There are dozens of card counting systems, but four cover 95% of what anyone will ever use: Hi-Lo, KO, Hi-Opt I, and Hi-Opt II. They differ in two main ways: how hard they are to use accurately, and how much betting-correlation edge they deliver. More complex systems give you more edge — but only if you can count them without error, which most people can’t. For nearly everyone, the right answer is Hi-Lo.

Hi-Lo (balanced, level 1)

By far the most popular system. Every card is worth +1, 0, or -1. At the end of the shoe, the count returns to zero, which is why it’s called “balanced.”

Who it’s for: Everyone starting out, and most people long-term. Hi-Lo captures about 97% of the theoretical edge of more complex systems while being dramatically easier to count accurately at speed.

Trade-off: You have to convert running count to true count by dividing by decks remaining. This isn’t hard, but it does mean you need a rough sense of how much of the shoe is left.

KO (Knock-Out, unbalanced, level 1)

Same values as Hi-Lo for everything except the 7, which counts +1 instead of 0. This unbalances the system on purpose — you never need to convert to a true count because the unbalanced structure does the conversion implicitly.

Who it’s for: People who find the running-to-true conversion mentally taxing. You give up a tiny amount of theoretical edge in exchange for a simpler decision process.

Trade-off: You have to remember that the count does not start at zero and does not end at zero — the “pivot point” depends on the number of decks. It’s simpler in one dimension and more confusing in another.

Hi-Opt I (balanced, level 1)

A slightly different card-value assignment that ignores the 2 and the Ace. It’s a “level 1” system (all values are +1, 0, or -1) but tweaked for better playing-strategy correlation.

Who it’s for: Counters who care more about accurate playing deviations than aggressive betting. You’ll likely need a separate ace side-count to keep betting correlation competitive, which adds complexity.

Trade-off: Better playing correlation, worse betting correlation than Hi-Lo unless you side-count aces.

Hi-Opt II (balanced, level 2)

A “level 2” system — values include +1, +2, 0, -1, -2. More information per card, better betting and playing correlation than any level-1 system.

Who it’s for: Advanced counters who can execute a level-2 count at full casino speed with zero errors and who are willing to add an ace side-count.

Trade-off: Dramatically more error-prone than level-1 systems. The theoretical edge gain is small — usually under 0.1% — and one miscounted hand per shoe wipes it out. Most players should not use this system.

Which one should you actually pick?

Pick Hi-Lo. That’s not a hedge — it’s genuinely the right answer for almost everyone. The 0.1–0.2% edge you might gain from a more complex system is only real if you count it without error, and error-free execution at casino speed is much harder than people realize. The counters who actually make money long-term tend to use Hi-Lo (or something similar) and compensate for any lost edge with better game selection, better bet spreads, and more hours of play.

Our trainer supports all four systems, so when you’re ready to experiment you can practice them head-to-head. But start with Hi-Lo, and don’t switch until you can execute it perfectly.

Honorable mentions

Red Seven: Like KO, but with a twist — only red sevens count +1, while black sevens are 0. Rarely worth the cognitive overhead.

Zen Count: A level-2 system with slightly different card values than Hi-Opt II. Similar trade-offs.

Omega II: Another level-2 system, technically the most powerful published balanced count. Almost nobody plays it accurately at speed.

Wong Halves: A level-3 system with fractional values (+½, +1, etc.). Extraordinary theoretical edge, roughly impossible to execute cleanly. More a mathematical curiosity than a practical tool.