Poker rooms in the United States have settled on 5% as the effective ceiling for cash-game rake, and that number is doing more damage to win rates than most regulars realize. At $1/$2 no-limit, a 5% rake capped at $3 per pot means a player winning 8 bb/100 before rake keeps closer to 3.5 bb/100 after it — and that's before the jackpot drop. The cap, not the percentage, is what separates a beatable game from a treadmill.

Where the 5% Actually Bites

The rake percentage matters far less than the cap structure once you're above micro stakes. A 5% uncapped rake is a different animal from 5% capped at $3, and both are different from the time collection used in some card rooms.

At $2/$5, a $5 cap at 5% gets hit on any pot over $100. In a typical live session with 30 hands per hour, that's roughly $90–$120 pulled off the table per hour across nine seats. Divide that by the number of players and you get the real number: somewhere between $10 and $13 per player per hour, win or lose.

Online, the same logic runs at higher hand volume. A 5% cap of $3 at $1/$2 six-max, with 85 hands per hour, extracts more per hour than the live equivalent — the cap gets reached more often because the hands move faster. The percentage is identical. The cost isn't.

The Cap Is the Lever

Operators know this. When a room raises its cap from $3 to $4 at $1/$2, it's not a 33% rake increase — it's closer to a 15–20% increase in total rake collected, because not every pot reaches the old cap. When a room lowers the cap from $5 to $4 at $2/$5, regulars notice within a week.

What Regulators Have Actually Done

State gaming commissions have been slow to touch rake caps directly. Nevada and New Jersey both allow operators to set their own rake within statutory maximums — New Jersey's Casino Control Commission caps rake at 10% or $5 per hand for poker, whichever is lower, but doesn't mandate the 5% most rooms use. Pennsylvania's Gaming Control Board sets a 10% cap with no dollar limit for online poker, which is why some PA-licensed rooms run caps above $3.

The 5% figure is a market convention, not a regulatory floor. That distinction matters: a room can raise it tomorrow without a rule change.

Why Regs Keep Firing Anyway

The counterargument is that recreational players don't calculate rake, and regs do. If a cap increase pushes the game from 3.5 bb/100 to 2 bb/100 for a winning player, the marginal reg leaves — but the recreational player who loses 15 bb/100 pre-rake doesn't notice the difference between losing 18 bb/100 and 20 bb/100.

Except the reg leaving changes the ecology. Fewer strong players means softer games, which means more recreational players stay longer, which means the room's total rake take can go up even as per-player rake on regs goes up. That's the trade operators are making.

The problem is the lag. A room that pushes the cap too far loses its regs faster than it replaces them with rec players, and the games get shorthanded and slow. By the time the room notices, the player pool has already thinned.

The Open Question

If 5% is a convention rather than a rule, what happens when a major operator tests 6% at $2/$5 with a $5 cap? The math says the rake increase is small enough that most regs won't leave immediately. The ecology says the ones who do leave are the ones the game needs most.

Watch the caps, not the percentages. The next rake fight won't be about the 5%.