The money flow
Understanding poker economics requires tracing where money actually goes, which is simpler than in other verticals and has less obvious consequences.
Money enters the ecosystem when players deposit. Overwhelmingly this comes from recreational players, since winning players withdraw more than they deposit by definition.
Money circulates between players as hands are won and lost. This movement generates no revenue for anyone and is the game itself.
Money leaves in two ways. The operator takes rake. Winning players withdraw their profits.
For the ecosystem to persist, deposits must exceed rake plus winning-player withdrawals. If they do not, the total money in the system declines, games get shorter, and eventually there is nothing to play for.
That is the entire economic model, and every decision about rake, rewards and ecology is a decision about this balance.
Rake structures
Percentage of pot with a cap is the standard cash game structure. The operator takes a percentage of each pot, typically a small single-digit figure, up to a maximum amount regardless of pot size.
The cap has a significant effect. At low stakes, pots rarely reach it, so the full percentage applies throughout. At high stakes, the cap is reached quickly and the proportional cost falls sharply.
The consequence is that low-stakes games are proportionally the most expensive, which is a genuinely awkward property. The players least able to afford it, and least likely to be skilled enough to overcome it, face the highest proportional cost.
No flop, no drop rules waive rake on hands that end before the flop, which reduces the cost of tight play and is now standard.
Tournament fees are charged as a separate amount alongside the buy-in, typically expressed as a percentage. A tournament advertised at a given entry consists of the prize pool contribution and the fee, and the fee is the operator's revenue.
Time-based rake charges for table time rather than per pot, which is uncommon online and removes the distortion the cap introduces.
Attribution methods determine how rake is credited to individual players for reward purposes. Dealt attributes equally to everyone dealt in. Contributed attributes in proportion to money put in the pot. Weighted contributed is the common standard and attributes proportionally among those who contributed. The choice affects who earns rewards and therefore which playing styles are subsidised.
Effective rake and beatability
The figure that matters to players and therefore to the ecosystem's health.
A player's result is their edge over opponents minus the rake they pay. A modestly skilled player with a small edge can be a net loser after rake, which means the rake level determines how good a player must be to break even.
At low stakes, where the cap rarely binds and the proportional cost is highest, the standard required to beat the game is correspondingly higher. This produces an outcome that is worth stating plainly: a substantial proportion of low-stakes players who are better than average opponents are still losing money, because the rake exceeds their edge.
Two consequences follow.
Recreational players lose faster than the underlying skill difference implies, because rake compounds their disadvantage.
The room's population skews towards stronger players over time, since the marginal players who might have persisted are pushed into losing.
Operators setting rake are therefore balancing revenue per hand against the survival of the player population, and the balance is easier to get wrong in the revenue direction because the cost appears slowly.
Reward programmes
Poker rewards exist because rake is high enough to make the game difficult, and because competing rooms offer them.
Rakeback returns a proportion of the rake a player generated, historically as a straightforward percentage. It reduces effective rake directly and was the dominant mechanism for years.
Tiered loyalty returns increasing proportions as a player's volume rises, which concentrates the benefit among high-volume players.
Points-based systems accrue on rake paid and convert into cash, tournament entries or merchandise.
Missions and challenges reward specific behaviours rather than raw volume, which allows targeting.
Randomised rewards, including chest and prize-drop mechanics, deliver value with a variable element. These are engaging and are structurally closer to casino mechanics than to a loyalty programme, which is worth noticing.
Freerolls and exclusive events provide value in the form of playing opportunities rather than cash.
The critical structural question is who the rewards flow to.
A programme paying the highest rates to the highest-volume players returns most of its value to professionals, who are already extracting money from recreational players. The subsidy accelerates that extraction and shortens the life of the ecosystem.
A programme weighted towards recreational players, or towards behaviours other than volume, keeps more money with the population that funds everything.
The industry has moved substantially in the second direction over the past decade, replacing pure volume-based rakeback with structures that reward variety, occasional play and engagement. This was unpopular with professional players, who correctly identified that their returns were being reduced, and it was a rational response to the observation that rooms with heavily professional populations were declining.
The high-volume player question
The tension that sits underneath reward design, stated directly.
High-volume professional players generate substantial rake. They fill tables during quiet periods. They are, in a narrow accounting sense, valuable customers.
They also extract money from recreational players continuously, and they extract it faster than recreational players expect to lose it. A recreational player who deposits expecting an evening's entertainment and loses in an hour to a professional playing twelve tables does not come back.
The commercial calculation is therefore not simply about rake generated. It is about whether a player's presence increases or decreases the total money in the ecosystem over time.
Operators have addressed this through the ecology measures described in the first lesson: seating restrictions, table limits, anonymity, restrictions on tracking tools, and reward structures that do not favour volume.
Professional players argue that these measures penalise skill, that the games are advertised as competitive, and that a business cannot invite people to compete and then restrict those who compete well. That argument has real force and it resembles the customer restriction debate covered in the Sportsbook Trading course.
The operator argument is that the product depends on recreational players continuing to play, that unrestricted professional activity depletes them, and that everyone including the professionals is worse off in a room that has emptied.
Neither position is straightforwardly correct. What is observable is that rooms which did not manage this saw their recreational populations decline, and that the industry's practice has moved towards management rather than away from it.
Setting rake
The practical decision, and the considerations that bear on it.
Competitive position. Rake is comparable across rooms and players who care will compare it. A room raking substantially above the market loses the players most sensitive to it, who are disproportionately the higher-volume ones.
Stake level. The cap structure means low stakes bear the highest proportional cost, and operators concerned about their recreational population sometimes adjust caps to reduce it.
Format. Faster formats generate more rake per hour at the same rate, which means the same nominal rake is a heavier burden.
Reward offset. Effective rake after rewards is what players actually experience, and a room with high headline rake and generous rewards may be cheaper than one with the reverse.
Ecosystem health. The measure that matters over time and the hardest to observe directly. Deposit volume relative to withdrawal volume, recreational player retention, and the proportion of the population that is net winning all indicate whether the balance is sustainable.
Monitoring the ecosystem
To close, the figures a poker operation should watch.
Net deposits, meaning deposits minus withdrawals, which shows whether money is entering or leaving the ecosystem.
Recreational player retention, which is the leading indicator of everything else.
The proportion of rake generated by high-volume players, which indicates how dependent the room is on professionals.
Winning player concentration, meaning what share of the population is net profitable and how much they extract.
Average session length by player type, since recreational sessions shortening indicates bankrolls being consumed faster.
Effective rake by stake level, after rewards, which is what players experience.
New player survival, meaning how long a first-time depositor lasts, which is the most direct measure of whether the games are hospitable.
A room where net deposits are declining, recreational retention is falling and rake is increasingly concentrated among a small number of high-volume players is in the decline described in the first lesson, and the decline is easier to arrest early than late.
A worked rake calculation
To make the cap effect concrete, an illustrative comparison.
A room takes 5% of the pot with a cap of £3.
At a low-stakes table where the average contested pot is £20, the rake is 5% of £20, which is £1, comfortably below the cap. The proportional cost is the full 5%.
At a mid-stakes table where the average contested pot is £60, the rake is 5% of £60, which is £3, exactly at the cap. The proportional cost remains 5%.
At a high-stakes table where the average contested pot is £300, the rake is capped at £3. The proportional cost is 1%.
The player at the high-stakes table pays a fifth of the proportional cost paid by the player at the low-stakes table.
Extending this to a player's overall economics, a low-stakes player contesting a hundred pots in a session at that structure pays substantially more, relative to the money they are risking, than a high-stakes player doing the same. The edge required to overcome it is correspondingly larger.
This is why the observation in the main text holds: a low-stakes player who is genuinely better than their opponents may still be losing money. The rake is doing more work against them than their skill advantage is doing for them.
Operators concerned about this adjust the structure at the bottom of the ladder, either by lowering the percentage, lowering the cap in absolute terms so it binds sooner, or applying reward rates that offset more of it. Each reduces revenue from the segment least able to generate it and improves the survival of new players, which is a trade with a long payback.
Reward design in practice
Some concrete guidance on structuring a programme.
Decide who you are subsidising. Every reward structure returns value to somebody, and the choice between volume-weighted and behaviour-weighted determines whether that is professionals or the general population.
Make the value legible. Players who cannot work out what they are earning do not value it, and complex points systems with opaque conversion rates deliver cost without appreciation.
Avoid pure volume ladders that pay dramatically more at the top, since these are the structures most associated with professional subsidy.
Include non-cash value. Tournament entries, exclusive events and status cost less than cash and are valued, particularly by recreational players for whom the experience matters.
Reward variety and engagement rather than only quantity, which directs value towards the players whose participation the ecosystem depends on.
Watch the randomised mechanics. Chest and prize-drop rewards are engaging and are structurally similar to casino products. Introducing them into poker imports a set of design considerations that belong to a different vertical, including those covered in the Product Innovation course, and they should be assessed accordingly rather than treated as a loyalty feature.
Measure what the programme changes. The holdout discipline described elsewhere in these courses applies here. A reward scheme that pays substantial value to players who would have played identically without it is an expensive way of transferring money.
Rake in tournaments
The structure differs enough to warrant separate treatment.
A tournament entry is split between the prize pool and the operator's fee. An event advertised at £50 plus £5 has a prize pool contribution of £50 and a fee of £5, giving an effective rake of roughly 9% of the total paid.
Several features follow.
The fee is charged once, at entry, regardless of how long the player lasts. A player eliminated in ten minutes has paid the same fee as one who reached the final table.
Proportional fees vary by buy-in level, and as with cash game caps they generally fall as buy-ins rise. Low buy-in events carry the heaviest proportional fees.
Rebuys and add-ons carry their own fees, which means the total paid by a player who rebuys several times is considerably more than the advertised entry suggested.
Overlay reduces effective rake to negative for the players in an event that did not fill, since the operator is contributing to the prize pool.
Satellites complicate the picture, since a player entering a large event via a satellite has paid the satellite's entry and fee rather than the target event's.
The player-facing implication is that tournament costs are less transparent than cash game rake, because the fee is bundled into a single advertised figure and rebuy potential is not stated as a total. Operators presenting the fee clearly and indicating the realistic total cost of a rebuy event are being more transparent than the norm, which is a low bar and worth clearing.
Where the model breaks
To close, the failure sequence that this lesson has been describing, stated as a progression.
Rake is set to maximise revenue per hand. Rewards are structured to retain the highest-volume players, because they generate the most rake. Fast formats are promoted because they produce more hands per hour.
Each decision is individually rational and each increases the rate at which money leaves the ecosystem.
Recreational players lose faster than they expected. Their sessions shorten, their deposits become less frequent, and some stop.
The remaining population is more heavily weighted towards strong players, which makes the games harder, which accelerates the departure of the recreational players who remain.
Liquidity falls. Games run less often, at fewer stakes, at fewer hours. The room becomes less usable, which drives further departures including of the professionals whose income depended on the recreational players.
The decline is difficult to reverse because the thing needed to attract players is players.
That sequence has played out at real rooms and is the reason ecology management became standard practice rather than a theoretical concern. An operator setting rake and reward policy purely on revenue per hand is optimising a number that will be smaller in two years because the population generating it will have gone.