A $1/$2 no-limit hold'em grinder who wins 5 bb/100 before rake keeps roughly 4.55 bb/100 after it — a 9% haircut that, over 40,000 hands a month, costs about $360 in pure profit. The rake cap is where that money goes, and the cap is set by the room, not the pot. Most US-facing low-stakes games cap at $3 plus $1 for the bad-beat jackpot at $1/$2, and that ceiling hits far more often than the number suggests.

The cap isn't a ceiling — it's a floor on your losses

Players picture the cap as protection: the house can only take so much. True, but the cap is reached at a pot of roughly $60 in a standard 5% structure, and at $1/$2 that's a routine hand. Any pot that gets to a flop, a turn bet, and a river call is already capped. In practice, the majority of hands that go past the flop at these stakes are max-rake hands, which means the rake is effectively a flat fee, not a percentage. A flat fee scales down badly as your win rate shrinks.

Run the math on a 5 bb/100 winner. At $1/$2, 5 bb/100 is $10 per 100 hands. A $3 max rake paid on, say, 28% of hands works out to $84 per 100 hands in rake — but only the portion attributable to your pots matters, and a winning player is in roughly a third of capped pots. That still lands the effective cost near 9% of gross winnings. A 2 bb/100 winner, which is a realistic long-run result for a competent low-stakes regular, loses closer to 22% of gross profit to rake. The rake doesn't care that you're better than the field; it takes the same $3 whether you win $300 or $30.

Why the bad-beat drop makes it worse

The $1 jackpot drop is the quiet killer. It's collected on every raked pot, capped or not, and it's not returned unless the jackpot hits — a variance event with a payout schedule that favors the room's float. At 30 hands an hour, four tables, that's 120 drops an hour, or $120/hour leaving the table in jackpot money alone. Over a 40-hour month, $4,800 exits the player pool. Some of it comes back in a jackpot, most doesn't. Rooms that dropped the bad-beat rake during the 2020–2021 online boom saw low-stakes win rates rise measurably; the reverse is equally true.

What a 9% haircut actually changes

Nine percent sounds survivable until you compound it. A player who could beat $1/$2 for 5 bb/100 is, after rake, a 4.55 bb/100 player — which means the bankroll required to survive normal variance grows by about 10%, and the time to move up to $2/$5 extends by months. That's the real cost: not the dollars, but the delay. Grinders who never escape the rake cap at low stakes aren't usually bad players. They're players whose edge is smaller than the fee.

The structural question is whether any US-regulated room has an incentive to lower the cap. Rake is the product. A room that cut its $1/$2 cap to $2 would attract volume, but volume at low stakes is exactly what's least profitable per table-hour. So the cap stays, and the 9% stays with it — which raises the uncomfortable possibility that the modern low-stakes ladder isn't a ladder at all, just a treadmill with a $3 cover charge.