# How much is the tournament rake we pay?

• Basic
Joined: 05.03.2008
The question I am posing to myself, but also to all you readers may seem simple, some may even consider it as a rhetorical one, assuming that the answer is obvious. Nevertheless, I will dare to claim that, just like most poker problems, this one's correct answer is “it depends”.
In a cash game, depending on the cap, the rake can be computed based on a sufficient sample. At the stakes where most people play the rake is usually around 5% of each pot that is played if the flop is seen and there is a cap that depends on the poker room, the stakes, the number of players at the table etc. The final percentage of money raked on average may drop to around 4% of the pot if the hand isn't over preflop.
Many of you may quickly claim that the tournament rake is around 9% for most of the tournaments that the majority of players take part in, unless we're talking about a heads-up tournament, where it is usually 4,76%. On my turn, playing the devil's advocate role I will then ask you: is the rake a player pays the same, either he plays a 2 minute blind interval turbo, or a 10' blind interval regular tournament? The poker room earns the same money at either circumstance, but what will its hourly profit be at each of them? How many hands has a player bought with the same amount of rake at the super turbo tournament as opposed to the normal one?
So, the answer isn't that simple as it may seem. Let's take an example of real numbers to make the problem more clear. Consider a winner-take-all Sit'n'Go tournament between 10 players with a \$10+\$1 buy-in. Let us also suppose that each player takes 1000 chips at the beginning of the tournament. The poker room will earn \$10 from all players' fees added together and will reward the winner with a \$100 prize. But actually, this game is a cash game where each player paid \$11 to receive chips worth \$10, that is each chip he bought is worth \$0,01. This particular cash game has special rules. Rake has been prepaid by the players when the bought their chips, blinds increase as time passes and nobody can stand up with his chips to cash them out until one of them wins all the chips of the table. Furthermore, no player can rebuy chips during the game. Let us know turn this tournament to a more imaginary one assuming it starts with a blind level of 1000 antes and 5000/10000 blinds and we'll see all players go all-in in the dark at the very first hand. And if we have one single winner at this first hand, then all the players will have indeed paid 9% rake (\$10) out of a \$110 pot, in contrast to a normal cash game where they would have paid around 4%-5% of the pot. This particular tournament rake is indeed an expensive one. But if we collect some statistical data, we'll see that in such a winner-take-all turbo tournament that lasts around 70 hands on average, the total sum of pots when the flop was seen will be around 40000 chips (which is equivalent to \$400) and thus the \$10 that the players will have paid to play all these hands will certainly be less than the 2.5% of the “money” that built these pots.
Imagine now how many more hands the players of a MTT play until its end, especially if it is about a non-turbo one. Also take into consideration that as the blinds increase, the “money” that enters the pots is even much more than the money paid as tournament entries.

To be continued...
• 13 replies
• Bronze
Joined: 03.12.2010
Huge post

I'd suggest you break it down a bit or users will see it as more of a chore to read than interesting. Looks like you put a lot of work into the post so it would be a shame to see people overlook
• Basic
Joined: 05.03.2008
Part #2

Many of you will quickly say that in practice, since these tournaments aren't winner-take-all ones, those numbers are false. And this would be obvious since chips' value is not linear to \$ value, as, when there are two players left, the chips in play represent less than the total money played in the tournament (due to the 3rd place prize already awarded). I will still state that based on the data I collected, if these tournaments had the same prize of \$33 for all 3 players that got in the money and thus there would again be a closer to the linear relationship between chips and \$, the actual tournament rake would still be less than 3.25%!
Now, randomly selecting 3 tournaments of 79 hands each from my sample, that is 237 hands in total, I saw that, if the tournament was a winner-take-all one, there would be postflop pots of total value of \$1630 in value , with just \$27 raked. At the same time, a random cash game session of a similar buy-in (\$10) 9-seat NL Hold'em with blinds of \$0.05-\$0.10 consisting of 235 hands played at the very same poker room, the amount of money raked was \$20.97, while the total magnitude of pots raked was \$319.74. We're talking about a 6.66% average rake (while at higher stakes it drops to 5% and the cap is sometimes reached more often than in 1 out of 100 hands played)! Another observation I made during this comparison, is that half the pots were raked at the cash game session, while only one third of the pots reached a flop at the Sit'n'Gos.

To be continued...
• Basic
Joined: 05.03.2008
Part #3

Another observation confirming some common clues when comparing the rake paid in these two types of play (tournaments vs cash games) is the contribution of a particular hero to the rake that the poker room earns. A player playing big pots on average, i.e. an aggressive one, or a player playing many pots (loose) contributes more to the rake that the poker room earns from a cash game. Similarly, a tournament player that goes deep in a tournament plays more pots and bigger ones in value (since the blinds increase when he goes deep in the tournament). Another interesting observation concerning a tournament player's contribution to the rake is that it is much bigger (per number of hands played) for the player that busts early, than the player that went deep and played many more and much bigger pots, having paid the same rake. Long term though, the deep runs a player makes in some tournaments makes it up for the times he busted during the early stages, thus making the average rake paid more strictly defined. It is also obvious that players who are better than the average one and make it to the middle and late stages of the tournament quite often, actually pay less rake than the bad players who tend to bust out early.

In conclusion, it becomes obvious that players who mostly play non-turbo tournaments, or larger field ones, pay less rake than the ones who mostly play smaller field and faster tournaments. Even though tournament chips aren't equivalent to \$, I think nobody can doubt that, given the competition met in cash games nowadays, we pay them much more than we pay for our participation in tournaments.
• Bronze
Joined: 30.09.2012
Totally agree. I skimmed through this post and incurred a 75% rake on my time and effort.
• Super Moderator
Super Moderator
Joined: 02.09.2010
Hi, SamGreekRNMD
I had a quick look over your post, and if I catch your drift, you are looking at the rake in terms of \$/hour in the various formats.

I think that you may be missing an element though.

If a player has a particular winrate (> 0) then the rake as a percentage of profit would actually be much less in the shorter, faster games.

On the other hand, rake as a percentage of profit, would be much higher in the huge MTTs, since the profit is non-existent unless you cash big, at which point rake becomes largely irrelevant.

Have you actually run the numbers on this?

A table showing the different formats vs the rake charged and hours of play might be a lot easier to follow than 3 text-heavy posts.

Cheers,
--VS
• Basic
Joined: 05.03.2008
Originally posted by VorpalF2F
Hi, SamGreekRNMD
I had a quick look over your post, and if I catch your drift, you are looking at the rake in terms of \$/hour in the various formats.
Actually that's not 100% true. I am looking at the rake in terms of \$raked/\$wagered. The \$wagered is hard to calculate though in tournaments, as the total chips wagered have to be expressed as \$tournament equity, so an equivalent of ICM has to be used. The comparison there between cash games and tournaments becomes quite hard, but can be approximated quite well if we are talking about winner take all tournaments, where chips correspond to certain cash amounts at any time. What you say here is closer to the truth when we compare regular tournaments with turbo ones, as the chips (and as a result the \$equity they represent) wagered in turbo tournaments are much less than those in regular ones.
• Basic
Joined: 05.03.2008
Originally posted by VorpalF2F

If a player has a particular winrate (> 0) then the rake as a percentage of profit would actually be much less in the shorter, faster games.

If I get it, you must be referring to tournaments here (as you mention faster games). The problem here is that in faster games the ROI gets lower, because the blinds rise faster, so there is no particular winrate referring to different speed tournaments. Turbo tournaments are ideal for mass multi-tabling regulars who will get a good rakeback bonus and at the same time increase their hourly profit due to multi-tabling (but not their ROI per tournament, which is what we examine here).
• Basic
Joined: 05.03.2008
Originally posted by VorpalF2F

On the other hand, rake as a percentage of profit, would be much higher in the huge MTTs, since the profit is non-existent unless you cash big, at which point rake becomes largely irrelevant.

With this way of thinking, cash game sessions that won't give you many buy-ins of profit also have a quite high rake. But this is not a correct approach in my opinion, since we have to look things in a more long term way. And in order to determine the rake that players pay in total (and not each one of them individually), we have to add all the chips' value (in \$) entering a pot (that the flop at least was seen) and compare it to the rake that was held from all the players. In other words, what the house earns compared to the pot contribution of all players in total. In cash games it is easy, as chips=\$, but in tournaments we have to make guesses and approximations (that I haven't done of course in such depth as to come to a 100% clear conclusion). These approximations, as I have said earlier, require ICM calculations for huge fields, especially in MTTs and are not feasible in any way...
• Super Moderator
Super Moderator
Joined: 02.09.2010
Hmmmmm,
My brain is bending here somewhat.

I think that the chips become irrelevant the moment they are purchased.

Here is why:
You decide to enter a tournament.
To enter this tournament -- whatever its size or structure -- you must put up your share of the prize pool and pay an entry fee.

We are used to viewing rake as a percentage of the prize pool amount, and we understand it on those terms, but there is no reason to have a fixed ratio of prize money to fee.

You then play the tournament to win the prize pool -- the fee part is gone.
If a large number of participants get relatively small amounts (less than the fee) that might be something to take into consideration.

The fee to enter the tournament is not connected in any way to the value of the chips.

If you divide the prize pool by the total number of chips in play, you have the value of a single chip at the beginning of the tournament, and everyone has an equal number, and thus an equal anticipation of reward.

ICM says that the bigger your stack, the greater your expected reward -- that is why chips you lose are worth more than chips you gain when you play a tournament.

The fee you paid to enter the tournament never enters into this calculation at all.
If you try to compare the rake to the value of your stack, then it changes every time you win or lose a pot.

This was a great exercise, but I'm going to just keep it simple and look at the percentage of the fee to the prize money.

Have a great weekend,
--VS
• Basic
Joined: 05.03.2008
I seem to have confused you with all these thoughts. I never compared rake to the players' stacks (only with the "money" entering the pots). I tried to estimate how much the tournament rake would be, if (hypothetically) it was held with cash game terms, i.e. if every time a pot was won (with the flop seen) the dealer would hold some of the tournament chips of the pot as house rake. Finally, the winner of the tournament would cash out his prize based on the total number of tournament chips left. The rest of the prize pool would have been the tournament rake (but held with cash game terms). And if it weren't a winner take all tournament, then the rest of the players would be paid according to the percentage of the final prize pool (after subtracting the rake from the buy-ins) that corresponds to their finish position. I will give you some numbers to make it more clear:

Imagine a tournament that starts with 10 players, \$10 buy-in and 1000 chips initial stack. Each chip is worth \$0.01. Let's say first one wins 50% of the prize pool, 2nd 30%, 3rd 20%. But that's a tournament that the rake is held in chips out of every pot, when players see a flop. So, let's say that after the end of the tournament, the winner accumulates 8000 chips (because the rest of them were held as rake by the dealer). These chips are worth \$80, so the winner will cash out \$40, 2nd place will cash out \$24 and 3rd player \$16. So, the rake of such a tournament would then be \$20 out of a total of \$100 (sum of buy-ins), so it is 20% of the total buy-ins. In this tournament the rake was held with cash game terms.

What I am trying to prove in my article, based on the numbers I have run, is that if there was a way to hold the rake of tournaments in the way I just described above, they would be much more expensive than they are now that we pay the rake in advance, as a certain percentage of the total buy-in, no matter how much "money" (chips) enter the pots we play, how many chips change hands without a flop seen etc. This way, I tried to prove that a certain edge a player has over his opponents can be more easily exploited in tournaments, because the rake becomes more beatable the way it is paid, in comparison to the cash games. I think this also proves why cash games have become more difficult nowadays, as well as turbo tournaments are harder to beat than regular speed ones. Some poker rooms also give players motives to play turbo SnGos by charging a smaller percentage of the total buy-in as a fee.

I hope I made it clear now

I wish you have a great weekend too!
• Super Moderator
Super Moderator
Joined: 02.09.2010
OK, I get it now.

I would be nice to see some hard data in table or graph form, though.

The "wall of text" format is tough to follow.

Cheers,
--VS
• Silver
Joined: 11.01.2009
It of course depends on the game type. In CAP NLHE, I guess rake is like 0.07 BI/100, and 0.1 BI/100 with normal stack sizes (100 bb), so if a tourney takes fewer than like 125 hands, then it's raked harsher.

It's known that SnG players are even more dependent on cashback than even PLOlers (whose rake in BI/100 is twice bigger). That's why the former tend to go for SNE more often.
• Basic
Joined: 05.03.2008
Originally posted by tonypmm
It of course depends on the game type. In CAP NLHE, I guess rake is like 0.07 BI/100, and 0.1 BI/100 with normal stack sizes (100 bb), so if a tourney takes fewer than like 125 hands, then it's raked harsher.

It's known that SnG players are even more dependent on cashback than even PLOlers (whose rake in BI/100 is twice bigger). That's why the former tend to go for SNE more often.
Yes sir, that is correct about the game type. But a cash game players buy-in is not representative of the amount of money he/she contributes in the pot, which depends on the time/hands played, as long as the avg pot size etc. For example, if I elect to play a 50bb stack strategy (and not start with a 100bb one) and leave the table after my stack goes up to 80bbs (or more) the stats you gave about rake compared to buy-in dramatically changes.

About SnG players achieving SNE status it also has to do with the ability to mass many more tables than a cash game player who plays at stakes requiring the same bankroll. This has to do with hand reading requirements, as well as decision making process, which is easier to do in SnG, as many hands require a single preflop decision, while at cash games you have to pay attention to flop-turn-river action much more frequently than in the SnG.

You are also correct about turbo tournaments lasting few hands, I have already analyzed my opinion about the topic. It doesn't matter the # of hands played, but the amount of "money" entering the pot.

Originally posted by VorpalF2F
OK, I get it now.

I would be nice to see some hard data in table or graph form, though.

The "wall of text" format is tough to follow.

Cheers,
--VS
That's good, you gave me motive to work on it within the next days, I will try to make a data sheet with numbers about it.