Ask keno players which pick count “pays best” and you get strong opinions: one or two numbers for steady hits, ten for the big multipliers. Both camps are usually arguing about the wrong thing.
Return to player, or RTP, is the share of every dollar wagered that a game pays back over the long run. In keno, that figure is set by the paytable, and each pick count has its own paytable. So the honest answer to “which pick count has the best RTP” depends entirely on how those tables were built. On a modern crypto board, the answer is often “none of them,” and the more useful question becomes what the pick count does to everything else.
Below, the math runs on a 40-number board that draws 10, using openly published paytables plus a few hypothetical edits. Probabilities are exact; session figures come from 20,000 simulated runs.
Where Pick-Count RTP Differences Come From
Traditional casino keno uses an 80-number board with 20 numbers drawn. Each spot count gets its own paytable, and operators calibrate each one separately. Published estimates put the house edge in land-based keno somewhere between 20% and 40%, and it can shift by several points from one spot count to the next. That is why old guides compare 4-spot and 7-spot tickets so closely.
Online and crypto versions changed the design brief. Many operators now publish a single RTP for the whole game and then build every pick count’s paytable to land on that number. When that is the case, the pick count stops being a lever on expected return. A well-run bitcoin casino keno table that advertises 98% should return close to 98% whether you pick one number or ten. Rounding multipliers to one decimal means no table hits its target exactly.
The Hit Odds Behind Every Pick Count
Before any paytable comes into it, the board itself fixes the odds. With 40 numbers and 10 drawn, each number you pick has a 1 in 4 chance of landing. Your average number of hits is simply your pick count divided by four: 0.25 hits on one pick, 1.25 on five, 2.5 on ten.
The chance of catching at least one number climbs fast as you add picks:
- 1 pick: 25%
- 3 picks: 59%
- 5 picks: 78%
- 7 picks: 89%
- 10 picks: 96.5%
The chance of hitting every number you picked collapses even faster. One pick is 1 in 4. Five picks is 1 in 2,611. Ten picks is 1 in 847,660,528. None of this says anything about RTP yet. RTP is the sum of probability times payout across every paytable row.
Three Published Tables, Three Near-Identical Returns
Here is what that sum gives for three paytables published on a live bitcoin keno page that states a 98% RTP:
- Ten picks, Low risk (1.1x for 2 hits up to 1,000x for 10): 97.93%
- Ten picks, Medium risk (1.5x for 3 hits up to 1,000x for 10): 97.97%
- Five picks, Low risk (1.5x for 2 hits up to 300x for 5): 97.95%
The spread between them is 0.04 percentage points. Over $10,000 of total wagers, that is a difference of about four dollars in expected return. Pick count and risk setting do not move RTP here; the gap to 98% is rounding.
The same arithmetic tells you what a fair one-pick table would look like. One pick hits 25% of the time, so a 98% table has to pay 3.92x on a hit. If a one-pick ticket somewhere pays 3x instead, it returns 75%.
Same 98%, Different Path: Hit Rate and Swing by Pick Count
If RTP stays flat, pick count changes two things: how often you get paid and how hard the balance swings.
- Ten picks, Low risk: pays something on 79.6% of rounds; standard deviation 0.69x stake per round
- Ten picks, Medium risk: pays on 48.5% of rounds; standard deviation 1.59x
- One pick at a fair 3.92x: pays on 25% of rounds; standard deviation 1.70x
- Five picks, Low risk: pays on 36.7% of rounds; standard deviation 6.08x
That last line is the surprise. Five picks on this table is almost nine times as swingy as ten picks on Low, even though both return the same 98 cents per dollar. More picks does not mean more volatility; the paytable’s shape decides it.
Which Paytable Rows Carry the RTP at Each Count
Breaking each table’s return down row by row shows where the money comes from.
On ten picks at Low risk, the 2-, 3- and 4-hit rows deliver about 90% of the entire return. The 1,000x prize for ten hits contributes roughly 0.0001 cents per dollar, close enough to zero that removing it would not change the RTP at the second decimal. Medium risk on ten picks moves the weight up a little: the 3- and 4-hit rows carry 74%, and the 5-hit row adds another 17%.
Five picks is different. The 2-hit row carries 42.5% of the return and the 3-hit row 34%, but the 300x prize for five hits carries 11.7% on its own, the same share as the 12x row below it. On a five-pick bitcoin keno game, the top prize is a real part of what you are being paid back. On ten picks, it is decoration.
One Shaved Row Can Cost More Than the Jackpot
Knowing which rows carry the return shows where a stingy operator could cut quietly. Two hypothetical edits:
On ten picks at Low risk, cutting the 2-hit payout from 1.1x to 1.0x would lower RTP by 3.1 percentage points, from about 98% to about 95%. Halving the 1,000x top prize to 500x would lower it by less than one ten-thousandth of a point. The jackpot is the row players look at, and the one that barely matters.
On five picks, the picture flips. Halving the 300x prize to 150x would cost 5.7 points of RTP, which is more than the 2.8 points lost by trimming the 2-hit row from 1.5x to 1.4x. So for high counts, check the small rows; for low and middle counts, check the top prize too.
What 500 Rounds Look Like at One, Five and Ten Picks
Here is 500 rounds at $1, across 20,000 simulated sessions per table. Expected loss in every case is about $10.
- Ten picks, Low risk: median result -$11.30; 20.6% of sessions finished ahead; the middle 80% ranged from -$28 to +$7
- Ten picks, Medium risk: median -$13.50; 34.2% finished ahead; middle 80% from -$53 to +$34
- One pick at 3.92x: median -$10.00; 39.6% finished ahead; middle 80% from -$57 to +$37
- Five picks, Low risk: median -$59.00; 20.4% finished ahead; middle 80% from -$110 to +$228
Same RTP, very different evenings. Five picks on Low produces the worst typical session and the best lucky one, because so much of its return sits in a prize that turns up once every 2,611 rounds on average. Ten picks on Low is the slow grind. With crypto keno, where a round settles in seconds, 500 rounds can pass quickly, so choose the session shape you want before starting.
How to Calculate RTP for Any Pick Count Yourself
A spreadsheet does this in minutes.
- Write down the board size (N), the numbers drawn (D) and your pick count (k).
- For each possible hit count h from 0 to k, compute the probability: COMBIN(D, h) x COMBIN(N – D, k – h) / COMBIN(N, k).
- Multiply each probability by the paytable multiplier for that row.
- Add the results. That total is the RTP for that pick count.
Check that your probabilities sum to 1 before trusting the answer. If totals across pick counts differ by more than a few tenths of a point, some counts are priced worse; avoid those.
On provably fair games, you can also verify each draw against the published seeds.
A Pick-Count Checklist Before You Play
- Confirm the RTP for the exact pick count and risk level you will use, not just the game’s headline figure.
- Expect average hits to equal picks divided by four on a 40/10 board; anything you read that implies more is wrong.
- Choose the count by session shape: frequent small returns, or rare large ones.
- On low and middle counts, check the top prize; on high counts, check the small rows.
- Size stakes so a 500-round stretch at the 10th percentile does not hurt.
To see how the published paytables look in practice, you can try keno at Jacks Club and run the spreadsheet check against the tables shown on the game page. The numbers should agree to within rounding.
Keno is a game of chance, and no pick count changes the long-run edge. Play only with money you can afford to lose. 18+ only; check that online gambling is legal where you live.