrected on 17 October 2026: the engine now values a resplit as two hands, so the standard game is 0.5117% (was 0.5153%), and the card-counting effects are priced from the depleted deck (the ace is third, not first). The benchmark named six decks; the engine is infinite-deck. Late surrender is worth 0.09 points, not 0.08, and the deck rows are labelled as quoted.