The board is physics; the payouts are a choice
A ball drops through 16 rows of pegs, and at each peg it goes left or right on a coin flip. To reach a far edge it would have to flip the same way 16 times running, which is fantastically unlikely, while landing near the middle just needs a rough half-and-half, which happens almost every time. So where the ball ends up follows a bell curve. The centre slots swallow nearly all the balls, the outer slots almost none. That's fixed physics, the exact same odds at every casino.
What each casino chooses is the payout table, the row of multipliers printed under the slots, smallest in the crowded middle, largest out at the lonely edges. Since the table fixes the average return, the player's only real lever is which table they play, a pure choice of how wild a ride they want. So comparing casinos means comparing their tables, and the catch is always the same. The biggest multipliers sit exactly where the ball hardly ever lands.
The whole idea comes down to one picture. Every peg is a 50/50 coin flip, left or right, and a ball's trip to the bottom is just a run of those flips. There are loads of ways to mix lefts and rights and end up near the centre (70 different routes into the middle slot), but only one way to reach a far edge. The ball has to flip the same way every single time. So before a single multiplier is attached, the shape of the board alone crowds nearly all the balls into the middle. (The route counts are Pascal's triangle, each number is just the two above it added together.)
So how rare is a jackpot edge? On the real 16-row board the ball has to flip the same way sixteen times running, the same odds as tossing sixteen heads in a row, about 1 in 65,536. Drop one ball every second and you'd wait 18 hours to see that single edge. Put another way, of the 3.58 million real Plinko balls dropped across every casino last week (Fig. 01), only around 110 would have reached a jackpot edge at all.
A week of the real thing
This isn't a thought experiment. In the seven days to 7 July 2026, twenty-six crypto casinos took 3.58 million Plinko bets between them, about 511,000 a day, and the daily count barely moved all week. It's one of the most-played originals in the category.
The play is heavily lopsided. gamdom alone took 32% of every Plinko bet placed anywhere, and the top five casinos accounted for two-thirds of them. The other twenty-one share what's left.
We hold verified payout tables for seven of these casinos (marked in cyan below), enough to compare the actual games running behind all that volume, which is what the rest of this page does.
The money lives where the balls don't
The bars below show where balls actually land, ~79% pile into the five centre slots. The two lines show what each type of casino pays at each slot. Both give you back less than your stake right where the balls pile up, and hide the four- and five-figure multipliers out at the edges, where fewer than 1 ball in 30,000 ever reaches. (The payout scale up the right-hand side climbs in ×10 jumps, so those edge numbers are even more extreme than they look.)
"High risk" is really two risks
Rank the seven casinos by volatility, how wildly a single ball's result swings from the usual small loss to the rare big hit, and they don't spread out evenly. They clump into two. The 1,000× tables are all about equally bumpy, the 10,000× "Expert / Nightmare" tables are roughly eight times wilder. There is essentially nothing in between. At the top tier you're choosing between two completely different games, not sliding a dial from calm to wild.
The wildest ride doesn't cost the same everywhere
Plot each casino's volatility against its RTP, how much of every dollar it hands back over time, and the picture is clear. The standard-table casinos huddle together on the left, calm, and paying back ~99%. The two extreme tables stand alone far to the right. Mind the gap between them, though. Stake Expert pays back 99% for that lottery ride, while BC.Game Nightmare pays back only 98% for the exact same wild ride, meaning BC.Game keeps twice the cut Stake does for an identical thrill.
Four casinos, one identical price list
Lay the actual payout rows on top of each other and the standard tier collapses onto a single line: Roobet, Shuffle, Rainbet and Duelbits all use exactly the same multipliers, the original Stake "High" layout, now an industry default. Duel nudged it up a touch (130→131, 9→9.1) to reach a 99.9% base, and the two Expert / Nightmare tables tower about ten times higher than everyone else out at the edges.
Four balls in five are losers
Strip away the spectacle and look at what one ball actually does. On both types of table, ~79% of balls land in the losing middle, but the size of that loss differs. The standard table hands back 0.2× (you keep a fifth of your stake), the extreme table just 0.1× (a tenth). You pay for the 10× bigger jackpot with a deeper loss, not a more frequent one.
Drop a thousand balls and watch
Everything above, in one place. Pick a table, drop balls at $1 each, and watch two things happen at once: the balls pile into the bell exactly as the board forces them to, and your bankroll bleeds as roughly four in five land in the losing middle. Keep an eye on the jackpot-edge counter, it barely budges. Switch to the extreme table for a wilder ride (it starts the run fresh).
Sorted by how wild each table really is
The highest-risk Plinko table at 16 rows for each casino, exactly as displayed. RTP is the operator's stated figure; "net" shows Duel's rakeback perk. Hover the volatility column for detail.
Displayed multipliers are UI-rounded, so RTP computed from them is approximate; stated figures are authoritative. Duel's 100% is a daily-capped 0.1% rakeback over a 99.9% base table, not a property of the game.