There’s no shortage of ersatz ‘experts’ trying to sell information and systems on how to make money gambling. Even at games where people *should* know better, there’s countless come-ons for systems to win at slot machines, roulette and even state and provincial lotteries. In this article, we’ll look at a popular keno ‘strategy’ scam that suggests people can win big money simply by tracking keno ‘patterns’.

It’s hard to understand why even experienced gamblers are able to convince themselves that there are ‘secret methods’ that promise to ‘crack the code’ and consistently win money at a game governed by complete randomness. I get at least a half dozen spam emails a day promising a system that will ‘bankrupt the American lottery system’. Last I checked, that hadn’t happened. Fortunately, I don’t get any ‘keno secrets’ spam emails but there’s plenty to be found online. Do a Google search and you’ll find thousands, most claiming that ‘keno patterns’ are a legitimate way to make a killing. One site promises a ‘special report’ of winning keno patterns for a mere $99.95. Another site offers a similar ‘special report’ for $49.95. Still more sites don’t offer suggestions but speak in hushed, reverent tones about the power of winning keno patterns. My favorite of the bunch starts their article by making note of the fact that a random number generator (RNG) distributes the keno numbers and then **completely ignores the reality of random number generation** to assert that ‘keno patterns’ are a legit thing:

“You have to know that Keno game uses an RNG system when extracts the numbers; this means is extracting random numbers, without any connection with previous or next extractions. Yet, as in any game, there are different patterns which can be observed after a long game monitoring.”

So even in a random game governed by a computer based RNG system there are ‘patterns’–not only in keno but in ‘any game’? Uh, no, there’s not….


It’s amazing how much time I spend in these articles explaining a) why ‘random number generation’ means just that and b) why this reality means that countless ‘systems’, superstitions and urban legends just aren’t true. So let’s go over some basics of computer random number generation in gambling machines once again. From a computational standpoint, having a computer generate a random number isn’t particularly difficult. But since slot machines, keno machines, video poker machines, etc. have plenty of computational power for this sort of thing as well as a vested interest in making sure everything is *completely* random they go out of their way to make this more difficult in a variety of ways. In most casino game RNG’s the generation of a number is a complex algorithmic function. In video keno, the ‘card’ has spots numbered 1 through 80. If you think the computer just spits out 20 random numbers between 1 and 80 you would be wrong.

When a gaming machine–slot machine, keno machine, video poker machine, etc–is turned on the random number generation component of the machine’s CPU immediately starts spitting out random numbers. Not between 1 and 80 but between 1 and several billion. It does this hundreds of time per second whenever the device is turned on. It doesn’t matter if the game is being played or not. It just keeps spitting out random numbers. When the player starts the game–pulling the handle on a slot machine, pressing ‘deal’ on a video poker or keno machine, the machine notes the random number that has been generated at that exact millisecond.

To make things even more complex and to further guarantee randomness, the random number generated by the computer is processed through another complex algorithm to determine where the reels stop on a slot machine and what numbers show up on a keno machine. For example, the number generated by the computer may be divided by a set number or a specific formula and the remainder from the product of that division used to determine which numbers on the game board get lit. The various algorithmic functions used to produce the initial random number and to ‘process’ it for further randomization are virtually impossible to reverse engineer. Even if you could, it would require timing down to the millisecond to use this information to influence the outcome of the keno machine.

So that’s why keno patterns simply don’t work. A keno machine isn’t ‘randomly generating’ numbers on the board, it’s generating long complex functions that are used to determine which numbers appear on the board. People that think they can ‘see patterns’ aren’t much different than people that think they can see faces in clouds. And even if you could ‘see patterns’ in winning keno cards keep in mind that it would equal ‘correlation’ not ‘causation’. You might be able to ‘backfit’ a theory that certain patterns produce more winners but there’s no proof of causality nor is their any guarantee that what had some correlation in the past will continue to have it in the future.