Why the chart looks the way it does

What basic strategy actually is

Basic strategy is not a feel for the game or a read on its rhythm. For every possible combination of your hand and the dealer's upcard, the expected value of each available play has been computed exhaustively, and the chart collects the plays with the highest value. It tells you the play that loses least on average when the same situation repeats.

There is one assumption built into it. Basic strategy knows only three cards: your two and the dealer's upcard. It does not track what has already been dealt and cannot predict what comes next. For anyone not counting cards, every departure from this chart lowers expected value.

Where it all starts: dealer bust probability

The dealer makes no decisions. Draw on 16 or less, stop on 17 or more. Because the procedure is fixed, the upcard alone is enough to work out how often the dealer breaks.

There is a cliff between six and seven — 42% drops to 26% in one step. That is why the chart splits so sharply into a 2–6 block and a 7–A block when you read it across. Against a 2 through 6 the dealer breaks roughly four times in ten; from 7 upward the dealer reaches 17 or better about three times in four.

Why you hit hard 12 against a 2 but stand against a 4

This is the cell people find most confusing. Hitting 12 busts you on any ten-value card — 16 of 52, about 31%. A dealer 2 busts about 35% of the time. At a glance, waiting looks better.

The trap is what happens when the dealer does not bust. Standing on 12 loses to any dealer total from 17 to 21. In other words 12 is a hand that only wins when the dealer breaks. A dealer 2 may be weak, but it survives 65% of the time. Giving up all of that is worse than accepting a 31% bust risk to build a hand that can win on its own.

Once the dealer shows 4, 5 or 6, bust probability passes 40% and the arithmetic flips: waiting now wins. That is why 12 reads S against 4, 5 and 6 only, and H against 2 and 3. From 13 upward, hitting busts you more often, so standing begins at 2.

Why soft hands play so aggressively

A soft hand cannot bust on the next card. Ace plus six is soft 17; draw a ten and you get 17, not 27, because the ace drops to one. It is a free attempt with nothing at stake.

So soft hands keep hitting far longer than hard hands, and against a weak dealer they double instead. Soft 18 is the clearest case. Eighteen sounds respectable, but against a dealer 9, 10 or ace the correct play is H — 18 rarely beats those upcards, and being soft means there is nothing to lose by improving it.

Splitting: three things to remember

Always split aces

Two aces together make soft 12, which is nearly useless. Split them and you have two hands starting from 11. Ten-value cards are the most common card in the deck, so each has a strong chance of reaching 21. It is the most profitable split in the game.

Always split eights — even against a ten

Kept together they make hard 16, the worst hand in blackjack: too low to stand on, too likely to bust if you draw. Starting from a single eight is a far better position, and you get two of them.

Splitting eights against a dealer 10 still loses money. It just loses less than defending 16 does. No cell illustrates better that basic strategy picks the play that loses least, not the play that wins.

Never split fives or tens

Two fives make hard 10. One ten-value card turns that into 20, and in most spots it is a double-down. There is no reason to break it into two weak hands starting from five. This is why the calculator does not treat a pair of fives as a pair at all and sends it to hard 10 instead.

Two tens make 20, which is already close to the best hand available. Splitting trades one strong hand for two ordinary ones.

Surrender is the skill of losing

Surrender runs hardest against intuition, because you hand over half your bet without playing the hand out.

The test is simple. If playing the hand to the end has an expected loss of more than half your bet, surrendering is the better outcome. Hard 16 against a dealer 10 is the classic case: hitting and standing both lose more than half the time, so taking a certain half-loss is cheaper.

Only a handful of cells qualify. At an S17 table it is 16 against 9, 10 and ace, plus 15 against 10; under H17, 15 against ace and 17 against ace are added. If your table does not offer surrender, fall back to hitting. Turn surrender off in the calculator's rule toggles and the chart recomputes itself.

Common misconceptions

"Someone else played badly and ruined my cards"

Another player's choices do not change your expected value over time. The card they took might have helped you or hurt you, and the two possibilities cancel exactly.

"Assume the dealer's hole card is a ten"

A widespread simplification, but not accurate. Ten-value cards are 16 of 52, roughly 31%. Basic strategy already weighs every possibility by its true probability.

"I have lost several in a row, so I am due"

Cards have no memory of the last hand. In a shuffled shoe the distribution of the next card is unaffected by what came before.

"Doubling my bet after each loss will get it back"

Martingale-style betting systems cannot change the house edge. Table limits and the size of your bankroll simply make the eventual loss larger.

"Basic strategy means I will win"

No. It reduces the house edge to roughly 0.5% and 0.5% is still negative. It is the slowest way to lose, not a way to win.

Situation by situation

The spots that come up most often, each worked through on its own page.