How slot machines are random
Last update: May 2020
What does “random” mean?
The basis of slots, and in fact every classic game in the casino, is that the result of every play is random. Random means that every choice is equally likely. For example, let's say you have five balls:
With a random selection, your chances of drawing any ball are the same as drawing any other ball.
Now, some of you are screaming "Nonsense!" because you think my explanation would mean that you're just as likely to hit the jackpot as to hit a non-winning combination. Well, keep your panties on, I'll explain. Looking at the five balls again, imagine that in our game you win only if ball #5 is drawn. Each ball has equal chances of being drawn, but you don't have equal chances of winning, because only one special ball is a winner.
Slot machines work the same way, but with much longer odds. A slot might have one million possible results (combinations of symbols), but only one of those is going to be a winner. Each of the one million results is equally likely, but only one of those million results is the jackpot. (more on jackpots...)
Now, some of you are jumping up and down, thinking that slots can't be random, because you know that slots are programmed to pay back a specific percentage, and you think that a slot that's completely random can't do that! Well, say it, don't spray it. Here's the explanation on that one. We'll use a game that pays back 90%, using a random result. Our game has 10 balls:
You pay a dollar for each draw. If you draw a blue ball, you win a dollar. If you draw the black ball, you win nothing. If the balls are drawn completely randomly, then 90% of the time you'll win a dollar. It should be easy to see that in the long run, this game pays back 90%. So, instead of randomness being incompatible with the 90% payback, it's the randomness that causes the 90% payback.
Here's proof. Click the button to run through 500,000 plays of the above game. The balls are chosen randomly, and the payback is 90%.
Slots work on the same exact principle, except rather than having only 10 possible outcomes, there are thousands to millions. The games are dressed up with music and symbols and multiple lines and reels, but when you crunch the numbers, random spins cause the 90% (or whatever) payback. (more on slot paybacks...)
When results are random, that means that nothing affects the result. Machines continually don't get looser and tighter to meet a specified payback, jackpots don't get more likely just because they haven't hit for a while, the presence or absence of a player's card doesn't change the results, etc. There are no patterns, there are no cycles. Every spin is completely random.
Some people write in asking, "But, but, does [thing X] affect the outcome? Because you didn't mention [thing X] specifically." The answer is still the same. I can't predict all the ways people will come up with that they think might change the spin results, so whatever you want to ask about, then yes, it's included, even if I didn't mention it specifically. Every spin has identical chances as the previous spin, period.
If you're satisfied with this answer, then you can stop reading here, and graduate to my article on how slot machines work. If you're still skeptical that slots are completely random, then read on.
The machines don't get looser and tighter to meet the payback
Let's expand on what we learned above.
Probably the biggest myth about slots is that they continually get looser and tighter to meet the specified payback. It seems that way because people who haven't run through the math think that a specified payback and the randomness principle are mutually exclusive, that you can't have a certain payback if the machine is random. In fact, it's the opposite: the randomness is what creates the payback.
Above we used a game with ten balls. We'll use another example here, a bag of marbles, 95 black and 5 white. You bet a dollar every time you pick one out of the bag blindly. When you pick a marble, if it's black you get your dollar back, and if it's white you don't. It should be pretty easy to see that this is a 95% payback game.
We could turn this game into a machine version by telling the machine to simply pick a number between 1 and 100, and pay you back if the number was between 1-95. I hope it's clear that the randomness is what ensures the 95% payback. The machine doesn't have to get looser and tighter. As long as the picks are random, you'll get back 95%.
Here's another example: Let's say you bet $1 on the flip of a coin, and you get 90¢ when you get heads (and you get your original dollar back), but when you get tails, you lose your whole dollar. This is also a 95% game. The machine version would simply pick a 1 or a 2, and pay you the $1.90 if it picked the 1. For two plays, resulting in one heads and one tails, you paid $2, and got $1.90 back. ($1.90 ÷ $2.00 = 95%.) Just as with the marbles, the randomness ensures that the long-term payback is 95%. The machine doesn't have to get looser and tighter to meet it. A random result means a 95% long-term payback.
While these examples are simple games, slots work the exact same way. The gazillion combinations just disguises the fact that the machines are a bad bet. After all, the example games above seem like obvious sure-losers, and you weren't interested in playing them, right? Who would? So the casino makes the games complex so they're more interesting and so the math behind them isn't as blatantly easy to see.
Randomness is the basis of every classic casino game, not just slots
It shouldn't be hard to believe that slots are random, because
all the other casino games are. If randomness works for
craps and roulette, why wouldn't it work for slots?
Classic casino games work on this basic principle:
- In every round there's a random result (from dice being thrown, cards being dealt, or reels being spun).
- When you win, the payout is less than the odds of winning.
There's no mystery about slots, just like there's no mystery about craps. Why would there be?
In both slots and table games, the casino wins because it has a built-in mathematical edge on the games. The casino doesn't have to screw with the dice to beat players at craps, they don't have to screw with the cards to beat players at blackjack, and they don't have to screw with the machines to beat the players at slots. The payline symbols are chosen at random, and it's the math that ensures that you're a long-term loser.
Randomness is dictated by law.
Every jurisdiction mandates that the outcome of non-skill games
be completely random. That goes for both table games and
slots. Slot makers therefore make the machines completely
random, because they couldn't sell them otherwise. (Many
jurisdictions actually test the games to make sure they're properly
random before the games can be put on the floor. That includes
Nevada.) So, even if the casinos wanted the machines to
operate other than randomly, they don't have a choice. First
of all, somebody else already made the game random, and even if it
were possible for the casino to change the programming, it would be
highly illegal to do so. For example, this is from Nevada
Regulation 14 (PDF):
"[A gaming device] must use a random selection process to determine the game outcome of each play of a game...Each possible permutation or combination of game elements which produce winning or losing game outcomes must be available for random selection at the initiation of each play....The selection process must not produce detectable patterns of game elements or detectable dependency upon any previous game outcome, the amount wagered, or upon the style or method of play."
There you have it. Slots are random, period.
Randomness is shown on PAR sheets.
When a manufacturer makes a slot, the first thing they do is create a PAR sheet. It's a list of the symbols for each reel, and the paytable for the winning combinations. From that, anyone with halfway-decent math skills should be able to calculate the payback for a simple slot without bonus rounds. (Bonus round math is a bit trickier, but still just as random.) It's thus perfectly clear from manufacturer's PAR sheets that the randomness of landing the available symbols, married to the paytable, dictates the payback. I have the most comprehensive list of par sheets available anywhere.
Slot machine simulators return the proper payback by operating randomly.
I programmed a virtual replica of a Blazing 7s slot machine using the original PAR sheet from the manufacturer. This slot, running completely randomly, and without ever getting looser or tighter, properly returns the expected payback in the long run. The randomness ensures the results.
Here's a mini-version of that simulator. Click the button, and it'll spin 704,000 times (a month's worth of full-time play in a casino), and your return will be pretty close to the expected 91.96%.
I programmed this the way all slots are programmed: to pick symbol stops completely randomly. I did not tell it to take previous results into account, and to tighten up if it had been paying out "too much" recently. Here's the actual code for picking the result for each reel:
rnd = Math.floor(Math.random()*72)+1;
Anyone who understands computer programming can verify that that's the code I actually used by checking the source code for this page.
Why you can trust this article
Some people argue passionately on the Internet that slot machines aren't random, and that this article is wrong. Here's what to know about that.
- Those spewing the myths and urban legends cite have no professional experience and cite absolutely no evidence. They pull B.S. out of the air and declare it to be fact. By contrast, I have the arsenal below.
- I've done the actual mathematical design of actual slot machines myself, professionally, for-hire.
- I programmed my slot machine simulator the exact same way that real slots work, with every spin being completely random. You can run it for thousands of spins, and see that it returns the expected payback, without any shenanigans like the slot continually getting looser and tighter over time, or going through "cycles".
- I use the very best sources, such as actual slot machine par sheets from the game manufacturers, and the actual Nevada Gaming Commission regulations on slots. The par sheets describe the actual probabilities of the games, and the laws dictate that spin results must be random. Those who spout B.S. (like that machines constantly get looser and tighter) cite no sources.
- I was an assistant to the legendary Wizard of Odds for ten years; he's done the programming for countless slots for online casinos (which work exactly like their real-life counterparts), and I helped him with his early slot research, even before he published his infamous survey in of how loose and tight various Vegas casinos were.
- I've written professionally about the industry, outside of this website. Casino Player magazine trusted me enough to run one of my articles as a cover story.
I also put my money where my mouth is. To show that there's no such thing as a winning slot system, for years I've offered $10,000 hard cash to anyone who could demonstrate a winning system. No system-seller has ever dared take me up on it, because they know their systems don't work. I also used to offer a separate challenge to those who claim that higher-paying slots are located in certain areas of the casino, but I retired that one after many years because it was hard to keep up with all ways people kept trying to find loopholes to exploit the test, and nobody ever took me up on it anyway.
Did you graduate?
If you're now convinced that slot spins are random, then congratulations! You've graduated and can proceed to my article on how slot machines work. If you still think that slots get looser and tighter in order to meet the specified payback, or that slots aren't random in some other way, and you want to convince me of that idea, then read on.
I probably can't convince you. Did you remember that I explained that I've programmed actual slots? And you haven't? And yet somehow you think you're more knowledgeable on this topic? That's some serious chutzpah, for sure. In any event let me suggest you actually learn enough to do the analysis yourself, so you'll have some actual evidence, rather than blindly making assertions as though they were fact:
- Write a program to generate a million random numbers. Use code similar to what I described above.
- Learn how to do statistical analysis. For example, learn how to do a chi-squared test. That will tell you whether the results you generated in Step 2 are actually random or not.
- Write a program to perform the analysis to test for randomness.
If you don't do this, then you're a non-programmer who doesn't understand statistics, trying to convince a programmer who does understand statistics that he's wrong. And Hillary Clinton is running a secret pedophile ring in the basement of a pizza shop, isn't she?
Play slots online
I suggest you play something other than slots because the slot odds are so bad. You could also play online with fake money, because then it doesn't matter if you lose. A good casino for free-play is Bovada, since it requires no download and no registration. (If you see a registration box, you can close it and continue without registering.) You can play with real money too, though I hope you won't (or at least won't bet more than you can comfortably afford to lose).
All my slot machine articles
- Slot machine basics. How much it costs to play, how much you can win, expected loss, why they're a bad bet, why they're popular, how you can limit your losses, speed of play
- How to play slot machines
- Slot returns. How much they pay back.
- The Randomness Principle. Slots don't continually get looser and tighter as they're played. They don't have to.
- How they work. Explains the randomness principle, and runs through the math to show how a game returns a particular payback percentage. There's a companion page on Par sheets.
- Slot Machine Myths
- Slot Machine B.S. Wrong info that's published elsewhere.
- Strategies. Tips for increasing your chances of winning, and saving money.
- Slot Jackpots. Odds of hitting the jackpot, progressive jackpots, and other jackpot topics.
- Skill-Based Slots. The scoop on the new games in which your results aren't entirely determined by chance.
- Slot Machine malfunctions. How and why slot machines screw up, causing players to think they've won the jackpot when they really haven't.
- Slot Machine Simulator. I programmed an exact replica of the Blazing 7s slot (odds-wise). Click it to play thousands of spins in one second and see how you do.
- List of good Bovada slots. I spent a full day surveying Bovada's voluminous offerings and extracted only the few with nice, modern graphics and mobile-compatibility.