Start with the draw, not the ticket
Keno probability feels opaque because the game mixes two separate pools of numbers. The house draws a fixed set of winning numbers from a large field, and your ticket marks a smaller set of spots from that same field. The question you are really asking is: given that the house drew N numbers from a field of F, what is the chance that exactly K of my S chosen spots are among them? That is a counting problem, and it has a single reliable method.
The method is the hypergeometric distribution. It applies whenever you sample without replacement from a population that is split into two groups — in keno, the numbers you picked and the numbers you did not. The formula counts how many ways the draw can include exactly K of your spots, then divides that by the total number of possible draws.
The formula, piece by piece
Write the field size as F, the number of spots you mark as S, and the number of balls the house draws as N. The probability of matching exactly K of your spots is:
P(K) = [ C(S, K) × C(F − S, N − K) ] / C(F, N)
Here C(a, b) is the combination function: the number of ways to choose b items from a set of a items without regard to order. The numerator has two parts. The first part, C(S, K), counts how many ways the draw can hit exactly K of your chosen spots. The second part, C(F − S, N − K), counts how many ways the remaining drawn numbers can come from the numbers you did not pick. Multiply them together and you have the number of draws that produce exactly K matches. Divide by C(F, N), the total number of possible draws, and you have the probability.
As an example, suppose a keno game draws 20 numbers from a field of 80, and you mark 5 spots. The probability of matching exactly 3 of them is C(5, 3) × C(75, 17) divided by C(80, 20). You do not need to compute this by hand; the formula is the point. The same structure works for any field size, draw size, spot count, and match count the game uses.
What changes when the game parameters change
Every keno variant changes one of the three inputs, and each change moves the probabilities in a predictable direction. It is worth seeing them side by side before you touch a calculator.
| Parameter | What it controls | Effect on match probability |
|---|---|---|
| Field size (F) | How many numbers the draw comes from | A larger field makes any specific match less likely, because each of your spots is a smaller share of the pool |
| Draw size (N) | How many winning numbers are drawn | A larger draw makes matches more likely, because more of your spots get a chance to be hit |
| Spots marked (S) | How many numbers you choose | More spots raise the chance of matching at least something, but lower the chance of matching all of them |
The interaction between S and K is the part most players misread. Marking more spots does not simply improve your ticket. It shifts probability mass toward middling match counts and away from the extremes. A one-spot ticket either matches once or not at all. A ten-spot ticket has many more possible match counts, and the most likely outcomes cluster in the middle of that range, not at the top.
From match probability to payout odds
Calculating the probability of a match count is only half the picture. The other half is what the operator pays for that match count. Keno pay tables are built so that the payout for a given match level is less than the fair payout implied by its probability. The gap between the two is the house edge on that ticket.
To see the structure, compute the probability of each possible match count for your chosen number of spots, multiply each probability by its payout, and sum the results. That sum is the expected return per unit wagered. If the game returns less than one unit per unit wagered on average, the difference is the house edge. The exact edge depends on the pay table the operator publishes, and pay tables vary between venues and online products. The game information panel or rules page is where the current table is posted.
Two tickets with the same spot count can carry different edges if their pay tables differ. A ticket that pays more for a rare high match count and less for common low match counts has a different variance profile than a flatter pay table, even if the long-run return is similar. Volatility does not change the expected return; it changes how lumpy the ride is.
Why the "due number" instinct fails in keno
Because keno draws are independent, no number is ever due. The hypergeometric formula assumes every number in the field has the same chance of being drawn, and that past draws do not alter the composition of the next one. A number that has appeared in ten consecutive draws is exactly as likely to appear in the eleventh as any other number, assuming the draw is fair and the equipment or RNG is sound.
This is a different situation from card games where removed cards change the remaining deck. In keno the field resets completely between draws. The probability of matching K spots on your next ticket is identical whether the last draw matched zero or all of your spots. Any system that tracks hot or cold numbers is recording history, not changing the next draw's odds.
Working through a full example
Suppose a keno game draws 20 numbers from a field of 80 and you mark 3 spots. You want the probability of matching all 3. The formula gives C(3, 3) × C(77, 17) divided by C(80, 20). The first factor is 1, because there is only one way to choose all three of your spots. The denominator is the total number of 20-number draws from 80. The result is a small probability, and the payout for a 3-of-3 match is set accordingly.
Now suppose you want the probability of matching exactly 1 of your 3 spots. The numerator becomes C(3, 1) × C(77, 19). This is a much larger number than the 3-of-3 case, because there are many more draws that include exactly one of your three numbers. The pay table reflects that: a 1-of-3 match typically pays nothing or a token amount, while 3-of-3 pays a multiple.
You can repeat this calculation for every possible match count from zero to your spot count. Summing the probabilities of all possible match counts always gives 1, because one of those outcomes must occur. That is a useful check when you run the numbers yourself.
What the formula does and does not tell you
The hypergeometric formula tells you the long-run frequency of each match count under a fair draw. It does not tell you which numbers to pick, because no selection method changes the probability when the draw is uniform. It does not tell you that a ticket is a good deal, because that depends on the pay table. And it does not predict what will happen on your next ticket, because probability describes distributions, not individual outcomes.
What it does give you is a way to read a pay table with clear eyes. Once you can compute the probability of each match count, you can compare the payout on offer against the fair payout for that probability. That comparison, repeated across the whole pay table, is how the house edge on a keno ticket is built. The arithmetic is not complicated, but it is exact.
FAQ
Does marking more spots increase my chances of winning at keno?
Marking more spots increases the chance of matching at least one number, but it lowers the chance of matching all of them. The probability mass spreads across a wider range of possible match counts. Whether that is better depends on the pay table and what you mean by winning.
Can I use past keno results to pick better numbers?
No. Each draw is independent, and the field resets completely between draws. A number's past frequency does not change its chance of appearing in the next draw, assuming a fair game. Tracking hot or cold numbers records history without changing the next draw's odds.
Is the probability formula the same for every keno game?
The formula structure is the same, but the inputs change. Different games use different field sizes, draw sizes, and spot limits. You must plug in the specific values for the game you are playing. The pay table is also game-specific, so the same match probability can correspond to different payouts at different operators.
Play responsibly
Gambling should be treated as paid entertainment, never as a way to earn income or recover losses. 1-800-MY-RESET (1-800-697-3738)
18+ or 21+ depending on where you are; follow the minimum age that applies to you.
This article is general information about how these mechanics work. It is not legal advice and not a recommendation to gamble or to use any particular operator. Availability and legality differ by jurisdiction — check the rules that apply where you are.