Hard 12 Against a Dealer 4
Updated
Stand — by 0.4% of a bet, which makes this the tightest non-surrender decision in basic strategy. It is also the cell that published charts get wrong more than any other, and we know that because we had it wrong ourselves.
- Stand. 20.9% lost, against 21.3% for hitting.
- The margin is 0.4% — the narrowest on the chart outside surrender.
- A dealer 4 busts 39.6% of the time, and that is finally enough.
- Because it is this tight, the composition of your 12 can flip it.
We had this one wrong, and it is worth explaining
Our own strategy tables gave hit for this cell in 48 of 80 rule combinations. The generator that produced them had been checked against a published chart for eight decks only, so the error survived everywhere else.
We found it while building this site, verified the correction against an independently written solver, and fixed all 48. We are describing it here rather than quietly patching it because a site whose whole claim is “our numbers are computed, not copied” owes you the failures as well as the successes.
It also explains why this cell is worth its own page. When a decision is decided by 0.4%, a solver that is slightly off produces the wrong letter, and nobody notices.
What each option is actually worth
Six decks, dealer stands on soft 17, averaged across every two-card 12. Bet of 1 unit:
Your own cards can flip it
A margin this thin is smaller than the effect of the cards in your hand. A 12 made of 10+2 has already taken a ten out of the shoe, so your next card is slightly less likely to bust you and hitting gains ground. A 12 made of 7+5 leaves every ten in place and standing is clearer.
Basic strategy ignores this on purpose: it gives one answer per total so it can be memorized. If you want the composition-aware answer for your exact cards, that is what the calculator is for.
Keep reading
By Martin Kosek · every number on this site is computed by us, not copied.
← Back to guides