- Joined
- Jul 11, 2008
- Messages
- 1,576
- Reaction score
- 184
- Points
- 63
This article is for anyone playing online casino games who gets frustrated when short-term results don't match what "should" happen mathematically.
The house edge tells you what happens over thousands of bets. Variance tells you what happens over the next ten. Most players obsess over the first number and completely ignore the second, which is backwards because variance is what you'll actually experience every time you sit down to play.
What Variance Actually Means
Variance measures how much results bounce around the average. Low variance means results cluster close to the expected outcome. High variance means they swing wildly.Roulette is relatively low variance. You bet red or black, you win roughly 48.6% of the time. Play 100 spins and you'll probably win 45-52 times. Not exactly 48.6, but close. The swings are predictable. You're not going to win 70 times or lose 80 times - the outcomes are distributed pretty tightly around the average.
Slots are high variance. You can lose 50 spins in a row, then hit a bonus round that pays 200x your bet. Your overall RTP might be 96%, but that 96% is made up of mostly small losses and occasional massive wins. The distribution is lumpy. Most of your spins return nothing or small amounts. A tiny percentage of spins return huge amounts.
This is why two players can play the same slot for the same amount of time and have completely different results. One person hits the bonus round twice and walks away up £500. The other never hits it and loses their entire £200 bankroll. Same game, same RTP, same house edge. Just variance landing differently.
Why This Destroys People Psychologically
Your brain isn't built to understand variance. You see patterns that don't exist.You lose eight blackjack hands in a row and you think the game is rigged. It's not. Eight losses in a row happens about 0.4% of the time in blackjack, which sounds rare until you realize you're playing hundreds of hands per session. Rare things happen constantly when you're making thousands of bets.
Or you win five roulette spins in a row and you think you've found a pattern or you're on a hot streak. You haven't. The wheel has no memory. The next spin is still 48.6% to hit red. The previous five wins mean nothing. But your brain insists they mean something, so you keep betting, and variance swings the other way, and you lose it all back plus more.
The other thing that kills people - they remember wins more vividly than losses. You hit a £800 win on a slot and you remember that session forever. You forget the fifteen sessions where you lost £50 each time. Your brain tells you you're up overall. Your bank account says otherwise.
How Variance Works in Different Games
Blackjack is medium-low variance if you're playing basic strategy. You win roughly 43% of hands, lose 48%, push 9%. The swings are there but they're not wild. You can calculate roughly how big your bankroll needs to be to survive the normal downswings - usually about 20-30 betting units for a few hours of play.Baccarat is similar. Low variance, outcomes cluster close to expectation. You're not going to have huge swings unless you're betting massive amounts.
Video poker sits in the middle. Most hands you lose or get your money back. Occasionally you hit a flush or full house for a decent win. Very rarely you hit a royal flush for a massive payout. The variance is higher than blackjack but not as insane as slots because the pay table is more evenly distributed.
Slots are all over the place depending on the game design. Low volatility slots pay small amounts frequently. You can play for hours on a small bankroll because you're constantly getting small wins that feed back into your balance. High volatility slots are feast or famine - mostly nothing, occasionally huge. You'll either blow through your bankroll quickly or hit something big.
The RTP can be identical on both types of slots. A low variance 96% RTP slot and a high variance 96% RTP slot will lose you the same amount of money over 10,000 spins. But the experience is completely different. One bleeds you slowly. One swings you around violently.
Sample Size and Why 100 Spins Means Nothing
This is the thing people struggle with most. Short-term results are almost entirely variance.You play 100 spins of roulette betting red. You win 40 times. You're down £20 on £100 wagered. Does that mean the game is rigged? No. It means you were unlucky. 40 wins out of 100 is within normal variance for a game where you should win 48.6%. Run it again and you might win 55 times.
The casino doesn't care about your 100 spins. They care about the 100 million spins happening across all their players. On that scale, variance evens out and the house edge dominates. Everyone's individual results are bouncing around randomly, but the casino's aggregate results are predictable.
This is why "I played 200 hands and I'm down 60% of my bankroll" doesn't mean anything. 200 hands is noise. You're still deep in variance territory. The house edge hasn't even kicked in properly yet. You could be down 60% because of bad luck, or you could be playing badly and variance is just making it worse. No way to tell from 200 hands.
Professional gamblers think in tens of thousands of bets. That's when variance starts to fade and edge starts to matter. Recreational players think in hundreds of bets and convince themselves the game is rigged because they lost more than they expected. The game isn't rigged. You just don't understand how much noise there is in small samples.
Bankroll and Variance
If you don't have enough bankroll to survive normal downswings, you're going to go broke even if you're playing a beatable game.Let's say you're playing blackjack with a 0.5% house edge. You're betting £10 per hand. In theory you should lose about £5 per 100 hands. But variance means you might be down £200 after 100 hands just from bad luck. If you only brought £150 to the table, you're broke before the edge even matters.
Slots are worse. High variance slots can easily eat 50-100 betting units before you hit anything significant. If you're betting £1 per spin and you've only got £50, you're probably going to bust out before you hit a bonus round. That doesn't mean the slot is unfair - it means your bankroll was too small for the variance.
The general rule - you need about 100-200 betting units for low variance games, 300-500 for high variance games. And that's just to survive a normal losing streak, not even accounting for the house edge which is grinding you down the entire time.
Most casino players bring way too little money for the bets they're making. They sit down with £100, bet £5 per spin on a high variance slot, go through twenty losing spins in the first five minutes, and they're already down 40%. Then they chase, increase their bet size to try to get it back, and blow out even faster.
Why Winning Streaks Feel Like Skill
You win twelve hands of blackjack in a row and you think you're playing well. You're not. You're just experiencing the positive end of variance.Twelve wins in a row in blackjack happens about once every 4,000 sequences of twelve hands. Rare? Yes. Impossible? No. If you're playing 80 hands per hour, you're seeing about six sequences of twelve hands per hour. Play for 700 hours and you'll probably see it happen once. Feels magical when it does. It's just statistics.
The problem is your brain interprets the winning streak as evidence that you're doing something right. You start to believe you've figured out a pattern or you're reading the dealer or whatever. Then variance swings back and you lose fourteen hands in a row, and now you're convinced the casino flipped a switch. Neither interpretation is correct. It's just noise.
Slots are even worse for this because the variance is so high. You hit three bonus rounds in thirty minutes and win £600. You think you've found a hot machine or you're playing at the right time of day. You haven't. The machine is just running through its RNG and you happened to hit three low-probability events in a short window. Next hour you could hit nothing. Same machine, same RNG, just different variance.
I don't know if there's a fix for this. Human brains aren't wired to accept randomness. We see causation everywhere even when it doesn't exist. The best you can do is know that your instincts are wrong and override them with logic. When you're winning, it's probably variance. When you're losing, it's also probably variance. The only thing that's not variance is the house edge, and that's working against you the entire time regardless of what's happening short-term.
The Gambler's Fallacy
Red hits five times in a row on roulette. You bet black because it's "due." You just fell for the gambler's fallacy.The wheel has no memory. Every spin is independent. The probability of black on the next spin is exactly the same as it was before the five reds - 48.6%. The previous results don't affect future results. This is obvious when you think about it logically, but your brain screams at you that black is due because five reds "shouldn't" happen.
Five reds in a row happens about 2.2% of the time. Not common, but not rare either. You're going to see it if you watch a roulette wheel for any length of time. Seeing it doesn't mean black is more likely next. It just means you witnessed a normal variance event.
The opposite fallacy is thinking you're on a hot streak and the trend will continue. Red hits five times so you keep betting red because it's running hot. Also wrong. Same logic. The wheel doesn't care what just happened. Next spin is still 48.6% for red, same as always.
Both of these mistakes cost people enormous amounts of money. They're betting based on patterns that don't exist, fighting variance with more variance, and the house edge is extracting its percentage the whole time while they chase ghosts.
Variance in Bonus Features
Slots with bonus rounds are high variance by design. Most of your RTP comes from the bonus, not the base game.You might play 200 spins and never trigger the bonus. That's normal. The bonus might trigger once every 100-300 spins on average depending on the game. If you don't hit it in your session, you're almost certainly losing money because the base game RTP is terrible - often 70-80% - with the rest coming from bonus round wins.
This is why slots feel so streaky. You're either in a bonus round and winning, or you're in the base game and bleeding. There's not much middle ground. The variance is intentional. The casino designs it this way because humans find high variance more exciting than low variance even when the expected loss is identical.
Some players prefer low variance slots for exactly this reason. They'd rather lose slowly and play longer than swing wildly and potentially bust out fast. Neither approach beats the house edge, but the experience is completely different. Know which one you're playing.
Can You Use Variance to Your Advantage?
Not really. Variance is random by definition.The only way variance helps you is if you quit while you're ahead. You sit down, you run hot for thirty minutes, you're up £300, you leave. You just rode positive variance and got out before it swung back. That's not exploiting variance, that's just luck plus discipline.
Most people can't do this. They're up £300 and they think they're on a roll. They keep playing. Variance swings negative. They give it all back and then some. The proper play was to leave, but leaving feels wrong when you're winning, so they stay and eventually lose.
The casino doesn't care if you quit while ahead. They know most people won't. And even if you do, you'll probably come back next week and give it back then. They're playing the long game. You're playing a single session. Variance helps you in single sessions occasionally. The house edge helps them across millions of sessions. They win.
If you're serious about minimizing variance damage, you play low variance games and you quit after a fixed number of bets regardless of whether you're winning or losing. But that's boring. Most people are playing for entertainment, not to optimize their expected loss, so they play high variance games and ride the swings and tell themselves stories about why they're winning or losing. That's fine. Just don't confuse the stories with reality.
FAQ
Q: If variance is random, why do I always seem to lose in the long run?Because the house edge isn't random. Variance creates the short-term swings - you win some sessions, lose others. But underneath all that noise, the edge is grinding you down by 1-5% of everything you wager. Over time, variance averages out to zero and the edge is all that's left. That's why you lose long-term. Not bad luck. Math.
Q: Can I reduce variance by making smaller bets?
Smaller bets reduce the absolute size of your swings in pounds, but they don't reduce variance as a percentage of your bankroll. If you're betting £10 per spin on a high variance slot, you might swing £500 in an hour. Bet £1 per spin and you'll swing £50 in an hour. Same variance, smaller stakes. The only way to actually reduce variance is to play lower variance games or make fewer bets total.
Q: Why does it feel like I lose more when I increase my bet size?
Because you probably do. Not because the casino is cheating, but because you're usually increasing your bet size when you're chasing losses, which means you're already in a negative variance swing. You're betting bigger into bad luck, which amplifies your losses. The correct play when variance is running against you is to bet smaller or stop playing, not increase your stakes. But that's counterintuitive so most people do the opposite.
Last edited:
