Keno odds and RTP: the maths of every number of spots

Início » Keno in the UK: How It Works, What It Pays and the Odds » Keno odds and RTP: the maths of every number of spots
Last updated:

Odds and return to player answer two different questions, and keno shows the difference more clearly than most games. Odds tell you how likely a result is, such as catching all five spots. Return to player, or RTP, tells you what share of every chip staked the game gives back, averaged across a very long run, and the house edge is simply what is left. This guide on keno odds and RTP works out both across the ticket sizes from 1 to 10 spots in our model, a draw of 20 from 80 with a published sample paytable, and then sets the results beside what published surveys found for real games. We calculated every figure in code, so the sums can be repeated, and the surveys give a sense of how far a real game may sit from our sample.

Key Takeaways: Keno Odds, Returns and House Edge

  • RTP vs. Probability: Win frequency does not determine long-run value. True expected return comes strictly from multiplying each catch’s exact chance by its awarded paytable prize.
  • Edge Variance Across Spots: House edge changes drastically solely through the paytable structure: a 1-spot ticket loses 25.00%, while a 6-spot ticket cuts the edge to 5.01%.
  • Live vs. Video Discrepancy: Real-world casino surveys show live keno holds heavy house edges of 20% to 50% (50%–80% RTP), whereas electronic and video keno generally offer better returns (84%–95%).
  • Volatility Trumps Hourly Edge: High single-round swings (up to 36 chips standard deviation on 7 spots) mask the underlying 5% house edge across short sessions, though the cost accumulates steadily with game speed.

Odds Versus Return

A result can be rare and still be a good deal, or common and a poor one, and only the return tells you which. Catching 2 of 2 spots has a chance of 6.01%, which is 1 in 16.6, and the table pays 15 chips for it. Catching 5 of 5 has a chance of 1 in 1,551, yet it pays 838 chips.

What links chance and prize is the return. For each result, multiply its chance by its prize. Add up every result and you have the average number of chips the ticket returns for a stake of 1 chip, and that is the RTP.

The Return Sum in Three Steps

Take a 5-spot ticket as the example, with the chances from the spots and catches guide and the prizes from the paytable guide.

  1. List each result that pays, with its chance: 3 catches at 8.39%, 4 catches at 1.21% and 5 catches at 0.064%.
  2. Multiply each chance by its prize of 3, 13 and 838 chips, which gives contributions of 25.18%, 15.72% and 54.04% of the stake.
  3. Add them up. The total is 94.95%, so the house edge is 5.05%.

The results that pay nothing add zero to the sum, so they are left out, but they are why the total is below the whole stake.

What the Edge Costs per Game

An edge of a few percentage points sounds small until you apply it to every game. The table gives, for each spot count, the average number of catches, the average prize on the games that pay, the typical swing of one game and the edge. The swing is one standard deviation of the chips returned by a 1-chip ticket.

SpotsAverage catchesAverage prize when paid (chips)Typical swing per game (chips)House edge
10.253.001.3025.00%
20.5015.003.579.81%
30.756.185.625.65%
41.003.665.635.22%
51.259.8221.335.05%
61.505.8819.335.01%
71.754.0136.475.08%
82.009.2730.145.10%
92.256.1229.696.40%
102.5014.4117.666.80%

On a 1-chip ticket the expected loss per game is the edge expressed in chips, so 5.01% at 6 spots means about 0.05 chips a game. The swing is far larger than that, which is why a short run can end well ahead or well behind while the long-run edge stays the same.

Why the Edge Varies by Spot Count

Nothing in the draw makes one spot count better. The differences come from the paytable alone. With 1 spot the table pays 3 chips for a chance of 25%, a return of 75.00%, and the edge is 25.00%. From 3 spots upward the edge sits between 5.01% and 6.80%, and the sample table is least generous at 9 and 10 spots, where it returns 93.60% and 93.20%.

The lesson is general. The people who write a paytable choose each prize one cell at a time, and a small change in one rare cell, such as the top prize, moves the return by whole percentage points. That is why two games with the same spot count can have different returns.

What the Surveys Found

A gambling-maths reference reports three surveys of real games, shown in the table beside our sample.

GameRange of returnsSource
Live keno65% to 80%Survey in Las Vegas in 2001
Live keno50% to 74%Survey in Laughlin in 2012
Video keno84% to 95%Survey in San Diego in 2008
Draw of 20 from 8075.00% to 94.99%Our sample table

In words, keno in Las Vegas in 2001 returned between 65% and 80%, keno in Laughlin in 2012 returned between 50% and 74%, and video keno in San Diego in 2008 returned between 84% and 95%. The same reference also says that legitimate games are fair in the sense that every ball has a 1 in 80 chance of being drawn each game, and that the only choice that matters is where to play and how many numbers to choose.

An encyclopaedia entry gives a wider range for the house edge, with under 4% to over 35% for online keno and 20% to 40% for in-person keno. Set beside those figures, our sample table, with returns from 75.00% to 94.99%, is kinder than the live games in the surveys and close to the top of the video range. That makes it a fair model for a game that returns a high share, and a poor guide to a game that does not.

Why a Survey and a Model Can Differ

A survey measures the tables that particular casinos chose at a particular time, and a model measures one table in full. The two agree only if the table is the same, so a game that returns less than our sample is not misbehaving; it simply uses a different table. The reverse also holds. A game that returns more than our sample is making its prizes richer somewhere, and the paytable shows where.

The edge is also not the whole price. A live game that draws once every few minutes charges the edge on far fewer chips an hour than a machine that draws every few seconds, which the strategy and budget guide prices.

How Often Anything Pays

The return is an average, and the chance that a ticket pays anything at all is a separate fact. Across our table the chance of any prize runs from 6.01% with 2 spots to 25.89% with 4 spots, and the figures for 5 and 10 spots are 9.67% and 6.47%. Overall odds are published by lotteries too: a Virginia Lottery page lists overall odds of winning a prize from 1 in 2.8 to 1 in 16.6.

A high chance of some prize does not mean a high return, because the prize may be as small as the stake. The strategy and budget guide turns these chances into the cost and swing of an hour.

Checking a Real Game

Four checks tell you how a game compares with this model.

  1. Find the pool and the draw. Our figures assume 20 from 80.
  2. Read the paytable for your spot count, and find any printed return to player.
  3. Check for features that change the price or the prize, such as a multiplier or a pooled prize.
  4. Compare the listed return with the 75.00% to 94.99% of our sample, and treat a lower one as a higher price.

If the game does not print a return, the table still lets you compute it with the chances in the spots and catches guide.

The whole series is listed at the keno hub, and the pages below sit beside this one.

An Edge Is Not a Forecast

The house edge is a long-run average and says nothing about the next game. The safer gambling page lists the tools for setting limits and the places to turn if play stops being a choice.

❓ Keno Odds and RTP: Frequently Asked Questions

What is the house edge in keno?

Our sample table puts the edge between 5.01% at 6 spots and 25.00% at 1 spot. Surveys of real games found far wider ranges, with returns from 50% in one live game up to 95% in a video game. The edge depends on the table, not on the draw. It depends on the table.

What is a good RTP for keno?

Higher is better for the player, and our sample table reaches 94.99% at 6 spots. A gambling-maths reference found video keno returns of 84% to 95% in one survey and live returns of 65% to 80% in another. A game that does not print its RTP can still be priced from its paytable. Higher is better, and checkable.

How do you calculate the RTP of a keno game?

Multiply the chance of each paying result by its prize and add the results. On a 5-spot ticket the three paying rows contribute 25.18%, 15.72% and 54.04% of the stake, which sum to 94.95%. The chances come from the draw and the prizes from the paytable. Chance times prize, added up.

Can keno odds change from one draw to the next?

No. Every draw uses the same pool, and every number has a 25% chance each time. The odds of a ticket depend only on its spot count and the paytable, not on what was drawn before. A previous result tells you nothing about the next one. Each draw starts fresh.

Author:

Lucas Portela

Lucas Portela

Owner, BoldGambler • Avanhandava/SP

Oddsmaker, affiliate and content creator in the iGaming industry.