By CardCount Academy · Reviewed · 10 min read
There are dozens of published card counting systems, and beginners often lose weeks choosing between them. The honest summary is that the differences in results are small, and the difference between counting accurately and counting with mistakes is large. This guide lays out the 7 systems available on this site, explains the terms used to compare them, and ends with a recommendation.
The card values, side by side
| System | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | A |
|---|---|---|---|---|---|---|---|---|---|---|
| Hi-Lo | +1 | +1 | +1 | +1 | +1 | 0 | 0 | 0 | −1 | −1 |
| KO | +1 | +1 | +1 | +1 | +1 | +1 | 0 | 0 | −1 | −1 |
| Red 7 | +1 | +1 | +1 | +1 | +1 | +1 red | 0 | 0 | −1 | −1 |
| Omega II | +1 | +1 | +2 | +2 | +2 | +1 | 0 | −1 | −2 | 0 |
| Halves | +0.5 | +1 | +1 | +1.5 | +1 | +0.5 | 0 | −0.5 | −1 | −1 |
| Zen Count | +1 | +1 | +2 | +2 | +2 | +1 | 0 | 0 | −2 | −1 |
| Hi-Opt II | +1 | +1 | +2 | +2 | +1 | +1 | 0 | 0 | −2 | 0 |
The 10 column includes jacks, queens and kings. In Red 7, red sevens count +1 and black sevens 0. This table is generated from the same values the drills and quizzes use.
What the tags are trying to copy
Every system is an approximation of one underlying fact: removing a card from the deck changes the player's expectation by a specific amount. Peter Griffin calculated these "effects of removal". For a single deck they are approximately:
| Card removed | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | A |
|---|---|---|---|---|---|---|---|---|---|---|
| Player gains | +0.38% | +0.44% | +0.55% | +0.69% | +0.46% | +0.28% | 0.00% | −0.18% | −0.51% | −0.61% |
The 5 is the most valuable card to see leave; the ace and the tens are the most costly. A level-one system rounds all of this to +1, 0 and −1. Higher-level systems use +2 or fractions to follow the real numbers more closely — you can see Halves giving the 5 its extra weight and the 9 a small negative tag.
Three ways systems differ
Level
The level is the largest tag. Level one uses only ±1 (Hi-Lo, KO, Red 7). Level two uses ±2 (Omega II, Zen, Hi-Opt II). Level three goes further (Halves, usually doubled in practice to avoid fractions). Each level is more accurate on paper and noticeably harder to keep at table speed.
Balanced or unbalanced
A balanced system adds up to zero over a full deck and needs a true count conversion. An unbalanced one adds up to a positive number and is used straight from the running count, starting from a negative number that depends on the decks in play. The KO guide explains how that works.
How the ace is counted
The ace is a strange card: it is very valuable for betting (blackjacks pay 3:2) but behaves like a small card when you are deciding how to play a hand. Systems that count it as −1 or −2 are better at sizing bets. "Ace-neutral" systems — Omega II and Hi-Opt II count it as 0 — are better at playing decisions, and are meant to be used with a separate side count of aces for betting. That means tracking two numbers at once.
The efficiency numbers
- Betting correlation (BC) — how well the tags track the effect of removal above. This drives how accurately you know when to bet more. In shoe games it matters most.
- Playing efficiency (PE) — how well the count identifies when to deviate from basic strategy. It matters most in single- and double-deck games.
- Insurance correlation (IC) — how well the count predicts whether insurance is a good bet.
| System | Type | BC | PE | IC |
|---|---|---|---|---|
| Hi-Lo | Level 1, balanced | 0.97 | 0.51 | 0.76 |
| KO | Level 1, unbalanced | 0.98 | 0.55 | 0.78 |
| Red 7 | Level 1, unbalanced | 0.98 | 0.54 | 0.78 |
| Omega II | Level 2, balanced | 0.92 | 0.67 | 0.85 |
| Halves | Level 3, balanced | 0.99 | 0.56 | 0.72 |
| Zen Count | Level 2, balanced | 0.96 | 0.63 | 0.85 |
| Hi-Opt II | Level 2, balanced | 0.91 | 0.67 | 0.91 |
A value of 1.00 would be perfect. These are the figures usually quoted for each count without a side count; published values vary slightly by source and by how they are calculated. With an ace side count, the betting correlation of Omega II and Hi-Opt II rises to roughly the level of the others.
Read the table with two cautions. First, the betting correlations of the popular level-one counts are all within two points of each other. Second, the figures describe a player who never makes an error. A complicated count kept at 95% accuracy underperforms a simple one kept at 100%.
Each system in a paragraph
Hi-Lo
Introduced by Harvey Dubner in 1963 and the reference point for everything else. Most published index numbers, books, software and team play are built around it. See the complete Hi-Lo guide.
KO
Hi-Lo with the 7 counted as +1, published by Vancura and Fuchs in 1998. No true count conversion. An excellent choice if division at the table is what worries you.
Red 7
Arnold Snyder's unbalanced count: only red sevens count. Its pivot sits at zero, right where the advantage begins.
Zen Count
Snyder's balanced level-two count. It counts the ace as −1, so it works without a side count, and offers a real gain in playing efficiency over Hi-Lo for a moderate increase in effort.
Omega II
Bryce Carlson's level-two, ace-neutral count. Strong playing efficiency; needs an ace side count to bet well.
Hi-Opt II
Another level-two, ace-neutral count, with the best insurance correlation in this list. Traditionally used by single-deck specialists with an ace side count.
Halves
Stanford Wong's level-three count from Professional Blackjack. The highest betting correlation here, at the price of half-point tags.
How to choose
- If you are starting out: Hi-Lo. It is the best documented and everything else is explained in terms of it.
- If the true count is the obstacle: KO or Red 7.
- If you already count Hi-Lo without effort and want more: first add index plays and better game selection. Both are worth more than a new system. After that, Zen is the gentlest step up.
- If you play mostly shoe games: betting correlation is what counts, and level one already has it.
Whatever you choose, stay with it. Switching systems resets months of automatic recall, and the gain between any two systems in this list is far smaller than the gain from finding a table with better rules and penetration.
Educational content only. Card counting does not guarantee winnings, casinos can limit or refuse play, and laws differ by location. Read the disclaimer and, if gambling stops being fun, our responsible gambling page.
Sources and further reading
- Peter A. Griffin, The Theory of Blackjack (Huntington Press).
- Stanford Wong, Professional Blackjack (Pi Yee Press).
- Olaf Vancura and Ken Fuchs, Knock-Out Blackjack (Huntington Press, 1998).
- Arnold Snyder, Blackbelt in Blackjack (Cardoza Publishing) — the Red Seven count.
- Bryce Carlson, Blackjack for Blood (CompuStar Press) — the Omega II count.
- Wizard of Odds — Effect of Card Removal in Blackjack