House edge is the long-run expected loss per bet. Variance is how far your actual results swing from that long-run average in the short term. Both are built into every casino game, and confusing one with the other is why players think Plinko high risk is 'rigged' when it's just behaving exactly as the math predicts. This pillar explains house edge and variance separately, how they interact, why a 99% RTP game can still burn a bankroll in 30 minutes, and how to read the combination to pick games that fit your goals. All testable free on FakeStake.
House Edge — The Long-Run Cost
House edge is the percentage of every bet the casino keeps on average. On a 99% RTP game, the edge is 1% — every $100 wagered returns $99 on average after enough rounds. That's the game's built-in price. You can't strategy your way around it on casino originals (Mines, Plinko, Crash, Dice) because the payouts are calibrated specifically to hit that RTP. Blackjack is partially an exception — basic strategy is required to REACH the quoted 99.5% RTP, and non-basic play bleeds additional percent. Similarly in video poker with perfect play.
Variance — Short-Term Swings
Variance measures how spread out your actual results are around the expected value. Two games can have identical RTP and completely different variance. Flip (99% RTP) wins or loses one unit per round with equal probability — low variance, slow grind. Plinko 16-row High (99% RTP) wins up to 1000× or busts at 0.2× — extreme variance, lottery-shaped outcomes. Same expected return, radically different session experience.
House Edge × Variance — The Full Picture
A low-edge high-variance game (like Plinko High) can drain a bankroll fast even though the math favors it in the long run. A high-edge low-variance game (like slots at 94% with flat payouts) grinds steadily but predictably. The combination determines how casino games actually feel.
| Game | RTP | Variance | Session feel |
|---|---|---|---|
| Blackjack (basic strategy) | 99.5% | Low-Medium | Steady with occasional big wins |
| Flip | 99% | Low | Slow grind, 50/50 rhythm |
| Dice / Limbo (moderate target) | 99% | Medium | Balanced — some streaks but manageable |
| Plinko 16 rows High | 99% | Very High | Long bust streaks, rare huge wins |
| Mines 20 mines, 5 picks | 99% | Very High | Jackpot hunt — mostly lose, rarely multi-× |
| Slots (typical) | 94-97% | High | Grinding with bonus-round spikes |
Why 99% RTP Still Burns Bankrolls
1% house edge sounds tiny. It isn't — it compounds. If you bet a total of 100× your bankroll across a session (say $1 bets for 100 rounds with a $1 starting balance), the expected loss is 100 × 1% = 1 bankroll. On a high-variance game the actual result swings wildly around that expectation, but the central tendency is to lose everything given enough rounds. This is why 'just play the EV and you'll break even' is wrong — EV assumes infinite rounds with no ruin; reality has finite rounds and ruin ends the session.
Picking Variance to Match Your Goal
If the goal is long entertainment per dollar, pick low-variance high-RTP games — Blackjack with basic strategy, Flip, Dice at moderate win chance. If the goal is chasing a specific big win (with the understanding you'll probably bust first), pick high-variance options — Plinko 16 rows High, Mines with many mines and many picks, Pump Expert, Limbo at high target multipliers. Matching variance to goal matters more than the exact RTP difference, which is usually sub-1%.
Testing Variance Free
Pick any game on FakeStake, run 200 flat-bet rounds in a row, record your ending balance. Do it three times for the same game. The spread of final balances IS the variance. Low-variance games cluster tightly around expected return; high-variance games produce one near-zero session, one near-break-even, and one with a huge win. This experiment takes 20 minutes and teaches more about variance than any table of numbers.
You might also like
