Best Keno Strategy: The Four Choices That Matter and the Myths That Don't

The best keno strategy isn't about which numbers you pick. Every set of numbers of the same size has exactly the same chance of being drawn, so "lucky" picks, hot numbers and clever patterns change nothing. What a keno strategy can do is control what the game costs you and how your results swing. That comes down to four choices: the paytable, the number of spots, your stake and how many games you play.

This guide works through each of those choices with the exact probabilities we calculated for 1 to 10 spots, and explains why "how to beat keno" systems don't work. If you're new to the game, our beginner's guide to free keno games covers the basic rules first.

Why your choice of numbers doesn't matter

Keno draws 20 numbers from 80 without replacement. The chance of getting k hits with n spots depends only on n and k, not on which numbers you chose:

That's the hypergeometric distribution, the standard formula for drawing without replacement. Look at what's missing from it: there's no term for whether your numbers are low or high, in a row, scattered, or drawn last game. A ticket of 1–2–3–4–5 has the same chances as 7–23–41–58–77.

Draws are also independent. A number that came up in the last five games is no more or less likely in the next one. Assuming otherwise is the gambler's fallacy, which Britannica describes as falsely treating independent plays as if they were linked. Our guide to keno patterns takes the pattern myths apart one by one.

Choice 1: the paytable (the biggest lever)

Hit probabilities are fixed by the draw. The paytable decides how much of each bet comes back to you:

Here is a hypothetical 1-spot example to show how it works. With 1 spot you hit 25% of the time. If a game paid 3.6 units back (including your stake) on a hit, RTP = 0.25 × 3.6 = 90%, a 10% house edge. If it paid 3.2 units, RTP = 0.25 × 3.2 = 80%, a 20% edge. Same draw, twice the cost.

We don't have verified paytables for specific games, so we won't rate one. What Wizard of Odds reports is a wide range:

The gap inside each range matters more than anything else you can do. Moving from an 84% game to a 95% game cuts your expected loss from 16 to 5 USDT per 100 USDT bet, about two-thirds less.

How to use it: look for the RTP or paytable in the game's help screen before you play. In Great Britain, the Gambling Commission's technical standard RTS 3 requires operators to make information on the likelihood of winning available before play. If a game doesn't show it, choose one that does.

Choice 2: how many spots (sets the ride, not the edge)

Spot count changes the shape of your results. We calculated the exact distribution for every spot count from 1 to 10. The table below shows three things for each: the most likely number of hits, the chance of hitting at least half your spots, and how rare a full catch is.

What this means for strategy:

• Low spot counts (1–4) give frequent, small results. With 2 spots you catch at least one about 44.0% of the time.
• High spot counts (8–10) are long-shot tickets. With 10 spots the most likely result is just 2 hits (29.5%), and 5 or more hits comes up in about 6.47% of games (the sum of the 5-to-10-hit rows). The paytable concentrates its return in those rare results, so you'll spend long stretches losing small amounts.
• Spot count doesn't remove the edge. Whether a 4-spot or a 10-spot game is cheaper depends only on that game's paytable for each count, and Wizard of Odds' figures are ranges, not per-spot guarantees.

Choice 3: stake size

Your stake doesn't change your odds. It scales everything else:

A smaller stake means more games for the same budget and the same expected cost per unit bet. That's useful in keno because results are lumpy: with high spot counts, a budget that covers only a few dozen games can easily run out before any big hit arrives. Decide your session budget first, then divide it by the number of games you're willing to play to set your stake.

Choice 4: pace and number of games

Online and video keno can run one game after another, sometimes with an auto-play button. Expected loss grows with every game, so speed is a cost.

Here's a hypothetical illustration at a 10% edge and 1 USDT a game: 20 games cost about 2 USDT on average; 200 games cost about 20 USDT. Nothing about the odds changed, only the number of bets. Live keno draws are typically slower, but its edge range is higher, so slow doesn't automatically mean cheap.

How to beat keno? What doesn't work

These are the claims behind most "how to win keno every time on machine" searches. None survives the math:

• Hot and cold numbers. Each draw is independent. Past results don't shift future probabilities.
• Patterns and shapes on the ticket. Rows, columns, corners and clusters have the same odds as any other set of the same size.
• Waiting for a machine to be "due". A 10-of-10 catch stays about 1 in 8,911,711 every game, however long it's been.
• Raising stakes after losses. That's a Martingale-style progression. It increases the amount bet, and therefore the expected loss, without improving any single game.
• Playing more cards to "cover" more numbers. Each extra card is another bet with the same edge. Our 20-card keno guide shows how multi-card play multiplies the cost.

Zero hits is common. With 10 spots, 4.58% of games produce no hits at all, so roughly one game in 22 your ticket misses all 10 numbers.

A keno strategy checklist

1. Find the RTP or paytable. Prefer the highest RTP you can find for the spot count you want.
2. Choose your spot count for the ride. Few spots if you want frequent small results; many if you accept long dry spells for rare big hits.
3. Set a session budget you can afford to lose entirely.
4. Set the stake so the budget covers the number of games you planned.
5. Avoid auto-play or set a hard game limit if you use it.
6. Pick your numbers any way you like. Birthdays, random picks or quick pick: the odds are identical.

Keno vs lower-edge games for USDT players

If you play with USDT, a stablecoin keeps your bankroll steady in dollar terms, so the only swings come from the game. Keno's edge is high compared with crypto-native games. Crash games and dice publish a 1% edge, so 100 USDT of bets costs about 1 USDT on average versus 5–16 USDT at video keno. Our dice odds calculator shows the expected loss for any win chance at a 1% edge. Mines uses the same hypergeometric math as keno, applied to a 5×5 grid.

Common mistakes

• Spending time on number selection and none on the paytable.
• Assuming more spots means better value because the top prize is bigger.
• Treating a recent dry spell as a sign a big hit is coming.
• Raising the stake or adding cards to win back losses.
• Letting fast or auto-play games run past your planned number of games.

For the full odds tables and all our keno guides, visit the keno hub.