Every row is one coin flip, and the board is just a stack of them
The mechanism, step by step
The ball does not choose a slot. It makes one left-or-right decision per row, and the slot is whatever those decisions add up to. On a 16-row board that is 16 decisions and 65,536 possible sequences, of which 12,870 finish in the centre slot and exactly one finishes at each edge.
| Operator | Licence | Coins listed | Studios listed | Amount before ID check | Countries barred | Go |
|---|---|---|---|---|---|---|
| Vaveour partner | Curaçao Gaming Authority · OGL/2024/1676/0905 · Active | 6 | Casino and sportsbook on one balance | Case by case, clause 8.7 | Open to this site’s readers | Visit |
| Vave is the partner of this site, and the one operator here you can look up in a register in under a minute: licence OGL/2024/1676/0905, held by Latcas B.V., issued 19 May 2025 and active. Six coins at the cashier — BTC, ETH, USDT, LTC, DOGE and TON — and a 150% welcome bonus up to 1.5 BTC with 100 free spins across the first four deposits. The link is advertising, and Vave takes no part in the counts on this site. | ||||||
| Wild.io | Curaçao Gaming Authority | 12 | 80 | no amount published | 45 | |
| Bitcasino.io | Curaçao Gaming Authority | not read | 120 | 2,500 EUR | 1 | |
| Rocketpot | Curacao | 13 | not read | US$2,500 | 34 | |
| Rainbet | Anjouan Gaming Board | 9 | 33 | no amount published | 11 | |
| Wolf.bet | Government of the Autonomous Island of Anjouan, Union of Comoros | 13 | 32 | no amount published | 48 | |
| DuckDice | Anjouan Gaming Board | 10 | 19 | no amount published | 16 | |
| Flush | Anjouan Gaming Board | 9 | 85 | not read | list not read | |
| Bitsler | Curaçao Gaming Authority | not read | 41 | no amount published | 1 | |
| BitStarz | Curaçao Gaming Authority | not read | 62 | no amount published | 74 | |
Vave
Wild.io
Bitcasino.io
Rocketpot
Rainbet
Wolf.bet
DuckDice
Bitsler
BitStarzThe ball never picks a slot
The most common way to misunderstand this game is to think of the slot as the target and the fall as an attempt to reach it. That is backwards. The slot is not chosen; it is calculated, one row at a time, and only exists as an outcome once the last row has been passed.
Here is the whole mechanism. The board is a triangle of pegs, arranged in rows. A ball enters at the top, directly above the apex. At the first peg it goes left or right. At the second row it meets another peg and again goes left or right. This repeats once per row. When it falls past the last row it drops into a slot, and which slot that is depends on nothing except how many of its decisions were rights.
Zero rights means the far-left slot. All rights means the far-right one. Everything else lands somewhere in between, in the position given by the count.
That is worth stating plainly because it disposes of a whole category of belief about this game in one line: the ball has no destination, and there is nothing along the way that steers it.
A board with n rows has n+1 slots and 2^n paths
Two numbers follow directly from the description above, and they are the only two you need.
A board with n rows gives the ball n decisions. Each decision has two outcomes, so the number of distinct decision sequences is 2 multiplied by itself n times. Eight rows gives 256 sequences. Twelve gives 4,096. Sixteen gives 65,536.
The number of slots is different and much smaller: n+1. Eight rows, nine slots. Sixteen rows, seventeen. That is because the slot depends only on how many rights the ball made and not on which rows they happened at, and the count of rights can be anything from zero to n.
So thousands of sequences map onto seventeen results, and they do not map evenly. That unevenness is the entire behaviour of the game, and how it changes with the row setting is worked through on the page about row counts.
Why the middle is crowded: it is a counting fact, not a tendency
There is exactly one way to make sixteen left decisions in a row. There is also exactly one way to make sixteen rights. But there are 12,870 different orders in which eight lefts and eight rights can occur, and every one of them lands in the same centre slot.
That number is not an estimate and not a simulation. It is the binomial coefficient "16 choose 8", the number of ways of picking which eight of the sixteen rows were the right-hand ones.
The row of those coefficients for a 16-row board runs 1, 16, 120, 560, 1,820, 4,368, 8,008, 11,440, 12,870, 11,440, 8,008, 4,368, 1,820, 560, 120, 16, 1. Add them up and you get 65,536, which is the total number of paths, as it must be. Divide any one of them by 65,536 and you have the probability of that slot. The complete set, for three different row counts, is tabulated on the odds page.
The practical consequence is that the crowding in the middle is not a tendency of the ball, a property of the pegs, or a bias in the software. It is arithmetic that would hold for beads, coins, or anything else making independent binary choices.
What the animation is, and what it is not
On a physical Galton board the bounces are genuinely physical, and no two runs are identical. On a gambling site the outcome is generated first, by a random number generator, and the falling ball you watch is a rendering of a result that already exists.
This matters for one reason only, and it is not a suspicious one: it means the question "is this fair?" is not a question about pegs at all. You cannot inspect the physics because there is no physics. What you can inspect, on the operators that offer it, is the seed mechanism — a hashed server seed published before the round, a client seed you supply, and a reveal afterwards that lets you recompute the result. That is a narrow, real guarantee, and its exact scope is set out on the provably fair page.
It also means the speed of the animation, the sound, and the visible bouncing are presentation. None of them are inputs.
The one thing that changes without touching the grid
If the pegs decide the slot and the slot probabilities are fixed by the row count, then the only remaining variable is what each slot pays. That is exactly what the risk control changes, and nothing else, which is why it deserves its own treatment on the page about risk levels.
The distinction is worth holding onto because a lot of writing about this game blurs it. A setting that changes the payouts and a setting that changes the probabilities would be two very different things. Only one of them exists here.
Where the extreme payouts sit on the board, and how often the ball reaches them, is on the multipliers page.
Nothing here is a way to play
It is worth being explicit, because the arithmetic on this site is precise and precision is easily mistaken for advantage. Knowing the exact distribution of a board tells you what shape the results will take. It does not tell you a good time to bet, a good number of rows to choose, or a good level of risk, and no page on this site will suggest one. The counting is here so that the picture in a reader's head matches the machine, not so that anyone can act on it.
Who builds these boards — the casino itself or an outside studio — turns out to be a readable fact where the operator publishes a studio list, and it is handled on the page about who makes the game. The rules governing everything printed on this site are on the methodology page.