Keno guide
Cosmos Keno: six numbers, seven possible outcomes
The name Cosmos is decoration for this page and nothing else. Below we look at a six-number ticket in 40/10 keno, then run a seeded simulation with a made-up pay schedule.
Where a six-pick ticket usually lands
With six picks the expected hit count rises to 1.5. The distribution peaks at one hit (37.13%), with two hits (32.13%) a clear second. Zero hits still occurs in about one game in 6.5.
| Hits | Probability | As odds |
|---|---|---|
| 6 | 0.0055% | 1 in 18,278.0 |
| 5 | 0.1970% | 1 in 507.7 |
| 4 | 2.3799% | 1 in 42.0 |
| 3 | 12.6929% | 1 in 7.9 |
| 2 | 32.1288% | 1 in 3.1 |
| 1 | 37.1266% | 1 in 2.7 |
| 0 | 15.4694% | 1 in 6.5 |
A hand calculation: three hits
Three hits on six picks means three numbers from the drawn 10 and three from the undrawn 30. The drawn side gives C(10, 3) = 120. The undrawn side gives C(30, 3) = 4,060. The product is 487,200. The total number of six-number tickets is C(40, 6) = 3,838,380, so P(3 hits) = 12.69%.
For the bottom line, C(30, 6) = 593,775 tickets miss entirely. As a share of 3,838,380, that is the 15.47% above. Each extra pick multiplies the miss count by a smaller fraction, which is why blank tickets fade slowly as you add numbers.
The upper tail
Five hits come up 1 in 507.7 games and a full six-of-six 1 in 18,278.0. Combined, four or more hits happen in 2.58% of games. Any pay schedule that puts most of its value there will have a long gap between meaningful returns.
A seeded simulation
To see that tail in motion, the simulation below plays six sessions of 150 games each with a hypothetical schedule: 0x for 0 to 2 hits, 2x for 3 hits, 10x for 4, 60x for 5 and 500x for 6. That schedule returns 63.74% of stakes on average. It is invented for teaching and is not any operator's table.
| Session | Rounds won | Net (units) | Lowest point | Longest losing run |
|---|---|---|---|---|
| 1 | 28 | -46.00 | -62.00 | 20 |
| 2 | 24 | -62.00 | -62.00 | 18 |
| 3 | 18 | -90.00 | -95.00 | 16 |
| 4 | 29 | -52.00 | -52.00 | 15 |
| 5 | 19 | -22.00 | -22.00 | 17 |
| 6 | 13 | -108.00 | -108.00 | 32 |
Seeded pseudo-random simulation (Python random.Random, seed 606), 6 sessions of 150 one-unit rounds. Average session result -63.33 units. This is a simulation, not a record of real play; a different seed gives different numbers.
This is a pseudo-random simulation, not a record of anyone's play. The sessions differ a lot, and the longest losing runs are long, because a game only counts as a win here with 3 or more hits, which happens 15.28% of the time. A different seed would give a different set of rows; the long-run average would not move.