Bankroll & Variance Simulator
See what perfect basic strategy actually looks like over many hands — streaky and unpredictable in the short run, but grinding toward the house edge in the long run.
Bankroll basics
Expected loss is simple multiplication
Over the long run, what a game costs you is the total amount you wager times its house edge. Wagering $5,000 over an evening at blackjack's roughly 0.5% edge costs about $25 on average; the same $5,000 on double-zero roulette at 5.26% costs about $263. The edge is small on any one hand — what changes the bill is how much action you put through a game and how big its edge is.
In the short run, variance dwarfs the edge
A single blackjack hand has a standard deviation of roughly 1.1 to 1.2 bets. Over 100 hands at $10, the house edge predicts an average loss of about $5 — but the standard deviation of your result is around $115. That's why plenty of players finish a session ahead even though the math runs against them, and why a hot streak says almost nothing about the next hand.
Size your bets to your bankroll, not your hopes
Your bankroll is money you can afford to lose, kept separate from rent, bills, and savings. The smaller each bet is relative to it, the more hands you can play before an ordinary losing streak ends your session — which is why the simulators above let you change bet size and watch how quickly runs go bust. No bet size makes the edge disappear; it only decides how long you get to play.
Set limits before you sit down
Decide your loss limit and your time limit before the first hand, when you're thinking clearly, and treat both as fixed. Chasing losses by raising your bets is the same trap the betting-systems tab illustrates: a bigger bet against the same edge loses more, not less. If gambling is affecting your life, the National Problem Gambling Helpline is 1-800-522-4700.
Why the lines don't agree with each other
Each colored line is one simulated player, playing the exact same strategy against the exact same odds. Short-run results swing wildly — that's variance. The bold gold line is the average across all runs, and it drifts slowly toward a loss because blackjack's ~0.5% house edge is baked into every hand, even under perfect play. This is the entire reason casinos can offer a "fair-feeling" game and still profit reliably: they only need the average across millions of hands and players, not any single outcome.
The assumptions behind this simulation
This uses approximate published outcome frequencies for basic-strategy blackjack: about a 42.3% win rate, 8.5% push rate and 49.2% loss rate. Roughly 10.8% of wins are a 3:2 natural blackjack. (About 4.8% of all hands are a natural, but naturals almost always win, so their share of the wins is over twice that.) It ignores bet-size changes from doubling and splitting and assumes a flat bet every hand. So it is a simplified picture of variance. It is not a betting system, and not a way to predict any individual session.
Flat betting
The same bet every hand, win or lose.
Martingale
Double your bet after every loss; reset to base bet after a win.
Fibonacci
Step forward in the Fibonacci sequence after a loss, back two steps after a win.
% of runs that went bust
Why progressive systems don't beat the house edge
Martingale and Fibonacci feel like they should work — you're "owed" a win after a losing streak, so betting bigger seems safe. But every individual bet still carries the same house edge, regardless of how much came before it or how much you're betting. These systems don't just change the
of your risk (lots of small wins, punctuated by rare but catastrophic losing streaks) — run the head-to-head comparison above and you'll typically see Martingale and Fibonacci post a
average ending bankroll than flat betting, on top of a dramatically higher bust rate. That's because doubling down during a losing streak means betting more money against the same negative edge, which loses more in dollar terms, not less.
Why the table maximum matters
Martingale specifically requires doubling forever to guarantee a win eventually — but every real table has a maximum bet, and a long losing streak (which happens more often than most people expect) can hit that ceiling before you recover your losses. That table max isn't an accident: it exists specifically to cap Martingale-style systems.