I'll say it straight: I don't like where online MTTs are heading. First PKO, then Mystery Bounty, now Ocean KO. Every new format piles more randomness on top of a game that's already random enough, and at some point poker as a mathematical model is gone and what's left is gambling. Ocean, in my view, is the last nail in the coffin.

What annoys me separately is how many schools and staking teams moved away from the classics: long guaranteed tournaments, a normal structure. Why is clear enough. PKO and bounties bring money back fast. You bust two or three players, your buy-in is already back, makeup is smaller, everyone's comfortable. But a player has to learn the classics first. That's the basis of any teaching: Aristotle first, then Hegel, never the other way round. Learning bounty tournaments without understanding the depth of regular MTTs, main events and multi-day tournaments misses the point.

An example from my own live PKO. Final table, I have a big stack, I cover three players, and they have big bounties. I have KTo, and the solver says I have to go all in. I go. Against me: JJ, QQ and AT. K10 has about 18% equity in that four-way pot. A king came, I took all three.

By the solver it's the right decision. But without a foundation, the only lesson you take from a hand like that is "shove K10, get lucky". I think that rots the basics of the game.

Okay, enough emotions. Now the numbers, all from public sources.

Ocean KO multiplier table for levels 1-10 (PokerNews, pokerfuse):
400x 0.002%
100x 0.08%
10x 2.4%
2x 2.5%
1.5x 2.7%
1x 20%
0.7x 22%
0.5x 26.2%
0.4x 24.1%
Add it up and 72% of knockouts pay less than their average value. The average is still exactly 1: on average you get the same as in a regular PKO.

Almost a third of the bounty value (32.8% by pokerfuse's count) sits in multipliers of 10x and up. For Winamax Space KO it's 15%, for Betclic Smash KO 10%. GG built the wildest version of the three.

I worked out the standard deviation of a single knockout from these tables, in units of its average value:
Ocean KO 3.63
Space KO 3.28
Smash KO 1.81
In a regular PKO it's zero. You bust someone, you get exactly your half of the bounty.

Now in buy-ins, which is the interesting part. The starting bounty is half the buy-in, and an average regular gets about two knockouts per tournament (aJarov's figure from a GipsyTeam article). Then the multipliers alone add a spread of about 2.6 buy-ins to each tournament result. For scale: in PokerStrategy's PrimeDope breakdown, a 100-player MTT at 15% ROI had a standard deviation of $1720 over 2000 tournaments at an $11 buy-in, which is about 3.5 buy-ins per tournament.

So, honestly, the noise from the multipliers alone is almost as big as the whole spread of a regular 100-player tournament. Put them together: 3.5 and 2.6 give about 4.3 buy-ins per tournament, and the variance goes up roughly one and a half times.

What that means for you and me. To tell a 10% ROI from zero with about 95% confidence you need roughly (2 × σ / 0.1)² tournaments. For a regular 100-player tournament that's about 4900. For the same tournament with Ocean multipliers, about 7500. Same skills, and it takes one and a half times longer to prove them to yourself.

And the bankroll, the question I started with under the GipsyTeam news. The cushion you need grows with variance divided by ROI. Variance went up one and a half times, so you need about one and a half times the bankroll. In the same PokerStrategy example, a player with 15% ROI in 100-player fields needed 200 buy-ins to survive the worst downswing. For Ocean with the same numbers it's about 300. And that's before you remember that Ocean fields aren't a hundred people, they're tens of thousands.

Then the jackpots. In a Sunday Ocean with 23,443 entries, the chance that a 400x hits anyone at all is about 37%. Anyone, meaning one person out of twenty-three thousand.

And they snowball. Half of a knockout, the jackpot ones included, doesn't go to your cashier, it goes onto your own head. The article's example: in a $10 tournament you bust someone for 100x and become a "Whale", and your own head is now worth $255 in expectation. Sooner or later someone busts you, and one time in forty that's another 10x. In Sunday Ocean fields a chain like that happens in roughly every second tournament.

There's a funny paradox too. A bounty becomes "legendary" once it passes $10,000 and 50 buy-ins. Legendary levels have lower multipliers, so the max ends up around $500,000. But a $9,999 bounty isn't legendary yet, and in theory it can still roll a 400x, almost $4 million. The fattest head at the table isn't the most expensive one. It's the one that fell just short.

Strategy is more interesting than it looks. The average multiplier is 1, so while there are no big tokens in play the math is the same as a regular PKO, and Lorem says in the article straight out that the knockouts are capped from above. Everything changes once someone carries a big head: people start calling for it with hands that would be an easy fold in a normal PKO. And that's exactly where Chris Brewer worries about collusion, with friends, backers and stables helping "their guy" reach a fat token.

Who benefits is pretty clear. The $5 Ocean Opener drew over 420 thousand entries, a world record for an online tournament. Recreational players like it, the site likes it. Regulars are left playing longer to see their edge through the noise.

Honestly, I think this is exactly what GG wanted to build. And I even get it: money likes speed. Look at Macau. By Inside Asian Gaming's count from April 2026, there are 46 poker tables left out of 6000 gaming tables, 0.77%, and permanent poker rooms exist in only two casinos, the Venetian and MGM Cotai. The Wynn Macau poker room closed. Operators say outright that capped tables are better used for baccarat, and that poker is "a much less profitable game" for them. Baccarat runs 60-100 hands an hour, and the casino keeps about 1% of every bet. From poker it only gets the rake.

The one who runs the game always wins. Ocean is no exception. According to pokerfuse, the $5 Opener had 420,084 entries and $1,932,386 in prizes and bounties. That's $4.60 per entry. The other 40 cents, about 8%, look like the rake. Plus traffic and recreational players having fun.

For fields like these I'd count a bankroll not in hundreds but in a thousand buy-ins or more. Even in the PokerStrategy 100-player example the worst downswing was 138 buy-ins. In fields of tens of thousands it's several times bigger, and the multipliers shake it more on top.

To be fair: mathematically it's still not a slot machine. A knockout is worth the same on average, and a strong player keeps his edge. But that edge drowns in the noise, and the whole format is built around one feeling: what if a 400x drops right now.

My view is simple: classic freezeouts should stay. No lottery on top.

Anyone who learned MTTs straight on PKO and bounties: did you move to regular tournaments later? What felt the most unfamiliar there?