True Count Blackjack: The Conversion That Makes Counting Work

The running count tells you how many extra high cards are left in the shoe. It does not tell you how concentrated they are — and concentration is what actually moves your edge. Converting the running count to a true count is the step that turns a Hi-Lo count into real betting and playing decisions.

Why the running count alone misleads in shoe games

Suppose your running count is +6. Is that good? It depends entirely on how much of the shoe is left.

With five decks remaining, that surplus of six high cards is diluted across roughly 260 cards. The next card is barely more likely to be a ten than normal. With one deck remaining, the same +6 is packed into about 52 cards — a genuinely ten-and-ace-rich deck where blackjacks, strong doubles, and dealer busts all become more likely.

Same running count, wildly different situations. In a single-deck game the running count is close to the true count, which is why old-school single-deck counters could nearly skip the conversion. In a six- or eight-deck shoe, betting off the raw running count means systematically overbetting early in the shoe, when the count is diluted. The true count fixes this by measuring surplus per remaining deck instead of total surplus.

The conversion: divide by decks remaining

True count = running count ÷ decks remaining. Remaining — not dealt. Worked examples:

• Running count +9 with 4.5 decks left: 9 ÷ 4.5 = +2.
• Running count +7 with 3 decks left: 7 ÷ 3 = +2.33, which floors to +2.
• Running count +4 with 1 deck left: 4 ÷ 1 = +4. Late in the shoe, small running counts carry big weight.
• Running count −6 with 3 decks left: −6 ÷ 3 = −2. Negative counts convert the same way, and they matter — they tell you to bet the minimum or sit out.

Notice the third example: as the shoe depletes, the divisor shrinks, so the same running count produces a bigger true count. This is why deep penetration — how far the dealer deals before shuffling — is so valuable to counters. The high-true-count situations live at the back of the shoe, which is exactly why casinos cut penetration as a countermeasure.

Reading the discard tray

Nobody tells you how many decks remain. You estimate it from the discard tray — the clear plastic stack of already-played cards beside the dealer.

One deck is 52 cards, a stack of predictable height. At a six-deck game, if the discard tray holds about two decks, roughly four decks remain. Ignore the handful of cards currently on the table; they wash out. Estimating to the nearest half deck is plenty of precision for betting decisions.

Train this at home: stack one, two, three decks and study the heights until you can call any pile to within half a deck at a glance. It takes an evening, and it is the difference between a real true count and a guess.

Floor or round? Pick one and stick to it

7 ÷ 3 = 2.33. Is that TC +2 or +2.33? Bet ramps and deviation indexes work on whole numbers, so you either floor — always drop toward negative infinity, so +2.9 becomes +2 and −1.3 becomes −2 — or round to the nearest integer.

Both work if applied consistently. Flooring is slightly conservative on the bet side: you never bet a level your true count has not fully earned. That is the convention our free trainer uses for its bet ramp. What hurts you is switching methods mid-session, because your bets and deviations stop matching any coherent system.

What a true count point is worth

Each true count point shifts the edge by roughly +0.5% toward the player. Basic strategy alone plays at about −0.5% against the house, depending on rules — if your basic strategy is not automatic yet, drill it on the interactive strategy chart before worrying about conversions. That means the game is close to break-even around TC +1, and by +2 you hold roughly a half-percent edge over the casino.

Across full sessions on a good 3:2 shoe game with a proper bet spread, a competent counter runs roughly +0.5% to +1.5% over the house. One giant caveat: this assumes 3:2 blackjack payouts. A 6:5 payout adds about 1.4% to the house edge and makes counting effectively worthless. Never play 6:5 — walk.

Betting the true count: the ramp

Since your edge scales with the true count, your bets should too — Kelly-style sizing means betting in proportion to your edge. Our trainer's default ramp, using the floored true count:

The ramp is where the money is made: minimum bets while the house has the edge, real bets out when you do. Sizing the unit itself is a bankroll question — serious counters keep 200–400 max bets behind them, which the bankroll guide covers in detail.

Insurance at +3 and playing deviations

The true count also changes how you play hands. The most valuable deviation by far is insurance. It is a side bet that the dealer's hole card is a ten, and off the top of the shoe it is a losing bet. At TC +3 or higher (Hi-Lo), the shoe is ten-rich enough that insurance becomes profitable — take it every time.

Beyond insurance, the Illustrious 18 is the standard set of index plays; the numbers below come from the common six-deck H17 set. A few examples: stand on 16 vs 10 at TC 0 or higher; stand on 15 vs 10 at +4; stand on 12 vs 3 at +2; double 11 vs ace at +1. Some indexes are negative — you keep standing 13 vs 2 down to −1 and hit below it. Each deviation fires only when the count says the shoe's composition justifies it; otherwise you play straight basic strategy.

Putting it together

Keep the running count, glance at the discard tray, divide, floor. Bet from the ramp, take insurance at +3, apply your indexes. That whole loop has to run in a few seconds per round while you chat and stack chips — and none of it is illegal when it happens in your head, though casinos can still back you off; see the legality guide for the details.

The place to build that speed is the free trainer, which shows live EV as you play and runs million-hand simulations demonstrating that the ramp bettor beats the flat bettor long-run. Drill the conversion until dividing by 3.5 feels boring.

Educational content only. Practice with fictional chips on our free trainer — no real money, no casino affiliation. 18+.

Frequently asked questions

How do you calculate the true count in blackjack?

Divide the running count by the number of decks remaining in the shoe — remaining, not dealt. For example, a running count of +9 with 4.5 decks left gives a true count of +2. Estimate decks remaining from the discard tray to the nearest half deck.

Should I floor or round the true count?

Either works, as long as you apply one method consistently to both your bets and your deviation indexes. Flooring — always dropping toward negative infinity, so +2.7 becomes +2 — is slightly conservative on bet sizing. Our free trainer floors the true count for its bet ramp.

How much edge is one true count point worth?

Each true count point shifts the edge roughly +0.5% toward the player. Basic strategy alone plays at about −0.5% depending on rules, so the game turns approximately break-even near TC +1 and gives the player about a half-percent edge around +2. On a good 3:2 shoe game with a proper spread, a counter's overall edge runs roughly +0.5% to +1.5%.

When is insurance a good bet in blackjack?

Using Hi-Lo, insurance becomes profitable at a true count of +3 or higher, because the remaining shoe is rich enough in tens. At any lower count it stays a losing side bet and should be declined. It is the single most valuable count-based deviation.

Does the true count matter at 6:5 blackjack tables?

No — a 6:5 blackjack payout adds about 1.4% to the house edge, which overwhelms the roughly 0.5% per true count point a counter gains. That makes counting effectively worthless at 6:5 tables. Only play games that pay 3:2.