Home
Bet-spread simulator

Know your edge before you sit down.

A hand-by-hand Monte Carlo simulator for the bet spread you actually plan to play. Set your count, deviations, table rules, and bet ramp, then read the edge, the variance, and the risk of ruin. These are the figures you size a bankroll on.

Open the simulator
The bet-spread simulator showing EV per hour, standard deviation, risk of ruin and the long-run outcome band for a configured run.
A finished run: EV, standard deviation, risk of ruin, and the outcome band.
How it works

It plays the cards.

The simulator shuffles real shoes and plays every hand to the end: hitting, doubling, splitting, taking insurance, and playing one or two hands at a time against the dealer. Every seat is dealt from one shared shoe, so effects that a closed-form formula has to approximate instead fall out of the play itself. That includes how two hands move together, the extra variance that doubles and splits add, and the way your edge shifts with the count.

Everything you can set

Configure the whole game.

The count

  • Built-in Hi-Lo, Zen, and Hi-Opt II, or any balanced counting system you have saved (like CAC2!).
  • Deck divisor for the true count: whole, half, or quarter deck.
  • Choose how decks-remaining is rounded before dividing: round up to match the app, or round to nearest to match CVCX.

Strategy

  • Textbook basic strategy, or play your own custom basic-strategy chart.
  • Deviations off, a named H17/S17 preset set, or one of your own custom deviation sets.
  • Insurance is taken on its own index, including the index from a custom deviation set.

The table

  • 1, 2, 4, 6, or 8 decks, with penetration set by how many decks are dealt before the cut card.
  • Dealer hits or stands on soft 17, double after split, resplit aces, and up to 2–4 split hands.
  • Surrender, European no-hole-card, and a double rule of any first two cards, 9–11, or 10–11.

The bet

  • A full bet ramp by true count, from −2 (and below) through +17 (and above).
  • One or two hands at each true count, so multi-hand play and its correlation are modeled.
  • Set your bankroll and rounds-per-hour to turn per-round results into hourly figures.

How you play it

  • Play every round, or back-count: wong in and out of the same shoe, or leave for a fresh table when the count goes negative.
  • Set the true count you sit out below, and the dead time it costs to find a new table.
  • Seat 0–6 other players at the table to deplete the shoe the way a real game does.

What it reports

  • EV per hour and per round, standard deviation per hour, risk of ruin, and hours to N0.
  • A per-true-count breakdown: how often each count occurs, your average bet there, and its EV.
  • A live progress read-out as the run converges, with the effective rounds-per-hour when wonging.
The simulator's configuration panel: count system, deviations, table rules, bet ramp by true count, wonging and other players.
Every input the engine accepts: the count, the rules, the ramp, and how you play it.
What it tells you

The numbers you size a bankroll on.

Every run reports your edge per hour and per round, standard deviation per hour, risk of ruin for your bankroll, and the hours to N0 — the point where expected profit equals one standard deviation. The result is drawn as a long-run outcome band: the expected bankroll trend, the ±1 and ±2 standard-deviation range around it, the risk-of-ruin floor, and N0 marked on the timeline.

  • A shoe buy-in: the bankroll needed to absorb the drawdown of a single shoe, taken from the per-shoe drawdowns the engine measures. Other bet-spread simulators do not report it.
  • A per-true-count table: how often each count comes up, your average bet there, and the EV it earns.
  • Trip risk for a fixed cash-on-hand over a set number of hours — expected return, the chance of busting, the chance of ending down, and a likely ending-bankroll range.
  • Sample variance trajectories so you can see what a run of hours can really look like, not just the average.
The long-run outcome band with its expanding standard-deviation cone and N0 marker, alongside the per-true-count breakdown of frequency, average bet and EV.The long-run outcome band with its expanding standard-deviation cone and N0 marker, alongside the per-true-count breakdown of frequency, average bet and EV.
The outcome band with N0 marked, and the per-true-count breakdown behind it.
Accuracy

Checked against the benchmarks.

The default configuration reproduces a published double-deck Hi-Lo benchmark, and its EV/hr, SD/hr, risk of ruin, and N0 land within a fraction of established tools. The methodology and a side-by-side comparison are shown on the simulator, so you can see how it lines up before you rely on a number. You can save and name runs to compare configurations, and an identical run returns from cache instead of recomputing.

Validated against CVCX →Compared with PBS and blackjack.tools →

FAQ

Is this a formula or a real simulation?
It is a hand-by-hand Monte Carlo simulation. The engine shuffles real shoes, deals the cards, and plays every hand out: splits, doubles, insurance, and multiple hands. Because the hands are dealt from a shared shoe and dealer, the things a formula has to approximate come out of the play itself: two-hand correlation, the extra variance from doubling and splitting, and insurance.
How accurate is it?
The defaults reproduce a published double-deck Hi-Lo benchmark, and the EV/hr, SD/hr, risk of ruin, and N0 land within a fraction of established tools like blackjack.tools and Pro Betting Software. Larger run sizes tighten the estimate. The methodology and benchmark comparison are shown right on the simulator.
What is N0?
N0 is the number of hours of play at which your expected win equals one standard deviation — roughly, how long until the edge is expected to pull clear of normal swings. On the outcome band it is marked exactly where the unlucky (−1 SD) path is first expected back to break-even.
Can I use my own counting system and deviations?
Yes. Any balanced counting system you have saved, your custom deviation sets, and your own basic-strategy chart can all be pulled into a run, so the simulation reflects how you actually play rather than a textbook default.
Can I model back-counting and a full table?
Yes. You can back-count in and out of the same shoe or leave for a fresh table when the count goes bad, set the true count you sit out below and the time a table search costs, and seat other players whose play depletes the shoe.
Does it save my runs?
You can bookmark and name runs to compare configurations later, and an identical run is served instantly from a cached result instead of being recomputed.

Run your spread.

Set up the game you play and the way you bet it, and see the edge, the variance, and the risk before any money is on the table.

Open the simulator