Hello all,

I and several others will travel to the Isle of Man next week for discussing PLO rake with Pokerstars: it is commonly believed that PLO rake is too high, and stars will discuss this issue with us.

As we are representing the player community, I'd like to solicit player input about the subject of PLO rake. Many threads have already been written on the subject, but I think that another one leading up to the meeting could help.

In this thread we can discuss anything rake-related. I am personally particularly interested in the following questions:

1. what amount and distribution of rake does the community consider to be "fair"?

2. how does the poker ecosystem work and how does rake affect it?

3. What are some solutions to the rake problem that do not involve actually decreasing the rake percentage, but that the community will still be happy with?

4. What is the right way to compare rake in PLO to rake in other games?

5. What are the most convincing talking points for why rake is too high and should be lowered?

6. How would you try convince stars staff that PLO should be raked differently than NLH (in terms of percentage and rake cap)?

I have my own suggested answers to questions #4 through #6, which I'll share later after the thread develops.

Thanks in advance to anyone who chimes in!

1. make it same as NL

4. by bb/100

Thanks for doing all this and going to the isle of Man, also thanks to all the other reps going.
There is so much stuff that was discussed that it's easy to miss the forest for the trees, so hope my post sums up a bit.

1. Same as NL in terms of bb/100 rake.

2. Money should flow upwards, if rake is too high the higher games will dry out because very few will eventually move up/take shots, regs will play lower or/and move to other sites.
Midstakes will get tougher as well when those regs start playing there and dry out as well after some time.

3. More VPPs which I think would be even better than reducing rake %.
People like VIP levels (myself included ).
They start chasing higher VIP levels & bonuses which were out of reach before.

4. bb/100 is the only way that makes sense to me.

5. This has gotten longer so I'll start with the summarization:

- Recs getting on heaters / getting hooked to the game & making more deposits and even telling their friends about poker & their success.
- Recs losing at a slower pace -> they keep trying & want to improve
- Regs playing other regs because it's not totally senseless and they can actually win (and be it by rb only).
- Recs & Regs chasing higher VIP levels -> more volume
- Recs & Regs happy


Stars main goal besides making profits is to make them longterm.
They don't want to "hit n run" recs, they want them to stay!
The keywords are frustration & motivation.

Frustration:
Recs are paying even more rake due to their higher vpip, combined with tougher getting games they often get pretty much pulverized and demotivated right away and are unlikely to make another deposit.

Motivation:
If recs can win more often, especially when they start playing they are way more likely to redeposit because they got hooked, they won't redeposit if they get stomped .
Even if they don't catch a heater and only lose at a fairly slower pace, there might be those who want to actually improve.
Instead of saying "ok no chance for me" they might still have motivation left
because they don't feel totally owned.

Another motivation factor would be both the higher VIP levels & bonuses, if Stars goes that way instead of reducing the rake %.
They want to reach their next VIP level and they start saving up for those bigger bonuses for gold or even platinumstars which would have been out of reach for them before.
If they bust during their chase they still have their VIP level & their unfinished bonuses in mind, a lot more than right now where they
have close to no chance of getting anywhere near it.

This will attract many players, not just losing recs.

What is also very important is to not focus entirely on recs but also on winningplayers/regs.
Lowering rake or increasing VPPs will also lead to less tableselection and will solve
a fair bit of the problem of bumhunting/hardcore selecting which is again good for recs because they feel less hunted.
It's especially good for stars because our winrate will be low and they get their rake, but at least we can actually win and have the feeling of a challenge again
by playing decent players, which would be something refreshing.

The state in lowstakes PLO is very sad. I'm selecting like a bitch atm and I see other good regs doing so as well (actually I select more than I need, perhaps a bit out of protest & frustration over rake...).
Best example besides myself that comes to my mind is 'foldngst8n' (CR instructor) who I respect a lot and he snapsitsout
when he is still very likely the best player at the table but there is no fish...

There is no sense in playing even when you have the edge, this sucks bigtime and promotes bumhunting.

The other regs will eventually realize this as well and start doing the same.
Right now I see a mix between reggames running and games that are breaking after a fish left, so it's a perfect time to change something because it's not too late.

Lower rake will therefore keep games between regs going more often when they can actually win over another.

The last point would be the result of happier recs & regs.
Happy recs as explained in length and also happier regs because we
got heard by Stars and don't feel like cash cows.

Looking back there was a lot going on, from eventual strikes over to switching sites all the way up to creating a new site and what not.
The easiest and best way is to simply work with Stars and come to some good changes which will be good for everyone!

3. The slower you play and the lower your VPIP is, the more rake you pay relatively. It would fasten up the games and make the recs lose a bit slower.

Cliffs: recurrent recreationals are kings and rewards should revolve around them; PLO should be as profitable as other games given the same bankroll, but not more (NL converts won't let it happen by the market laws).

3. People like rewards more than equivalent rake reductions, and they like to fight for these rewards, not just have them granted in an opaque way such as Essence. Make promotions more oriented towards recs, encouraging small but consistent volume like Iron Man, Rush Week (the next week btw ) or Take Two, losing playing styles like Boom! Award merchandise, or tickets to big sport events or MTTs suffering from chronic overlays , or reload bonuses, or temporary VIP level upgrades (like FTP's Level Up), instead of instant cash.

Give awards for crazy plays that were lucky to result in a win, e.g. the weakest 5-card hand with badugi hole cards out of those that won a 500+ bb pot (1000+ bb at ante tables) in a particular week. Make jackpots for bluffers popping often enough to give anyone a chance, e.g. triggered by winning a hand with pocket quads.

Punish unethical behaviour in the chat and at the tables in general severely and monetarily (FPP fines). Reward Facebook and Twitter shill posts.

The goal is to make recreational players more willing to return to the tables consistently, pleasing the community is a secondary priority: it will be delighted automatically when it sees recs return more and more, which will boost regs' bottom lines more than reduction in rake.

Revenue = deposits minus cashouts. Cashouts (winning regs' earnings) can't be increased while keeping the site's revenue the same or bigger, unless more deposits are encouraged. But recs will bring more money in only if they're more satisfied with the playing experience. They normally have a weekly / monthly allowance of money they can afford to lose and won't increase it unless they get more pleasure.

The delight from going on a heater and moving to higher stakes may be diminished by facing unethical behaviour, which will inevitably be harsher at high limits because more money is at stake and to be won. Recs' losing speed can be reduced in two ways: by lowering the bb/100 rake... and by making them play lower stakes with the same delight (the 888 / Bovada model), e.g. offering lower but weekly reload bonuses. Variance is not the only component of pleasure for them - the social microclimate is also important - but it can be boosted by introducing lotteries into cashback (Santa Clauses, milestone hands and fancy play jackpots on a more regular basis than now).

Btw the path of small but consistent giveaways is what Ultimate Poker has recently chosen too. The euro networks are following this trend too.

1, 4. Imho a good measure of a game's profitability for a reg is the amount of money that can be squeezed out of the same bankroll on average with a set risk of ruin (say, 5%) or, say, basing on a Kelly strategy. So what should be compared is $/hand winrates at not nominally equal stakes, but at corresponding stakes in BRM terms.

Standard deviation of PLO in bb/100 are twice (or how by many times?) bigger than those of equal NL stakes, hence a twice lower stake should be played at PLO than NL given the same roll, and post-reward bb/hour winrates should be twice bigger in PLO. But not even bigger, otherwise there will be a lot of NLHE converts, whom personally I don't want to face at the tables because they're weak just upon the switch but serious about learning and gain strength fast, I'd like to face clueless casino customers instead.

While average PLO100 regs earn 4 bb/100 on average post-rewards according to OP, aren't average NL200 (corresponding stake) regs breakeven as well, earning as little as 2 bb/100 post-rewards? (I'm not sure, analysis similar to OP's should be done, at least on small available SSNL DBs.) Don't forget that NL is dead too, and by advertising PLO (and Stars as opposed to other sites) too much by lowering the rake by too much, you'll just rebalance the reg pool but not necessarily enhance the fish pool.

the uber micros are painful. being raked 23bb/100 is not fun, especially when it means the difference between being profitable and losing.

It would be nice to see them try to make the road from PLO2-PLO25 a reasonable option for beginners who want to progress and not get raked into oblivion.

Thanks for the posts so far! I reply to cbt and to coo.n in detail below.

I'd like to ask everyone who thinks bb/100 is the right way to measure rake, if they could explain why they think so. Steve, for example, claims that percentage-and-cap is the right system. I know why I disagree with him, but I'd like to hear other opinions.

Also, SlimyF, can you explain your post? I didn't understand it.

Reply to coo.n

Hey coo.n, thanks for the post. It's full of very interesting ideas. Some of them are not rake-related, but I'll make sure that they are raised in some capacity. I also disagree with a few of the things you said, but nothing too critical. Here are some specific replies:

But how do we compare $/hand winrates? After all, PLO is a high-variance game, and no one, including stars themselves, knows anyone's true winrate. As you know, I've been working on finding true winrates of the player community and am having a hard time even coming up with a decent guess, so how can we come to stars and tell them that this is the right criterion, seeing as no one in the world can compute it?

Also, a couple of smaller issues:
1. I'd compare $/hour, not $/hand. Which probably means dividing by like 2, since PLO is both slower and less susceptible to multitabling than NLH.
2. The Kelly criterion says you need a smaller bankroll the higher your winrate is (IIRC this goes linearly: if you double your winrate, you halve the required bankroll). So assuming PLO winrates are typically higher, this needs to be taken into account as well in your method.

I don't know. I hope to do this analysis in the future.

Can you elaborate? What is the 888 / Bovada model?

Two issues I had in general with what you wrote are:

1. I'm skeptical of your claim that PLO and NLH will balance each other based on market forces. There are all kinds of barriers to entry which can stop NLH players from converting.

2. You suggest, generally speaking, to prioritize attracting fish over decreasing rake. Your general argument is that more fish will make the games softer, and will increase winrates. My issue is that if the rake is too high it's hard to make the games beatable even if you attract fish. I suspect that what would happen is that as the number of fish increases, the number of regs will also increase, so you'll end up with a reg-to-fish ratio similar to today, and regs still struggling to have a decent winrate, and not being able to move up.


Reply to cbt

Thanks for the kind words, cbt! Thanks also for writing up the talking points. They're a good summary of the arguments for why to reduce rake, and they'll be useful to me.

Can you explain why you think bb/100 is the right way to measure rake? vI have my own arguments, but I'd like to hear yours first.

Just meant that if you play loose and fast (mean time one takes to sertain action), like most of the recs, it would be fair for them to pay less rake and that way they would go busto a bit slower.

Also 7823 tabling nitregs always have some kind of rb, which recs basicly don't have on Stars, and that way they could compensate some of the lost rake with rb..

Pretty sure that this idea won't get much love from the regs, but no one can't say that it wouldn't be fair. Should be thought at least on lower limits.

I'm pretty sure Stars can compute winrates of certain groups of players (e.g. Supernovas, those who play at weekends only, net depositors that have been active for more than a year) pretty accurately as they have access to tons of hands.

All right, we have to take into account that PLO players have to cash out more $/100 to make a living as they play fewer hands a month... but with that amendment (subtracting the cashout rate from the winrate, obtaining the profit (bankroll growth) margin), I still find it more convenient to use bb/hand for RoR calcs.

The discrete version of the Kelly criterion, which I should have called the Chen-Ankenman formula, says that one should play in games that maximise 2*winrate*bankroll - variance (all in $ and for the same sample size; by winrate, I mean the profit margin here, the amount you leave in the bankroll to have it grow and insure yourself from the ruin).

If you assume that winrate = w * stake and variance = v * stake * stake (where w, v are constants), which is true for trading and betting that are scalable but not poker (where you can't play PLO24 or PLO27 and also games are tougher are swingier at higher stakes so BRM is stricter), you'll find that the maximum is at stake_optimal = bankroll * w / v (the classical Kelly criterion) and hence winrate_optimal = w * stake_optimal = bankroll * w * w / v.

Since v_plo = 4 * v_nl (v = stddev^2 and stddev_plo = 2 * stddev_nl), we have to attain w_plo = 2 * w_nl to have winrate_optimal_plo = winrate_optimal_nl.

Oops, these are two quite opposite models (social 888 vs all-anonymous Bovada), my bad, but both are rec-centered and the sites do their best to deter regs (tiny VIP rewards and fixed costs per acquisition, not revenue share, for affiliates).
http://pokerfuse.com/features/editor...-888s-success/
http://pokerfuse.com/news/industry/b...-leader-04-01/

Fairly enough but the most cited barrier is the PLO rake.

Fewer regs will appear if poker is marketed not to them (i.e. not as a skill game that requires hard studying to succeed) and efficient learning tools are less advertised (instead, there should be, and there are, fake poker academies run by the operators themselves that satisfy recs' minimum curiosity).

Hey Eldodo. Thanks for fighting the good fight on our behalf. I don't play Stars much, but I think as the industry leader their changes will have meaning.

I think Stars is very unlikely to make any significant changes in the short term, but will give way on the sort of changes they have done in the past.

I think in the short term we should bargain for:

-Increase VPP multiplier for low stakes.
-Decrease requirements to reach and or maintain silverstar, goldstar, and platinum star.

Medium term:

-Rake reduction at lower stakes
-Increase frequency of low value promos.
-Increase of rewards to recreational, losing players. Maybe bump up the BE players but certainly not to the same degree.

Long term:

-Rake reduction to at least NL rates across all stakes.
-Evaluate different models of rake / rake back.

Well an alternative might be to look at hourlies, but that would include other aspects than just the rake.
I remember Steve saying that bb/100 was not the right measurement but he also didn't really come up with an alternative .

Tyrannic, thanks for the ideas. I noted them down and will bring them up.

I think Steve did suggest an alternative. IIRC he thinks the right way to measure rake is percentage-and-cap. By which standard, NLH and PLO and raked the same.

Steve correctly claims that measuring in bb/100 or bb/hour is not good enough either because in PLO good players can win more (in bb/100) so it makes sense that they pay more rake (in bb/100). I tend to agree with him: measuring bb/100 isn't good either.

I'll write something longer about it soon, but here's the jist of my objection to use percentage&cap to measure winrates:

Take a game which is just like NLH, except that every time someone wins the pot, a coin is flipped and with probability half each player gets back all the money they put in the pot (minus the rake). In this modified NLH, pre-rake edges are obviously doubled. But the modification probably made the game unbeatable because rake is taken even on the "canceled hands". Well, PLO is a bit like this: hand equities run closer, so the ratio of rake to typical equity is way larger than in NLH. That, I think, is the problem with measuring rake just by percentage-and-cap.

I want to find a sensible and fair way to measure rake, such that it can easily be compared between different games. One suggestion is to compare the total winnings of all winning players to the total rake paid. I'd love to have some other suggested methods.

would also be interesting to compare the % of players in NLHE and in PLO who are losing players ONLY after rake is considered.

Are the proportions different? should they be? what impact does this have?

What do you think of a system that calculates the rake depending on the equity? This way split pots should have nearly no (or no) rake, and all those typical flippings are way cheaper. I think this would help plo alot.

(This idea is from TomGrill of the german poker forum pokerstrategy.com and also told to the german representative for the stars plo meeting)

Here the German thread mentioned (read it with e.g. Google Translate). TomGrill didn't give an exact formula for e.g. multiway all-ins so I can't test it in PT4

But this will be a 25%-ish discount anyway because, according to my research, only 30-35% of total PLO20 table rake is taken in pre-river all-in pots.

I do think bb/100 is the best way we currently got to measure rake. Its a set amount of rake we know we pay for a sample of hands that we can compare across games. Steve want to factor max winrates also but think variance is overlooked both for whats possible (winrates) but also for that the extra risk there shouldnt be 'rewarded' with a premium rake price.

I think more vpp for PLO make most sense, and probarly best to have a cross stakes solutions so there is not weird effects in games played because of different rewards.

Yea, more VPPs for PLO for easier access to SN(E). That would be a big step forward.

Suggestions on appropriate multiplier? 2x easy/fair given NLHE comparison?

It still sucks for players that just learn the game and start from bottom. Its still hard to reach SN(E), cause your learn alot and play less, and second, the rb will come to account one year later. Normally Id expect to have left the micros/lows at this time.

The VPPs earned for SN(E)'s would have to be different of course then, otherwise they would receive well above 100% in rb .
Other than that I think it would be pretty awesome esp for the recs.

Just wanted to make sure everyone know that even if I didn't reply to everything, I read everything, and will re-read it tomorrow before the meeting.

Now, for some replies:

I'm open to being convinced otherwise, but this seems like a terrible system to me. The main problem is that it incentivizes players to go all-in early in the hand, since that will minimize the amount paid in rake. Therefore, you'll get more action, but not quite of the kind that makes poker a better game. Put another way, it deincentivizes playing multiple streets and causes there to be a difference based on when the money got in.

Well, I think that we need to take into account both winrates and variance. The issue is, though, that I think winrates matter more than variance: among the players that try to play PLO100, all of them need to learn to beat the fish, but only some of them are constrained by BRM (the others are overrolled and are constrained by skill). But in any case, I think you're giving a great argument for why bb/100 is *not* the right measure: it doesn't measure any of the things you specify.

Yeah, I agree. I hazard a guess that this is probably the kind of solution that will be agreed upon in the end.

So a 2x multiplier on only VPPs, or also FPPs? From your first paragraph, I'm guessing you're talking about multiplying VPPs, but leaving FPPs the way they are now. This makes some sense to me (I still think it might be too low), but the 2x multiplier seems too low. I think people need much faster access to SN. Maybe a 5x multiplier or so.

And, again, this seems to be a restricted-time solution: I suspect the games will become barely-beatable again in a couple of years. But it'd be good for now.

I've chatted a bit with the inventor of the system (TomGrill): he said he hadn't thought it over very well (and isn't even sure about the formula for the exact reduction as a function of the equity edge). I see little sense in it too.

I just threw the 2x out there, as I recall PLO rake bb/100 is roughly 2x that of NLHE. Someone needs to do some math on the effects accross stakes.

Is this true if only VPPS (not FPPS) is earned faster since u dont earn cashbonuses just milestones faster?

If you double VPPs without changing the FPP multipliers, then twice more FPPs will be earned per $1 of rake and hence twice more in cash bonuses. But even if you cut FPP multipliers in half for PLO, then SNEs will still get a lot that Stars would like - the bonus component of SNE returns 44% and the milestone one is 23.76%, so if only the milestone is doubled you'll still have 91.5% in returns for $7.5K in monthly rake - even iPoker doesn't offer such awesomeness.

But it bolis down not to return percentages, but to 1) what share of net deposits the room keeps and what share is given to net withdrawers, 2) what player behaviour is encouraged by the VIP system. The variant you proposed (putting more emphasis on milestones) encourages SNE grinding more, which might be bad for the game quality because, unlike SN, SNE requires true commitment and zillion-tabling, no matter which limits are played (lower than 25/50 accessible to 'chosen ones' only), because as rake caps are hit almost always at high stakes, VPP/hand rates are capped too.

10 billion hands / rake average (each format) x player ability micros
10 billion hands / rake average (each format) x player ability low
10 billion hands / rake average (each format) x player ability mid
10 billion hands / rake average (each format) x player ability high
Some where in there fits what your true edge playing perfect really is in poker and how small it is.

I'm sickened, and don't have to understand entirely the burden and logistics of imposed rake to know that cost is quite predatory with such a mass volume/turnover. If you saw 9 hours of every day with out rake, what would be an outcome? It's kind of a basic answer right? Why is that? Most people even those not mathematically inclined know that their edge, their ability is not going to make them money on any level when they paying someone to play, and most understand the amount simply as high. TO what extent perhaps they do not know. But equally all players reg or not, know that doing 6 times the work load in a rake period to a non rake period is going to please them/comfort levels playing. Looking at it, one would expect the site makes a incredulous amount of money per day/per week/ per financial year. They are a business but on the back of every hand, 120 per hour in some cashes, and every game at so high of a cost as to out weigh the service provided? What position do players really have, they don't have choice Rake x %(amount) rewards, rake <% (amount) no rewards. WE have to eat it up, and take it for good or bad. What about other choice, like more % pay outs in x amounts of games that are top heavy wieghted (doesnt matter that much if rake still chews you up)? And it hurts. Poker, cards is a damn expensive game. And for sites to bash all players so heavily over the head knowing that it's their right as business and players will take it is sickening quite.