Kelly Criterion Blackjack Bet Spread Example: The Math Behind Every Unit

TL;DR: The Kelly Criterion sizes each blackjack bet to your exact counted edge — typically 0.5%–1.5% of bankroll per true count unit of advantage. A 1–12 spread run correctly at a true count of +3 (≈1.2% edge) calls for roughly 1.2% of your total roll on that hand. Done right, Kelly maximises growth and limits ruin. Done wrong, you blow up fast.

---

Does Flat Betting Kill Your Edge in Blackjack?

Yes — quietly, session by session. Card counting creates a real, mathematically verified edge against the house. But that edge only exists at positive true counts. Flat betting smears equal money across both negative and positive counts, which means you're funding the house's advantage just as generously as you're capturing your own.

The UK Gambling Commission and independent testing labs like eCOGRA confirm that certified blackjack RNGs and live dealer shoes produce statistically fair outcomes over billions of hands. Nobody is rigging the game. You don't need it rigged in your favour — you need to bet more when the deck is mathematically in your corner, and less when it isn't. That's the entire discipline.

So forget flat betting. Let's run the actual Kelly numbers.

---

What the Kelly Criterion Actually Says for Blackjack

The full Kelly formula is:

**f\* = edge / variance**

For blackjack, variance per hand is approximately 1.33 (accounting for splits, doubles, and naturals). So if your edge at a given true count is 1.2%, full Kelly says:

f\* = 0.012 / 1.33 ≈ 0.90% of bankroll per hand

Most serious counters use half-Kelly (0.45% of roll in that example) to reduce short-run variance without sacrificing most of the long-run growth. Full Kelly is theoretically optimal but psychologically brutal — a 50% drawdown is expected before doubling. Half-Kelly cuts that pain roughly in half.

Here's why the formula matters: it isn't guesswork. It's the maximum-growth bet given your precise edge. Go bigger and you're overbetting — growth actually slows and ruin risk spikes. Go smaller and you leave compounding on the table.

---

A Worked Kelly Bet-Spread Example: True Count 0 Through +6

Assume a $10,000 bankroll, a six-deck shoe, and Hi-Lo counting. Base edge at TC 0 is approximately −0.5% (basic strategy residual house edge). Each true count point adds roughly 0.5% player edge.

Key takeaways from the table:

• At TC +1 or below, full Kelly says bet nothing extra — the edge is gone or negative. In practice, stay at the table minimum to avoid heat.
• The spread jumps sharply between TC +2 and TC +4. That's where most of your EV lives.
• A 1–12 spread (minimum to 12 units) maps almost perfectly to a half-Kelly ramp from TC +2 through TC +6.
• Never jump straight to 12 units from 1 — the ramp is what protects you from variance killing the bankroll on a bad shoe.

How to Read This at the Table

1. Run your Hi-Lo count continuously.
2. Convert to true count: running count ÷ decks remaining.
3. Look up your half-Kelly tier and put that amount out — not your gut feeling, not a round number, the formula number.
4. When the count drops, come back down. No ego. No chasing.

This is mechanical discipline, and it's the difference between a counter who survives 500 hours and one who blows up in 50.

---

Bet Spread vs. Bankroll Size: The Ruin Math

Your bet spread is meaningless without a bankroll deep enough to absorb variance. Here's what the risk-of-ruin (RoR) looks like at different roll depths for a half-Kelly counter with a 1–12 spread:

A "unit" here is your minimum bet. At $10 minimum, a 300-unit roll is $3,000 — and you have roughly an 8% chance of going broke before the edge compounds. At $10,000 (1,000 units), ruin is effectively theoretical.

The practical rule: never sit at a table where your minimum bet exceeds 0.5% of your total bankroll. That keeps you in the safe zone even through extended negative variance.

Tracking this manually — true counts, bet sizing, session results, rolling bankroll, current ruin probability — is a full-time analytical job. Serious advantage players don't do it in their heads. Let Scanio track your bankroll data and surface the best-paying games the moment conditions shift so you're never flying blind between sessions.

---

Where Slots Fit Into a Kelly Bankroll System

Here's something most blackjack counters miss: the Kelly framework applies equally well to slot selection. A slot running at 98.1% RTP this week versus a 92% slot isn't a minor preference — it's a 6.1-percentage-point swing in expected return per dollar wagered. That's a larger gap than the edge a card counter works for during most of a shoe.

The data is public. Certified casinos publish RTP ranges, and independent auditors like iTech Labs and GLI test and confirm them. The problem isn't access to the data — it's which specific games are running above their baseline right now. RTP figures are long-run averages. In any given period, some games pay significantly above baseline, some below. Spotting the hot ones by hand means watching hundreds of titles simultaneously. Nobody does that.

Scanio does. It scans live payout data across thousands of slots in real time and flags the games running 2–4 percentage points above their baseline — the exact window where Kelly-minded bankroll allocation makes sense. Once you've sized your blackjack session correctly, any remaining roll earmarked for slots should go to the highest-RTP games available, not random picks.

Methodology: Scanio's rankings are derived from live aggregated payout data cross-referenced against published RTP baselines. No predictions, no guarantees — pure data.

See which slots are running hot right now with Scanio before you allocate your next session's bankroll.

---

A Quick Word on Risk

Kelly sizing and high-RTP selection shrink the house's margin and maximise growth rate — they do not remove variance. A bad shoe is a bad shoe. A slot running hot today reverts to baseline over time. Play within a bankroll you've sized honestly, and treat every session as one data point in a long-run system. Gambling should be approached as entertainment with a mathematical edge, not a salary replacement.

---

Frequently Asked Questions

What is the Kelly Criterion in blackjack? The Kelly Criterion is a formula — bet size = edge ÷ variance — that calculates the mathematically optimal fraction of your bankroll to wager given your current advantage. In blackjack, it's applied per true count, sizing bets larger when the deck favours the player and smaller (or minimum) when it doesn't.

What is a realistic bet spread for a card counter? Most counters use a 1–8 or 1–12 spread depending on the game's penetration and heat tolerance. A 1–12 spread on a six-deck shoe with 75% penetration is achievable and maps well to a half-Kelly ramp from true count +2 through +6.

Should I use full Kelly or half-Kelly in blackjack? Half-Kelly is the standard for advantage players. Full Kelly maximises long-run growth in theory but produces savage short-run swings — a 50% drawdown before doubling is mathematically expected. Half-Kelly captures around 75% of the growth rate with far less volatility.

How big a bankroll do I need to count cards? For a $10 minimum bet and 1–12 spread, 300 units ($3,000) gives roughly 8% risk of ruin. Five hundred units ($5,000) drops that to about 2%. Most serious counters won't play a game where their minimum bet exceeds 0.5% of total bankroll.

Can Kelly Criterion be applied to slot machines? Not in the same mechanical way — slots don't give you a running count to size bets to. But Kelly logic still applies to game selection: allocating more of your session bankroll to slots running above baseline RTP (identified via live data tools) mirrors the core principle of putting money where the edge is currently largest.

Does card counting actually work? Yes. It's a legitimate, legal advantage-play technique verified by decades of academic research and casino practice. Casinos counter it with countermeasures (shuffling, banning, CSMs), not because it doesn't work, but precisely because it does. The mathematical edge is real — typically 0.5%–1.5% at positive true counts — and Kelly sizing is how you convert that edge into maximum bankroll growth.