How Wong Halves Works
Wong Halves is a balanced, level-three counting system developed by Stanford Wong. Where simpler systems sort every card into three buckets, Wong Halves uses six distinct values — including fractions — because the cards genuinely differ in how much they matter.
The reasoning comes from effect of removal: how much the player's edge shifts when one specific card leaves the shoe. Pulling a five out of the deck helps you noticeably more than pulling a two, because fives are the card that most reliably turns a dealer's stiff hand into a made one. Hi-Lo tags both as +1 and accepts the approximation. Wong Halves doesn't.
| Card | Value | Why |
|---|---|---|
| 2 | +½ | Mildly helpful to remove — small effect |
| 3, 4, 6 | +1 | Standard low cards, strong dealer-safe draws |
| 5 | +1½ | The single most valuable card to remove |
| 7 | +½ | Borderline — helps the dealer less than a six |
| 8 | 0 | Genuinely neutral |
| 9 | −½ | Slightly favours the player when it remains |
| 10, J, Q, K, A | −1 | Blackjacks, strong doubles, dealer busts |
The system is balanced, meaning a full shoe counted end to end returns to exactly zero. That's what makes true-count conversion valid. It also gives you a free error check: if you count down a full deck and don't land on zero, you made a mistake.
Running Count vs. True Count
The running count is the raw sum of every card you've seen. The true count is the running count divided by the decks remaining — and it's the only one that tells you anything about your actual edge.
Here's why the division matters. A running count of +6 with six decks left means those extra high cards are spread thin across 312 cards. The same +6 with one deck left means they're concentrated in 52. Identical running count, completely different situations:
| Running count | Decks left | True count | Your position |
|---|---|---|---|
| +6 | 6.0 | +1.0 | Roughly break-even |
| +6 | 3.0 | +2.0 | Small edge — start ramping |
| +6 | 1.5 | +4.0 | Strong edge — max bet |
In an eight-deck shoe, each true count point is worth roughly half a percent of edge. Since a typical shoe game starts you around half a percent behind, you need a true count near +1 just to reach break-even — which is why the neutral zone is not actually neutral, and why most of a counter's hands are played at a small disadvantage.
True-count conversion is also where most people's accuracy collapses. Tags become automatic within a few hours of practice; dividing a fractional running count by a shrinking, estimated remainder — while the dealer is waiting — does not. That's what the counter above is built to drill.
Why the Strategy Solver Recalculates
Almost every basic strategy chart online is an image, drawn for one specific ruleset, usually unlabelled. Play that chart at a different table and some of its cells are simply wrong. The solver on this page doesn't look anything up — it works out the expected value of hitting, standing, doubling and splitting for the rules you select, and colours the grid with whichever wins.
The rules that actually change your play:
- Soft 17. Whether the dealer hits or stands on soft 17 changes several cells and costs you about 0.22% when they hit. Hard 11 against an ace is the clearest tell: double against a hitting dealer, hit against a standing one.
- Double after split. Without it, several marginal splits stop paying — 4s, 6s against a 2, and 2s and 3s against low upcards — because they were only profitable for the doubling chances they created.
- Deck count. Single deck is a meaningfully different game, not just a better-paying one. Fewer cards means each removal matters more, and the chart shifts.
- Surrender. Adds a fifth option and reshapes hard 15 and 16 against high upcards.
- Charlie rules. A six-card automatic win is worth about 0.16% and changes how you play multi-card stiffs — with five cards in hand, hit anything through hard 16, and always hit soft 19 or 20 since you cannot bust.
Common Questions
Is card counting illegal?
No. Counting cards in your head is legal everywhere — you're using publicly visible information and thinking about it. What is illegal in Nevada and most gaming states is using a device to help you do it. Casinos are private businesses and may still ask you to leave, but that's a house decision, not a legal one. This site is a practice tool, not a live assistant, and shouldn't be used at a table.
Is Wong Halves worth the extra effort over Hi-Lo?
In shoe games, honestly, not much. Wong Halves has a betting correlation near 0.99 against Hi-Lo's 0.97, which translates to a few percent more win rate — not a few tenths of a point more edge. The gain is larger in single- and double-deck games where playing accuracy matters more. Choose it because you'll count it accurately, not because the numbers look better on paper.
What actually determines whether a game is beatable?
Penetration first, then bet spread, then everything else. How deep the dealer cuts determines how often the count reaches the range where your big bets live. A 1-to-10 spread at 80% penetration beats a 1-to-20 spread at 50%. And the ratio that matters is your top bet divided by your neutral bet — not your minimum — because that's where most of your hands are actually played.
How long before an edge shows up?
Longer than almost anyone expects. Blackjack's per-hand standard deviation is very large relative to the edge, so a modest advantage takes hundreds of thousands of hands to separate from noise. A real edge and a losing month are entirely compatible. Bankroll and patience matter more than counting speed.