Bet Spread for Card Counters: How Much to Raise When the Count Is High
Counting cards tells you when the shoe favours you. The bet spread is how you turn that into money — and it is where almost all of the edge comes from.
Why counting alone earns nothing
A card counter plays the same basic strategy as everyone else most of the time. The edge comes from one thing: betting more when the remaining cards favour you (lots of tens and aces left) and as little as possible when they don't. If you bet the same amount every hand, knowing the count changes almost nothing.
Our own bankroll model, on a typical 8-deck live table that shuffles around half the shoe:
| How you bet | Edge per hand you bet |
|---|---|
| Same bet every hand | −0.38% |
| Bet only when the true count is +1 or higher, ramping up as it rises | +0.78% |
Same table, same count, same strategy. The only difference is sitting out the bad shoes. That is the whole game.
What a spread looks like
A spread is written as "1 to 8" — the biggest bet is 8 times the smallest. The classic ramp used in counting books is:
| True count | Bet (in units) |
|---|---|
| +1 or lower | 1 (or sit out if you can) |
| +2 | 2 |
| +3 | 4 |
| +4 or higher | 8 |
A unit is your base bet. How big a unit should be depends on your bankroll — see Bankroll and Risk of Ruin. A quarter of the Kelly bet is a sensible default, because the true count you are acting on has an error of roughly ±0.5 and over-betting is punished much harder than under-betting.
Why +1 and not 0
At a true count of 0 the player has no edge on this kind of table — the house edge is about −0.5% off the top and each point of true count is worth about +0.5%. So +1 is roughly break-even, +2 is where the edge is real. Betting from +1 keeps you at the table more often than waiting for +2, and in our model it comes out slightly ahead per hour. Waiting for +3 looks better per bet but you bet so rarely that you earn less overall.
The catch at live online tables
- You cannot sit out forever. Live tables kick idle players after a few minutes. Most counters place the minimum bet every 3 to 5 hands to keep the seat. Those bets are negative-expectation and they eat into the +0.78%.
- Half-shoe penetration. Before about 36% of the shoe is dealt, the true count rarely reaches anything useful; the table reshuffles around 50%. The window where a big bet is justified is roughly a quarter of each shoe.
- Spreads are visible. The casino sees every bet. A clean 1-to-8 spread that follows the count is the easiest pattern to spot, and accounts get limited or closed for it. Many casinos' terms prohibit counting aids outright — check before you play.
- The count has an error bar. CountPilot shows the true count with its ± range. When the range straddles your ramp threshold, the smaller bet is the safe choice.
What CountPilot shows
The floating panel shows the true count, its error, and a suggested bet in units from the ramp above. With the Pro bankroll settings it converts units into money and shows the risk of ruin for your bankroll. It will say "don't bet" when the count is not trustworthy — for example when you joined the shoe mid-way. The decision to bet, and how much, is always yours.
Try it free in your browser