True count and bet spreads

The count tells you when the game is in your favour. The bet ramp determines how much you gain from that information.

A card counter who raises and lowers bets randomly has done nothing except count cards for entertainment. The edge comes from translating the true count into appropriately sized bets: small bets when the house has the advantage (negative or neutral count), large bets when the player has the advantage (positive count). The design of the bet ramp, how steep it is, how it tracks the count, and how it stays within the boundaries of what a London pit will tolerate, is the critical skill that sits between the count and the money. At the Hippodrome on Leicester Square, for instance, the floor runs at 60-80 hands per hour depending on occupancy, which gives you a realistic sense of how frequently the count will move into actionable territory in a two-hour session.

Running Count to True Count: The Essential Conversion

The running count accumulates as cards are dealt. In a six-deck shoe, it can swing from roughly -30 to +30 by the time the cut card is reached at 75% penetration. The raw running count, however, is not directly comparable across different stages of the shoe. A running count of +12 with four decks remaining means something very different from a running count of +12 with one deck remaining. The first represents a mild imbalance spread across a large pool of remaining cards; the second represents a significant imbalance in a small, fast-approaching pool.

The true count normalises this. Divide the running count by the estimated number of decks remaining. If the running count is +12 and you estimate two decks remain, the true count is +6. That +6 has a consistent interpretation regardless of where you are in the shoe: it represents approximately six extra ten-value cards per deck compared to neutral composition. In Hi-Lo, each unit of true count above neutral adds approximately 0.50% to player edge. A TC of +2 against a 0.50% house edge roughly produces an even game; TC +3 gives the player approximately 0.50% edge; TC +5 gives approximately 2.0% edge.

Estimating decks remaining accurately is a trainable physical skill. You need to be able to look at the discard tray and estimate its depth to the nearest half-deck. At 75% penetration in a six-deck shoe, the discard tray holds approximately 4.5 decks' worth of cards when the cut card appears. Practising with a physical shoe and discard tray builds the visual calibration. The card counter trainer also tests deck estimation accuracy as a standalone drill.

The Bet Ramp: Structure and Rationale

A bet ramp is a pre-committed schedule of bet sizes indexed to true count levels. You decide the ramp before you sit down; you don't improvise it at the table. The structure of a reasonable UK six-deck ramp might look like this: bet 1 unit at TC +1 and below, 2 units at TC +2, 4 units at TC +3, 6 units at TC +4, and 8-12 units at TC +5 and above. Slight variations exist, but the principle is consistent: the bet size increases with count, and the ratio of maximum to minimum bet is the spread.

The spread is the ratio of your maximum bet to your minimum bet. A 1-to-12 spread means your largest bet is twelve times your smallest. This is the critical variable because it determines how much edge you capture from the count's positive fluctuations. A 1-to-4 spread barely scratches the available edge; a 1-to-24 spread will produce immediate pit attention. In UK casino conditions, a spread of 1-to-8 is moderate, 1-to-12 is the upper boundary of what many experienced counters consider viable over multiple sessions at the same venue, and anything beyond that invites scrutiny on the first session.

The ramp should be smooth, not binary. A counter who bets exactly 1 unit for 20 hands and then suddenly jumps to 12 units the moment the count peaks is following a script that the pit can reconstruct after the fact. A ramp that introduces gradual increases through TC +2 and +3 looks more like natural bet variation by a player who is reacting to their perceived momentum.

Kelly Criterion and Half-Kelly in Practice

The Kelly criterion is a formula for optimal bet sizing: it says bet a fraction of your bankroll equal to your edge divided by the variance of the game. For blackjack, the variance is approximately 1.33 (standard deviation per hand is roughly 1.15 units, variance is the square), so the Kelly fraction for a 1% edge is approximately 1% / 1.33 = 0.75% of bankroll.

At a 1% edge and a £10,000 bankroll, full Kelly says bet approximately £75 per hand. That's the theoretically optimal single-hand bet size for maximum long-run growth. But full Kelly comes with a known practical problem: it maximises geometric growth at the cost of wild short-term volatility. A losing sequence at full Kelly can produce drawdowns of 50% or more of your bankroll, which is psychologically difficult to sustain and operationally awkward if the drawdown happens to coincide with the sessions where you've established yourself at a given venue.

Blackjack Forum Online's analysis of the Kelly criterion recommends half-Kelly as the practical standard for blackjack counters. Half-Kelly means betting half the theoretically optimal fraction: at 1% edge and £10,000 bankroll, your bet is approximately £37.50. Half-Kelly retains approximately 75% of the long-run growth rate of full Kelly while halving the variance. This is the right trade-off for most players: slightly slower bankroll growth in exchange for significantly smoother variance and a longer operational lifespan at any given venue.

The bankroll calculator at tools/bankroll-calculator/ lets you input your unit size, spread, edge estimate, and session length to model expected outcomes and risk of ruin under both full and half-Kelly assumptions. Use it before committing to a bet structure.

Key numbers

Sources: our Hi-Lo analysis system, Blackjack Forum Online Kelly criterion, spreads and bankroll lesson.