Poker rake caps at $3—the micro table pays it twice

A $3 rake cap looks like a ceiling, not a cost. But at a 1c/2c table on a major site, a 200bb pot with a $40 pot going to the flop can generate the full $3 in rake — and the same table will pay that cap several times an hour. The micro-stakes player isn't dodging the cap. They're hitting it more often than the $1/$2 regular sitting two tables over.

That's the part most rake discussions skip. Everyone knows caps exist. Fewer people model what a cap does to a 100bb stack at 2c/5c, where a 20c raise preflop and two callers already puts 60c in the middle before anyone sees a flop.

The cap isn't the number that matters — the pot-to-cap ratio is

A $3 cap on a $1/$2 table is generous. You need a $60 pot before the house takes its full slice, and $60 pots at $1/$2 are uncommon enough that the cap rarely binds. The same $3 cap at 1c/2c is a different animal. A $60 pot at 1c/2c is 3,000 big blinds. That doesn't happen. What does happen is a $6 pot — 300bb — which is entirely reachable when three players see a flop and one barrels two streets.

Here's the arithmetic that matters. On a typical no-limit hold'em cash table, rake is charged as a percentage of the pot, usually 5%, with the cap kicking in once the pot crosses a threshold. At 5% and a $3 cap, the cap binds at a $60 pot. At 1c/2c, a $60 pot is 30 buy-ins. At 5c/10c it's six buy-ins. At 25c/50c it's 1.2 buy-ins. At $1/$2 it's 0.3 buy-ins.

So the cap binds most often at the stakes where players have the least room to absorb it. A micro player who wins a 400bb pot — $8 at 1c/2c — pays $3 in rake on an $8 pot. That's 37.5% of the pot. The same hand at $1/$2 would be a $400 pot, rake capped at $3, or 0.75%.

This isn't a rounding error. It's the dominant cost at the bottom of the ladder.

Worked example: one orbit at 1c/2c

Six-max, 100bb effective, 80 hands per hour. Assume a modest 25% of hands see a flop and the average raked pot is $1.20 — that's 60bb, which is realistic when you account for the many small pots and the occasional larger one.

• 80 hands/hour × 25% = 20 raked pots
• 20 × $1.20 = $24 in total pot value
• 5% rake = $1.20 in rake per hour

That's roughly 60bb/100 in rake paid. Now compare to a $1/$2 table with the same 20 raked pots but an average pot of $18:

• 20 × $18 = $360 in total pot value
• 5% rake = $18, but capped at $3 per pot
• 20 × $3 = $60 in rake per hour

At $1/$2, that's 30bb/100. The micro player is paying double the big-blind rate for the same number of hands. The cap didn't protect them — it just stopped the bleeding at a level they hit constantly.

Why the cap hits micro players twice

There are two separate mechanisms, and they compound.

The first is frequency. The cap binds whenever the pot crosses the threshold. At 1c/2c, that threshold is low in absolute terms — a $60 pot is rare, but a $6 pot isn't, and at 5% the rake on a $6 pot is 30c, not $3. Wait — that's the trap. The cap doesn't bind at $6. It binds at $60. So the micro player isn't hitting the cap on every hand. They're hitting it on the hands that matter most: the big ones.

The second mechanism is the one that actually stings. When you win a large pot at micro stakes, the rake takes a disproportionate share of your win. A 500bb pot at 1c/2c is $10. Rake is $3. You collect $7. That's a 30% tax on your biggest win of the session. At $1/$2, a 500bb pot is $1,000, rake is $3, and you collect $997. The tax is 0.3%.

The psychological effect is worse than the mathematical one. Micro players learn to avoid marginal spots because the rake eats the edge. They fold hands that would be profitable at higher stakes, not because the opponents are tougher, but because the rake structure punishes thin value.

The break-even threshold shifts

If you're a winning micro player with a 5bb/100 edge before rake, and rake is 60bb/100, you're losing 55bb/100. The rake isn't a drag on your win rate — it's the entire result. You need a pre-rake edge of at least 60bb/100 just to break even.

At $1/$2, with rake at 30bb/100, a 5bb/100 pre-rake edge still loses, but the gap is smaller. A 35bb/100 pre-rake edge breaks even. That's a meaningful difference in what kind of player can survive.

This is why micro-stakes regs often have higher pre-rake win rates than mid-stakes regs and still lose money. They're not worse players. They're playing a game with a worse cost structure.

What the sites won't tell you about the cap

Rake caps are marketed as player-friendly. "We cap our rake at $3" sounds like a concession. In practice, the cap is a pricing mechanism that scales inversely with stake size. The lower the stakes, the more often the cap is the actual price you pay.

Some sites have started offering reduced caps at micro stakes — $1 or $1.50 instead of $3. That helps, but it doesn't fix the ratio problem. A $1 cap at 1c/2c still binds at a $20 pot, which is 1,000bb. Still rare. The real fix would be a percentage-based rake with no cap at micro stakes, or a cap that scales with the big blind. Neither is common.

What you can do as a player is track your own rake paid per 100 hands. Most poker trackers will do this if you tag the stat. If you're paying more than 40bb/100 in rake at 1c/2c, you're in the range where the cap is actively working against you, and moving up to 5c/10c — where the same cap represents a smaller share of the pot — might improve your effective win rate even against tougher opponents.

That's the counterintuitive part. Moving up in stakes can reduce your rake burden in relative terms, because the cap stays fixed while pot sizes grow. The micro table isn't cheap. It's just cheap-looking.

The open question is whether sites will ever price rake by stake level rather than applying a flat cap across the board. Until they do, the $3 cap will keep doing what it's always done: protecting the house from variance at the top and extracting a premium at the bottom. If you're grinding 1c/2c for profit rather than practice, the cap isn't your friend. It's the reason your graph looks flat.