Plinko looks like a toy — a ball dropping through a triangle of pegs into one of the slots along the bottom. Underneath it sits a binomial distribution, and that distribution is the entire explanation for why the outer slots carry the largest multipliers while the middle ones hand back a fraction of the stake. Here is where those numbers come from and what the risk and row controls actually do.
How a slot gets chosen
At every peg row the ball goes left or right. With n rows a drop is a sequence of n binary decisions, and the landing slot depends only on how many of them went right. That gives n + 1 slots, and the chance of landing in slot i is C(n, i) / 2^n.
Hence the shape of the board. An outer slot is reachable by exactly one path, the middle slot by thousands. On a 16-row board there is one path out of 65,536 to the far edge and 12,870 paths to the centre, close to 20% of all drops.
Why the edges pay more
Multipliers are assigned so that the sum of probability × payout across every slot equals the stated RTP. Plinko on Tonza is set at 97%. The far slot on a 16-row board carries a 0.0015% chance, so even a very large multiplier contributes only a sliver of the total return — the payout can be that big precisely because almost nobody lands there.
The flip side is that the slots you hit constantly pay under 1x. That is not a malfunction, it is the design. Most drops return part of the stake and the entire upside lives in the tails of the distribution.
What rows and risk levels change
The row count sets how wide the distribution is. Three board sizes as an example; the probabilities follow from the formula above and owe nothing to the operator.
| Rows | Slots | Edge slot odds | Centre slot odds |
|---|---|---|---|
| 8 | 9 | 1 in 256 | 27.3% |
| 12 | 13 | 1 in 4,096 | 22.6% |
| 16 | 17 | 1 in 65,536 | 19.6% |
More rows means more slots and a far thinner edge probability, which is what allows the top multiplier to be raised. Fewer rows compress the distribution, rare outcomes become less rare, and the ceiling comes down. The stated return does not move; only the shape of the payouts does.
Low, medium and high rewrite the payout table while leaving the probabilities alone. On low risk the central slots return close to the stake and the edges stay modest, so results cluster in a narrow band. On high risk the centre drops well below 1x and the edges climb steeply.
The trade-off can be checked arithmetically: raising the payout of a rare slot forces a cut somewhere frequent, or the sum stops converging on 97%. No risk setting has a better expectation than another. They differ in variance, and therefore in how fast a bankroll swings.
The animation is not the game
The falling ball renders a result that already exists. Under provably fair each left or right step is derived from the server seed, the client seed and a nonce, so the path is fixed before the first frame is drawn. Timing, clicking and drop speed have no influence on it — the provably fair page shows how to recompute a round once the seed is revealed.
Variance versus the house edge
The practical takeaway is that Plinko's settings are a comfort dial rather than a return dial. Many rows on high risk produce long stretches of sub-stake returns broken by an occasional large hit; few rows on low risk turn a session into a near-flat line drifting down at the rate of the margin. The average return is the same either way, only its spread over time differs.
Stakes run from 0.1 to 10000 in your account currency, and Plinko sits with the other in-house titles in Tonza Originals. For a comparable risk profile with a live decision instead of a fixed drop, there are the crash games.
