Rake is how every cardroom earns money from a game where players compete against each other instead of the house. Without it there would be no incentive to run the tables, but every dollar taken out of a pot is a dollar that never returns to the players who put it in, which makes rake the single most reliable drag on a poker win rate that any regular grinder faces.
01How a percentage and a cap combine
The advertised number, something like 5 percent, only tells half the story. The cap decides where that percentage stops mattering. A room running 5 percent with a $3 cap takes five cents on every dollar up to a $60 pot, then flattens out entirely: a $200 pot and a $2,000 pot cost the same $3 in absolute terms, even though the second one is ten times larger. That single mechanic is why high-stakes players often care more about cap size than the headline percentage, since a low cap at higher stakes can mean an effective rake far below one percent of the average pot.
Many rooms also run a "no flop, no drop" rule, waiving rake on any hand that ends before the community cards are dealt. This rewards preflop-heavy, tighter styles slightly over loose, multiway games, since a hand that never sees a flop returns one hundred percent of the pot to the winner regardless of how the room's rake table is structured.
02Turning rake into a bb/100 number
The cleanest way to see the damage is in the same units used for a win rate: big blinds per hundred hands. Suppose a $1/$2 six-max game averages an $80 pot on hands that go to showdown or a raised pot, with rake landing near the cap on most of them, call it $3 a hand across roughly forty percent of hands dealt. Over a hundred hands that is 40 raked pots times $3, or $120 total, which at $2 per big blind works out to 60 big blinds per hundred hands paid in rake alone. A genuinely strong winning player might beat the game at 5 to 8 bb/100 before rake, so the house is taking a multiple of the player's actual edge out of every session.
That gap is exactly why a small edge can vanish once the accounting is done properly, and why a serious player should run their own numbers through a tool like the win rate converter instead of trusting a felt-level intuition about how much a session actually cost in rake.
| Pot size | 5% rake, no cap | 5% rake, $3 cap |
|---|---|---|
| $20 | $1.00 | $1.00 |
| $60 | $3.00 | $3.00 |
| $200 | $10.00 | $3.00 |
| $500 | $25.00 | $3.00 |
03What rakeback actually returns
Rakeback is a partial refund of rake already paid, typically ranging from around 10 percent at the low end to 40 or 50 percent on more generous reward schemes, sometimes delivered as cash and sometimes as points that convert to cash or tournament tickets. Applied to the earlier example, 30 percent rakeback on that 60 bb/100 rake bill returns 18 bb/100, which is meaningful, but it is a refund on a cost, not new profit. A player losing 3 bb/100 after rake and before rakeback is still losing money with rakeback added, just losing less of it.
Where rakeback changes outcomes most is at the margin. A player who is roughly break-even at the table, the kind whose skill edge is thin enough to be swallowed entirely by rake, can be pushed into clearly positive territory once a decent rakeback deal is applied. That is a real and durable gain, but it depends entirely on the player already having a workable game; rakeback multiplies an edge, it does not manufacture one from nothing.
Tournament rake works differently and is usually simpler to reason about: it is a flat percentage added on top of the buy-in, commonly 6 to 10 percent, taken once at registration instead of pot by pot, so it shows up directly against a player's tracked overall profitability rather than being buried inside individual hands.
04Reading a site's rake table before you sit down
Comparing two rooms on percentage alone is close to meaningless without also comparing the cap, the stake level it applies at, and whether the room offers rakeback or a loyalty scheme instead. A site advertising a lower headline percentage but a higher relative cap at small stakes can cost a low-stakes grinder more per hundred hands than a site with a higher percentage and a tighter cap. The only reliable check is running the actual rake table for the stakes you play against your typical pot size, not trusting the marketing number on the lobby screen.
None of this fully settles the harder question of whether a given game is worth playing after rake, since that also depends on the strength of the opposition sitting at the table, which no rake table can measure.