Plinko odds by row, with a worked example board

Every row of pins sends the ball left or right with even odds, so most drops pile up in the middle and the edges stay rare. On 16 rows, four balls in five finish in the middle five slots, and a single edge slot gets hit once in 65,536 drops. The multipliers printed under the slots set the RTP. Stake lists 99% for its own version.

By the Beltastic team · Updated

Where the ball lands

A board with 16 rows has 17 slots and 65,536 possible paths. The far left slot can only be reached one way, by bouncing left at every pin. The slot next to it has 16 paths, because the single right bounce can happen on any of the 16 rows. The counts follow one line of Pascal's triangle, and dividing each count by 65,536 gives the chance.

Chance per slot on 16 rows (each side, edge to centre)
Slot from the edgePathsChanceAbout 1 in
Edge10.0015%65,536
2nd160.0244%4,096
3rd1200.1831%546
4th5600.8545%117
5th1,8202.7771%36
6th4,3686.6650%15
7th8,00812.2192%8
8th11,44017.4561%6
Centre12,87019.6381%5

Own calculation, C(16, k) ÷ 2^16. Every slot except the centre exists twice, once on each side.

How tightly the drops bunch up, by number of rows
RowsSlotsMiddle 3 slotsMiddle 5 slotsEither edge
8971.09%92.97%1 in 128
121361.23%85.40%1 in 2,048
161754.55%78.99%1 in 32,768

Own calculation from the binomial distribution with p = 0.5 per row.

More rows don't spread the ball as much as you'd think. Going from 8 to 16 rows doubles the number of slots, yet the middle five still catch almost 80% of drops. What really changes is how far out the edges sit, and that's where casinos park the big multipliers.

Our example board and its RTP

We made up the board below to show the method. It isn't any casino's table. Each multiplier gets weighted by how often its slots are hit, and the weighted values added together give the RTP. This one comes out at 98.99%, and nearly a quarter of that return comes from the two x1 slots that just hand your stake back.

Invented 16-row board, slot pairs from edge to centre
MultiplierChance (both sides)Adds to RTP
x1100.0031%0.34 points
x410.0488%2.00 points
x100.3662%3.66 points
x51.7090%8.55 points
x35.5542%16.66 points
x1.513.3301%20.00 points
x124.4385%24.44 points
x0.534.9121%17.46 points
x0.3 (centre)19.6381%5.89 points
Total100%98.99%

Invented numbers for illustration. On this board 54.55% of drops return less than the bet.

You can run the same check on a real board in a couple of minutes. Write down the multipliers for your rows and risk setting from the game screen, multiply each one by the chance in the first table, and add them up. If you don't land close to the RTP the game states, you've copied a number wrong or the board isn't what it claims. The Plinko calculator does the sums for any board you type in.

What a 10,000x edge is worth

Rows
8 to 16
Risk levels
Easy, Medium, Hard, Expert
Highest multiplier
10,000x, on Expert with 16 rows
Lowest on Expert
0.1x
RTP
99% (house edge 1%)

An edge slot on 16 rows comes up once in 65,536 drops. If both edges pay x10,000, those two slots alone account for 30.52 points of the RTP, so the other 15 slots have to share what is left. That explains a board whose low end goes down to 0.1x.

Over 1,000 drops you have a 3.01% chance of seeing either edge even once. Stretch it to 10,000 drops and it's still only 26.30%. Most people who play the top risk setting for an evening never see the number that decides how the board is priced.

Strategy, simulators and predictors

Where you drop from doesn't matter on a seeded Plinko game, because the seeds settle the path before the ball even appears on screen. Each row is one coin flip read from the hash, 16 of them for a full board. Timing your clicks, waiting for a cold streak or dropping several balls at once changes nothing.

A simulator is fine for seeing how bumpy a setting feels before you put money on it. A predictor is a different thing. Anything that promises to tell you the next landing slot either has your server seed, which it shouldn't, or is lying, so don't hand over money, logins or seeds for one. Past drops are a different story: once you rotate, the revealed seeds go into a provably fair verifier and the 16 bounces come back in order.

18+. Plinko has a built-in house edge, and no row count or risk level removes it. Set a budget before you play. Help: Responsible gaming. Not affiliated with Stake. Names belong to their owners.

Questions

How likely is each Plinko slot?

Each row is a 50/50 bounce, so a slot's chance is its number of paths divided by 2 to the power of the rows. On 16 rows, 19.64% of drops land in the centre slot, while each edge slot gets one in 65,536.

Which RTP does Stake Plinko have?

Stake gives its Plinko a 99% return, so the house keeps 1 cent per dollar on average. Other sites price their boards differently. Check the rules there first.

Does the number of rows change the RTP?

Not when the game shows a single RTP for all settings. Rows and risk only move the return around the board, so losses get more or less frequent and the rare wins grow or shrink.

Is there a Plinko strategy that works?

No. The seeds decide the path, and each row is an even split. You can choose how bumpy the ride is, through rows, risk and bet size, but not the average result.

Do Plinko predictors or hacks work?

No. Nobody sees a drop's seed unhashed until you swap seeds, and by then the drop is history. Anything sold as a Plinko hack is a scam.

Sources

  1. Stake: Plinko game page (Wayback, 3 Jul 2026) · current page on stake.com. 99% RTP, 1% house edge, rows 8 to 16, risk Easy to Expert, 10,000x max on Expert 16 rows, Expert minimum 0.1x. Live page returns 403 to our fetcher. Checked 12 November 2026
  2. Stake: provably fair game events (Wayback, 19 Jul 2025). Plinko: one left or right decision per row, float × 2 Checked 12 November 2026
  3. Stake: provably fair implementation (Wayback, 26 Sep 2026). Plinko uses 2 cursor increments per game to cover 16 decisions Checked 12 November 2026

18+ · Play responsibly · gamblingtherapy.org