Did they even
Did they even

Baard,

I completely understand why you do not with to participate in this thread (and think that's completely fine if you wish to choose so). So I completely understand if this gets no response.

However, I would like to point out that my example was not a reference to the particular discussion that you are thinking. It was a description of possible systematic errors that I believe are present to certain extents.

Furthermore, what you are saying does not make sense:

because the denominator in would be negative it you looked at ALL the players.

So clearly, the only players you are including in "player winnings" are "winners" which produces exactly the biases that I described in my example.

This is simple to see: if in my example the EV winrate was -1.4bb/100, after 500k hands the output would still be ~1/6 players being a "winner" despite nobody actually beating the game.

No that's not what everyone wants. (besides Pokerstars and a few posters who think no one should win at poker)

Losers are supposed to lose and winners are supposed to win. That's part of the game.

If you say that recs should be put in an environment where external programs can't hunt them down. Then I agree with you.

If you say that recs deserve to play at tables where everyone isn't timing out everyhand. Then I agree with you.

If you say that players shouldn't be able to look at an external program to gauge their stats. Then I partially agree with you/fully agree with you.

But this absurd notion that recs should survive for the sake of it. No I don't agree with that. And if you think so then you should be demanding Pstars to get rid of Spin and Go's.

I honestly don't believe that they are modelling predictive data or running any advanced forecasts or sims to come to these conclusions. They may have key performance indicators on certain reports which they run on a routine basis that senior management hardly ever look at unless they want to change something - but the decisions that they've come to consistently over the past 2-4 years have been based on this simple business approach; Online poker is declining*, how do we counter as a business and their strat is simple. i) What money are the top X players extracting from the player pool - ii) how can we retain that money - iii) how can we rake everyone harder to compensate our decline in business.

I truly can't believe they are reviewing "consumer experience journeys" by splitting a player bell curve in 10% segments over 3 year sample sizes, by stakes, and reviewing how their proposed changes affect each segment of player bracket and then determining the best course of action overall.

We've seen this same tactic consistently from them when GGARJ & Joeri rep'd the PLO community at separate player meetings with the onus being on addressing the extortionate plo rake.

This same model was used, and its used as a mechanism to state; this is how fast the top sprinters on the planet are running, subsequently, on average everyone else runs this fast. Outting this data to the poker community, doesn't harm their business in any way or form... they own 66% of the market.

They need to hire a whole bunch of young poker players who have given their life to online poker and go through a massive culture overhaul and come to the table with a much more transparent approach.

* via players getting better, bots, net depositors shrinkage.

Beat me to it.

We're circling around the question, of course, of "just what is a winning player"?. And while I don't have much info to go on, it appears that the definition that was used was "anyone who was net positive in <some 12 month period>". And DD seems to be absolutely correct in that that's exactly as clear-cut a sampling bias as the more extreme top 2% / 4 BB/100 version from my previous post. You've just chosen $0 as the "cutoff"(!)

I'm not trying to goad you into joining the fray here (unless you think there's something of value you can add). Moreover, I know several folks who work at PokerStars who I'd consider quite sharp, and some of this stuff is so elementary that I'd be inclined to think there's simply no way that Intro To Stats-type errors such as these would have gone unnoticed. That said, I'd invite you to engage in one of the most effective exercises I know of to identify confirmation biases: call your team together and require everyone to brainstorm reasons why your calculations may be wrong. What underlying assumptions did you make that may not be true? What unforeseen circumstances might invalidate your entire model? Maybe you'll emerge more convinced than ever that you got it right the first time around? That's great. But as no stranger to these undertakings (and having witnessed even smart Stars folks get things glaringly wrong in instances past), I think you'll find it worthwhile.

Looks to me like Amaya execs just said to PokerStars guys "There must be xxx of cuts". Forget the rubbish that's peddled about "ecology" it is just to make Amaya financial reports look better (liabilities down etc - think devaluing of FPPs to StarsCoin).

PokerStars guys then had to make these cuts, in any way that they could find. I presume there was debate over where they could make them but there was no escaping the savageness that had to occur for Amaya.

But there will still be winners and losers. How on earth do you arrive at that conclusion? At a table full of new players/recs, someone wins and someone loses ffs.

The only difference is they won't be slaughtered before they even learn how to walk.

jfc you'd make Bernie Sanders blush.

What you are proposing is putting equally skilled players at the same tables. Of course on any given day someone will win and someone will lose. But the next day the winner will be the loser and vice versa. No one will be able to win consistently which means that Pstars will get all the deposits turned into rake... or at least a substantially higher percentage.

Explain to me how Pstars getting richer is going to help a pro or a rec or anyone in between? But kindly do it in another thread.

Since when are all net depositors of equal skill?

Actually in his example if you read it clearly 'winnings' must be a positive number by definition. I know that you are trying to claim losing as 'negative' winnings but you don't understand the equation then.

One way to prove that this will improve the poker ecology is to look at US sites who removed ALL rewards. Take Bovada/Bodog for example. The site became so fishy because over half the pros could not make a living there. The rec model flourished and they went from a low traffic site to biggest in america and likely 2nd biggest site in the world. Bovada is so fishy that people on this forum get mad if you post that it is.

What Chinamaniac said is a hard hard lesson we/all us players had to learn a long long time ago. Most regs I know in US don't play anymore simply because they can not make money on bovada. If they could they would but they were so dependant on rewards that that was over 80% of their income. The only way I can play and still make a cut is because I had to study like 15+ hours a week post blackfriday just to make 50% of what I make pre bf. It was a lot of work and no fun but I am 5 times the player I was then.

Take China for example. He is probably the strongest pro I have ever encountered in my games. He murders US sites and Merge for example has the exact same rake structure as poker stars except zero rewards. And rake is rake like he said. I know 'net effective' rake is what matters to you but the definition of rake can be found by looking at the rake charts period.

How do you think he murders the games? By grinding tons of hands and posting on 2+2? No, he had to work really really hard and get way better at poker than most if not all of the people in the thread. That's why he doesn't feel bad for any of you guys.

It's not breech of contract or unethical. Go ahead and hire a lawyer, but don't you think someone woulda taken the case by now? It's because there's no case. The reward structure can be changed at any time and you agreed on that by clicking t&c. You agreed to it.

The top of the iceberg will be trimmed for sure and this will surely slim down the market of pros. Only the excellent pros will survive and many of those will have to drop limits. But the site will get fishier as bovada and all US sites did. This is an improvement in the ecology. And like Chinamaniac said, rewards will surely be cut further once they see this positive effect on ecology.

You can post away or study. It's your choice.

Obviously they are not... but creating a group of tables for 1 tablers is going to have a much higher percentage of recreational players then regular tables.

On average players that 1 table are far worse then mass tablers. Creating this environment would greatly increase the amount of hands required for wealth to shift on the regular tables, meaning much more rake paid.

I don't think its a case of the stats guys being wrong. or missing something, I would guess if the stats guys were running the company they'd probably improve the longevity of the games and be thinking outside of the box and bringing some creativity to table. And there wouldn't be as much animosity from regs.

I'm sure the team is really intelligent and statistically gifted, with a solid grasp of poker and whats needed.

That said, if leadership doesn't have the right vision, be able to think outside of the box and put themselves in the "consumer" shoes... its gonna go wrong. I would bet my left nut, the leadership team wanted x changes, and the stats guys were asked to paint this picture. When it should really be the other way round.

At the moment they are only putting themselves in the "shareholders" shoes and they've made this massive oversight regarding loyalty rewards in 2016... which has a business critical impact on "pseudo rake" - it just blows my mind soo hard that they aren't thinking about pseudo rake implications.

Nobody has asked a lawyer to take the case afaik. No reason to believe a litigation lawyer would decline it.

Small print on some internet page doesn't allow you to act wholly unreasonably.

When you check in to a hotel you sign a registration card accepting their t&c. Do you think they could just bury in there a clause that lets them increase the charge for your second night to anything they wanted? If they demanded $100,000 for the second night would you just pay it and say "oh well, silly me, i ticked their t&c. I agreed to it."?

This is a bunch of garbage. Bodog has an massive established sports book and casino. These are vastly different situations.

Also, not sure how the Merge ecology can be called healthy when the traffic is like 20% of of what it it once was.

I am not proposing putting players anywhere. I am simply suggesting giving them an option.

They are free to go wherever they want.

Hey Baard As others have pointed out, the metric you describe has all the biases accused of it when looking at high stakes games. To put it in the simplest way possible, if you have 20 regs of equal skill and 4 recreational players who play 50/100 during the year, the correct method for determining how much money the regs are winning from the games would be to take the total winnings of the 20 players to determine how much the regs win and to take the total losses of the 4 recreational players to determine how much money is lost.

However, because of sample size and variance, the method you described does not work. The samples at 50/100 will be such that 5 of the equally skilled pros may run bad enough to lose money for the year at 50/100. Therefore you overestimate the winnings of the pros by cutting out the ones who ran the worst, giving you an overestimate of the average expected winrate of "winning players" in the game. Also, you overestimate the amount of money that is put into the games by losing players (because you are adding losses of the winning players who ran bad).

Putting these two things together gives you a very distorted picture and not at all what you are claiming to be looking at. This should be clear if you imagine multiplying the sample by 50. If your method were correct, you would have the same ratios. However, they would of course be different because now all the winning pros would be winning and the average winrate of players who actually won would be lower than your estimate of average winrate. To make it even clearer, imagine using your method over the course of a single session. The average winrate for "winning players" that it gave would be absurdly high.

So the method gets more accurate as the sample increases. 1 year for high stakes games is just not a big enough sample for this to be accurate though. There will be many players who were +EV but took failed shots etc adding to your "money lost by losers" with the ones who ran better staying in the games and adding to "money won by winners". If you have done analysis that shows the ratio gives answers close to the ratio that would be gained if you did it properly then fair enough (although why not just do it properly in that case) but if there is really just nobody at the company who understands the skew that is being put into the results using this method and it is actually being used to make important decisions, that's a really sad state of affairs

Apologies if I am just misunderstanding what you are using this metric for exactly and you realise it's limitations but I don't get that impression from your post and the posts from the people who were at the meeting.

It was really Monorail I thought was misunderstanding since he suggested that we only were looking at the top 2% winning players..

And regarding the method.. looking at the results of all the players using the formula mentioned makes little sense. The results of the whole population = rake and would therefore always be -1, when disregarding rewards.

Since Player Losses = Rake + Player Winnings, it seems to me that the formula gives a good representation of how the losers money are split between the cardroom and the winning players. And since the sample is all the players (over a long period), I think it can be argued that the amount won by the winners in the sample, should be very close to the combined EV of all the +EV players in this group.

I will be the first to admit that the method has flaws. For example, it does not take into account the effect the Player Winnings will have on other games. This is why we only used the numbers as an argument for why we removed rewards from the high stakes games, and not to say anything about the poker ecology as a whole.

Thanks,
Baard

Edit:

After reading Kanu's post (Hello Alex ), I realize that I should have added that the number of hands covered by the sample needs to be large for the metric to be useful. In the meeting, we looked at data from 5/10 and upwards, and at least that stake have a very significant number of hands played during a year.

I think "ONE YEAR!" just sounds like a big sample to everyone involved with Poker Stars.

Limiting table caps to one or two cash would be better. Then you could keep your hud which they say doesn't do much anyway...

congrats on the massive score btw

I'm going to give you a pass on this one, Baard, because I know you to be an otherwise reasonable thinker. But come on: ANY reasonable reader of my post would very clearly understand that the #'s I was using were purely hypothetical (which I stated repeatedly), in trying to understand just what type of error DD was describing. Using the 'absurd extreme' is just an easy way to do that (and I know you know that, so let's leave the silly straw men out of this.) Regardless, Kanu articulated in much clearer fashion than I did the problems with the methodology. In any case, thanks for popping into this thread, even briefly. As I've written several times: I'm not expecting perfection on this, but I am searching for signs of reasoned analysis. Any continued participation here by you or anyone else is very much appreciated, despite the occasional disagreement.

Surely in any given year at any given stake there will be a load of +EV regs who played between a few hundred and a few thousand hands but lost? At high stakes where samples are smaller this will probably have a more significant impact on your ratio. Perhaps the problem is summed up in your "(over a long period)". It's not a long period in terms of the variance involved and many of the players in your sample will probably have played a very small sample of hands.

Meh, after reading your edit, I guess it'll be hard to get anywhere without looking at the data in a bit of detail and trying to work out what the ratio should be compared to what the method gives. It sounds like the guys who visited were unconvinced but at least you have some awareness of the limitations.

One of the great all-time 2+2 posts (would be great if someone could dig it up) that most contributed to my understanding of variance was a very detailed statistical post by a poster skilled in math showing that what we think of as the "long term" is actually laughably short. The standard 2+2 response to any poster asking about winrates was "you need min 100,000 hands before you can tell anything". Then this math guy poster made a great detailed post showing that none of us will live long enough to even approach the kind of volume where we could state our "true" winrate with any degree of certainty. I'm making the following #'s up (< take note, Baard) since I don't remember them, but the poster made a very persuasive #-backed argument that if you had a 4 PTBB/100 winrate over 250,000 hands, there was still something like a 15% chance that you were actually a losing player. Would be great if someone also remembered this and could find it.

But yeah: "ONEYEARZOMG!!!!!!" is a scary small sample size.

EDIT: I see Kanu jost posted something to the same effect. Yep: sample size is a huge issue here, and will significantly contribute to / exacerbate any sampling bias.