Just for fun · Oct 2026
I recently went to Vegas, and that's where this page started. I love Vegas, but mostly for the attractions, the people watching, and the sheer scale of everything. I grew up on Vegas movies, and Casino and Vegas Vacation are near the top of my list. I'm under no illusion that you can make a living gambling. Most people lose, and that's the deal.
Table games feel straightforward: you know the odds going in. Slots are different, and they fascinate me. Walk into any casino and it's mostly slots, and YouTube is full of people hitting it big. So I wanted real answers to three questions, plus an excuse to build my own (very clean) slot machine, because I honestly think the animations are clever.
01 · What are the odds?
An example machine I made up, bonus round and all, with its paytable out in the open (real ones keep theirs secret). It pays back 92% over the long run, bonus included. Enjoy the lights, then check the meter.
Press spin. Bet is $1 a spin.
Luck meter
Spin to see where you land.
Compared with 1,000 simulated players who made the same number of spins on this exact machine.| Line | Pays | Odds |
|---|
Two bonus symbols land on the first two reels, the last reel slows to a crawl, and it feels like you almost got the bonus. You didn't. The result was decided the instant you pressed spin, and the reels are just the reveal. Real machines can make near misses look a lot more common than real wins through how symbols are mapped onto the reels. Mine flags them so you can count them.
When two bonus symbols land, my last reel slows down and glows, just like the real ones. It's great drama, and it means nothing: the outcome already exists. The slowdown is there to make the moment feel bigger, win or lose.
On my machine a single cherry pays 50¢ on a $1 bet. You get the "WIN," but you're down 50¢. On modern multi-line games this happens constantly: bet $2.50 across 50 lines, "win" $1.20, celebrate a $1.30 loss. The counter tracks every one. Flip the machine to 5 or 9 lines and watch how often it happens, then see the numbers just below.
Free spins feel like a gift. They're not extra money: a slice of the machine's return is set aside to pay for them. On my machine, 23.0 of the 92 points of payback come from the bonus round (25% of everything it pays back), so regular spins return only 69.0%. The house edge is the same either way.
The picker on the machine goes from 1 to 5 to 9 lines. More lines means more ways to win on every spin, so it feels like better odds. Does it change them? These numbers come from counting every possible spin of this machine, not from sampling.
At 1 line a spin costs $1. A line pays on 26.7% of spins, but 17.9% of all spins pay less than you bet.
| Lines | Cost / spin | Payback | Any line pays | Pays less than bet | Loss / hour |
|---|---|---|---|---|---|
| 1 | $1 | 92.0% | 26.7% | 17.9% | $48 |
| 5 | $5 | 92.0% | 64.4% | 42.7% | $240 |
| 9 | $9 | 92.0% | 68.6% | 48.8% | $432 |
An example machine, not a real game. "Loss disguised as a win" means a spin where a line pays something, but less than the total bet, and no bonus. Loss per hour is the cost per spin × 600 spins an hour (my assumption, the same one used elsewhere on this page) × the 8% the machine keeps.
The takeaway
More lines don't change the odds per dollar. They raise the cost of every spin, and they make small "wins" a lot more common.
Every line here pays exactly like the paytable, so a dollar on one line is worth the same as a dollar on nine: the machine pays back 92.0% either way, and the numbers above show it. What changes is the price of each spin and how often something lights up. At 1 line, 26.7% of spins pay anything. At 9 lines it's 68.6%, and about 7 in 10 of those wins pay less than the $9 you bet.
That's the trick: a "win" on 9 lines can be a loss, while the lights and sounds celebrate anyway. And because each spin costs more, the expected loss per hour grows in step with the lines ($48, $240, $432 at my assumed 600 spins an hour). Nine lines isn't a worse bet per dollar than one. It's just nine times the dollars.
One player is a story. A thousand players is data. Pick a machine, and I'll simulate a whole casino floor of people sitting down at it. Same math for everyone, with each spin drawn by a seeded random number generator.
Share of players still ahead after…
Example machines, not real games. Each one is a paytable I built so the long-run return matches your payback setting exactly. Players never run out of money and never quit early, which flatters the "still ahead" numbers if anything.
Why YouTube lies (by accident)
Run the crowd a few times and you'll always find someone way up. That's the person with the video.
Nobody uploads "I put $600 in and lost $50, like the math said I would." Thousands of people play, a handful hit big, and only those sessions become videos. The highlighted line above is that player. It's real luck, it really happens, and it says nothing about your odds. It's the same thing as Rusty winning four cars in Vegas Vacation while Clark loses everything.
Look at the bold line instead. Over a short session, results are all over the place. The longer everyone plays, the more tightly the crowd's average lines up with the dashed line: the house edge, to the decimal.
02 · Can you improve your odds?
Real numbers, not examples. Nevada's Gaming Control Board publishes how much its casinos' slots keep of every dollar played, by denomination and area. These are the twelve months from Sep 2025 to Aug 2026.
† Fewer than 100 machines in that group, so the number swings a lot from year to year. Payback = 100% minus the share kept. Multi-denomination machines let you pick the coin size, and the report doesn't split them out.
Statewide, penny slots kept 8.86% of what was played (91.14% back). $1 machines kept 6.26% and $5 machines 4.37%. On the Strip, pennies kept 10.53%. And a "penny" game rarely costs a penny: most bet across many lines per spin.
At an illustrative 600 spins an hour, $1 a spin on a penny game at 8.86% costs about $53 an hour on average. $5 a spin on a $5 machine at 4.37% costs about $131. A higher payback on a bigger bet still takes more dollars. The bet size matters more than the percentage.
All slots together kept 7.14% statewide, 7.75% on the Strip, and 8.36% downtown, versus 5.66% in Reno and 5.29% in Wendover. That's a real difference, but it's a few cents on the dollar. Every row is still a loss.
Nevada's nonrestricted casinos ran about 126,000 slot machines, and they won $11.0 billion on them over these twelve months, 68% of all gaming win. 71% of the machines are multi-denomination.
| Area | Machines | Slot win | Kept | Paid back |
|---|---|---|---|---|
| Statewide | 125,706 | $10,952M | 7.14% | 92.86% |
| Las Vegas Strip | 35,133 | $5,035M | 7.75% | 92.25% |
| Downtown Las Vegas | 10,001 | $641M | 8.36% | 91.64% |
| North Las Vegas | 3,968 | $272M | 7.63% | 92.37% |
| Laughlin | 5,743 | $457M | 8.27% | 91.73% |
| Mesquite | 2,797 | $185M | 6.73% | 93.27% |
| Reno | 8,837 | $684M | 5.66% | 94.34% |
| Wendover | 3,607 | $227M | 5.29% | 94.71% |
Low-volatility games pay small amounts often. High-volatility games go quiet for a long time and then pay big. Flip the volatility in the simulator above: the crowd spreads out a lot more, but the bold average line lands in the same place. You're choosing how the loss feels, not how big it is on average.
A random number generator picks every result the moment you press spin, with no memory of the last spin. A machine that hasn't hit in hours is exactly as likely to hit as one that just paid. The simulator checks this every run: the hit rate right after ten straight losses is the same as the hit rate overall, give or take noise.
Hitting the button to stop the reels early, playing at certain hours, switching machines after a jackpot, using a player's card or not: none of it changes the outcome. A player's card does get you comps and free play, which is real value, just not better odds.
The only real levers: the game's payback (which you can't see), denomination and location (the averages above), and how much money goes through the machine. Your expected loss is roughly bet × spins × house edge, so fewer, smaller spins is the one sure way to lose less.
03 · Does advantage play matter?
This is the part that surprised me. A few kinds of machines carry something over from the last player, and once in a while that makes the next spin worth more than it costs.
Some progressive jackpots must hit before the meter reaches a stated cap, like "must hit by $1,000." The meter starts at a reset amount and grows a little with every bet. The trigger point is somewhere between reset and cap, so the closer the meter is to the cap, the sooner it has to go and the bigger it'll be.
A few games really do save progress toward a bonus, like collected symbols or a filling meter, and it stays on the machine when a player leaves. Walk up to one that's nearly full and the last player already paid for most of your bonus.
Most games that look like that are just for show, though. On Mo' Mummy, the piles of gems over the reels grow and grow, but the bonus can trigger on any spin no matter how tall they are. Players call this "perceived persistence," and it's built to keep you spinning so you don't walk away from a "full" machine. The tell is the help screen: if it says the display doesn't affect results, believe it.
Assumes the hit point was picked at random anywhere between reset and cap, you're the only one playing, and you keep playing until it hits. Expected value = expected jackpot − base-game edge × expected dollars played. With these settings, it turns positive at a meter of –.
So, does it matter?
Yes, the math is real. In practice, it's a grind for a handful of people, not a trick for a weekend in Vegas.
Drag the meter down and the value disappears fast. It's only positive in the last stretch before the cap, and casinos know exactly which machines those are. The people who do this for real scout dozens of casinos, compete with each other for the same few machines, need a big bankroll to ride out the worst case, and still only earn a small edge per hour. If someone else is playing the same jackpot, your share of it shrinks too.
For everyone else, the honest answer to "can you beat slots?" is no. The good news is that knowing that makes them more fun: set a budget for entertainment, enjoy the lights, and treat a win as a bonus.
Data and method
Just for fun, and not gambling advice. If gambling ever stops being fun, the National Council on Problem Gambling runs the National Problem Gambling Helpline: call or text 1-800-MY-RESET. In the spirit of honesty: AI coding agents wrote most of the code here. I came up with the idea and the questions, and checked the numbers. Everything is vanilla JS, canvas, and SVG, with no libraries.