Here is an interesting experiment: Post how long you have been playing in the same poker room and how many winning regulars have stayed constant for 2+ years
I've been playing in the same room for five years, and in the same area for nine. There are three pros who have stayed constant, two losing/breakeven regulars, and two whales. There are around five people who used to play 2/5 but have dropped down to 1/3 or 1/2. Everyone else has turned over.

my point: even those who can make 50k a year will have years they make 35k and years they make 80k, and will probably live their life like t hey make 80k every year. When that bad year hits... there they go, out the door

No winning recreational regs?

I recently came back to a poker room after a 4 year break. out of the dozen or so full time grinders I remember there is only 1 left along with a new batch of full time pros.
I suspect they will be gone and replaced by new guys in another 3 years.

Majority of live pros is whoever is running hot at the time however no matter how good they are eventually the variance gets 95% of them and they get washed out of poker. DGAF talks about this a lot on his podcast.

Here is a more realistic experiment:

Note how many people actually track their winning/losing.

I know of two people.

If no one is tracking, perceptions are pretty irrelevant.

Here is a secret:

A lot of "pros" do not rely on poker as sole income.

Those who do...you should almost feel bad for them. It's a pretty tough grind.

I believe the document you'll want to look at is this one which is a bit more indepth: https://www.govinfo.gov/content/pkg/...cr-00414-1.pdf

If you look at figure 4 on page 27 it does appear that only the top 5 to 10% have a positive winrate. The statistician stated:

The dissenting econometrician who was trying to show that skill could not predominate over luck noted:

The numbers can start to make things a bit confusing here:

My problem with this study as it relates to our discussion is that this winrate data in the graphs was for 5/T online.

The statistician noted:

In regards to our winrate discussion, I completely disagree. The differences between 5/T online and live low stakes are stark. The greatest single skill one could have at 5/T online is probably game selection. While game selection is important in all formats, it's not nearly as important in live low stakes. LLSNL is a mostly casual game. 5/T online is a lot of dickwaving by grinders trying to move up the ranks to become the best. The margins in 5/T online are small, the margins in LLSNL are huge. Rake is higher in live poker. Also, online pros receive a percentage of rake back. Perhaps not so much in 5/T but certainly in the lower limits many of the "losing" pros were making quite a good living off rakeback.

Online, players multi-table which results in a higher percentage of grinders at any given table. I'm not sure how long a rec player or whale will play in an online session vs a live session (or in a given year). There are just a lot of different dynamics, such as the guy who is just walking through the casino that decides to play some poker.

The style of play is completely different. Live players have a much higher VPIP than online players. There is a lot more limping and cold-calling and way less 3-betting. Furthermore, I'm guessing the 5/T games referenced in this study are actually 6max games rather than full ring.

My point is simply that this data is not necessarily super relevant to LLSNL winrates. I'd expect the spread in Figure 4 to be even wider with winning players winning even more money and losing players losing even more in LLSNL. It also wouldn't surprise me if the top portion of the graph moved higher, resulting in an even higher percentage of winning players in live poker.

A player could have a bad couple months and blow his roll. Or a player could have a bad couple months, blow his roll get a job for a bit to build his roll back up. You don't know if that guy had a job, or was just taking a "vacation" or was playing somewhere else during that time or simply playing different hours than you.

You dont suddenly bust your roll in a couple of months, even when running like shyt.

If that happens you are either

1) Not sufficiently bankrolled, aka playing underrolled

2) Playing higher stakes than you are rolled for

3) Or, you simply have huge leaks in your game that is being exposed when running bad like tilting, weak mental game and so forth.

Grinding poker has a high turnover rate regardless. Grinders like to assume they are god's gift to poker and that past pros simply couldn't hang but I've known pros who have quit while at the top of the game. I've also known others who failed to adapt and whose margins were likely cut significantly so they left. I've known others who simply had unrealistic expectations for how much one could make in poker or simply didn't enjoy the lifestyle of a poker pro. Of course there are also many sunrunners who are here today and gone tomorrow. Still, poker is not a great long term career path for nearly anyone, regardless of ability.

Welcome back to reality where the vast majority of LLSNL grinders are under-rolled and have huge leaks.

Sure, but then you arent a fulltime player/pro or even a steady winning player in the first place.

No properly skilled winning player with a stable mental game busts his roll in a couple of months. That is what fish and degens do.

Theoretically that sounds great but in real life pros have all sorts of degenerate qualities. A large percentage of them are underrolled and/or are staked.

Part of my point though is that we can't necessarily differentiate between who is a pro nor how good someone is in all aspects of poker because the sample size we play with other players in live poker is so minuscule. Most rooms run multiple tables, our playing schedule is unlikely to be identical with many other grinders, and a large percentage of grinders play at multiple rooms.

In other words we really aren't going to have a very good idea of how much the other players in the room are actually winning/losing (even if we were to track exactly how much other players win/lose ever session which no one does). For the average grinder that you see who disappears in a few years you don't know how much that player actually made playing poker or necessarily if he was even a winning player. You aren't necessarily privy to all the reasons he quit poker, or even if he actually quit poker at all (perhaps he simply isn't playing where you play anymore).

A guy that grinds 40hrs a week at the poker table could easily be supporting himself with a full time job and you wouldn't even know it. The water is further muddied because players like engaging in dick waving contests both on the table and in the forums where they will flat out lie about results to further inflate their own egos.

Successful poker players don't remain in poker, even those who do move up.

They diversify so that they don't have to play poker for living.

IMO, nobody wants to play poker for living. Unless you're Postle and winning is guaranteed.

What I derived from these posts is that intelligence, skill, patience, discipline, and hard work are each important components to being a successful 2/5 NL grinder. Aren't these important at all levels of poker though? If you want to be able to make millions beating the biggest games in poker you just need to work hard, have a certain level of intelligence, and not have any major mental game leaks.

The problem is that everyone thinks they are intelligent and skilled. They are not. Everyone thinks they will put the work in. They won't. I don't even think most new pros even fully consider how other things like mental game will negatively impact their results.

Some posters paint a picture like becoming a successful professional poker player is easy. All you gotta do is work hard and have a certain level of intelligence. Easy enough right? Wrong. Most players will fail in this endeavor (at least that is my opinion based on little more than anecdotal evidence).

I never said it was easy though, did I? Having all those traits is hard to find.

Idk why everyone assumes someone who stops playing is broke lmao. I played full time, just thinking about playing now makes me want to vomit. Poker would be the perfect sideline for me too, but I just can't get myself the motivation to play again.

I made more in 1 real estate transaction than I did in almost 2 full years grinding full time. What a waste of time

Getting chippy and off topic in here. Take it easy, all.

The analysis for this legal case is super interesting. Makes me want to buy a few million hands and do similar analysis (am econometrician).

However, I don't think figure 4 is actually helpful for discussion of distribution of winrates. The expert came up with his own cute way of predicting skill level based on specific summaries of play style (I think) then drew that graph based on where the players are in that prediction. It sloped upwards, which means there's a good positive correlation between his prediction of skill and winrate. Good for him, the index he created actually predicts winrate to some extent. But it doesn't actually mean only that many players have a positive winrate. His predicted top tier can still have garbage players in it, pulling the positive winrate to zero and the predicted low tier can still have winners in it pulling the negative winrates up. What it does show though is that skillful play can be described mathematically to some extent, which is important in proving that poker is a skill game. If you said "it's skill, but I can't really say what the skill is..." then you'd have a bad time in court.

+1 million. If you have the combination of intelligence and discipline necessary to kill at poker you can make a lot of low/no variance money in a profession.

I haven't met many solid consistent winners who I didn't think had the aptitude to do something else for just as much or more money. They just ended up playing more poker because they got tired of the professional grind and having a boss or didn't want to/couldn't afford to hike up the experience/education ladder. My guess is they probably could've found a better employer when they were annoyed by the office grind or gotten a premium education but poker was too tempting. Grass is always greener blah blah.

I semi-pro'd for a year playing online 15-20hrs a week 2/4 and 3/6 limit 6-max pre black Friday (note: lol what a waste of time, the effort involved to kill limit at that level would've paid back so much more in no limit). My skill level increased rapidly because I worked up from .25/.50 to those stakes so sample size wasn't massive but I crunched near the end that my hourly at that level was $40+ or - $10. Went to graduate school and it took me 5 years (lulz PhD) to get back to that hourly wage but with a few years of work experience I'm way past that in guaranteed hourly and the risky part of my compensation is stock so even non-guaranteed money won't get "coolered" unless we have another great recession, it'll just not be as much as I want. Opportunity cost of that time in school would've been high if not for black Friday, but now I have a much better way to cover my nut than poker.

The turnover is admittedly concerning. I’ve only been doing this for 1.5 years so I haven’t really seen it but I hear the conversations between pros about how few people are left from 3-5 years ago.

I hope I’m not burying my head in the sand when I think I won’t be part of that statistic.

Thanks for this info. I was somewhat skeptical about how he was able to determine who the skilled players were.

Have a backup plan and have a separate life roll. Leave yourself an out

By fortuitous coincidence, the 95th percentile is very close to the two-sigma level of the positive side of the normal (Gaussian) distribution. This means that breakeven is two sigmas above the average loss rate of all players.

And what is the average loss rate for all players? It is, quite simply, what every player pays on average in rake.

In my local 2-3-5 game, $6 gets taken out of every pot that sees a flop. Something like 30 hands per hour are dealt, so $180/hour goes down the slot. For a ten-handed game, this means each player is losing $18/hour, or 3.6 big blinds/hour.

95% of players lose, so the two-sigma mark is at breakeven, so a good guess at a distribution of winrates for the player pool is a Gaussian whose mean is at -3.6 bb/hour and whose standard deviation is 1.8 bb/hr.

p(w) = 1/[1.8* sqrt(2* pi)] * exp[(w + 3.6)^2/(2 * 1.8^2)]

Suppose someone has played a specific number of hours of the game, and has wons a specific number of big blinds. The frequentist best guess for their winrate is the total bets won divided by the total hours played. But if we are good Bayesians, we can use this probability distribution of winrates to compute the most likely true winrate for the player, given this distribution and their measured results, using the methodology outlined on pp. 40-43 of The Mathematics of Poker by Chen and Ankenman.

No problem.

The more I read this the more interesting it is. DeRosa's arguments do not appear to be intellectually honest to me.

Figure 7 of cumulative winning from coinflips is hilarious because in one way it's dumb but an another way it's completely correct. Heeb's figure 1 of cumulative winnings and the discussion around it was overly simplistic, he walked right into the figure 7 counter argument. Both figure 7 and figure 1 are plots of random walks. So yeah, they effing look the similar if you take the top and bottom winrates. The thing is that figure 1 are a bunch of random walks with drift, drift being the winrates while figure 7 is a bunch of random walks with no drift so the spread out is random and any given upward line can suddenly crash downward forever just as likely as it would keep shooting upward. In other words, take any group of "consistent winners" from the coin flipping game and now just continue the plot from the end, they'll spread out randomly again with some looking like winners and some losers from the high starting point because they are perfect random walks just spreading outward randomly. If you took the consistent winners from poker in figure 1 then continued their plots they will continue on being positive unless the players go on mega tilt or take shots they should not be. That's because their time series are random walks plus drift, with the drift being their skill.

Lol at p36 about DeRosa not discussing Heeb's analysis showing winning players won money with QsJs while losing players lost money with it. Just lol.

DeRosa's argument that the skilled player not being guaranteed to win every hand means poker is not a skill game is some of the most incredible nonsense I've ever read from an expert testimony. He couldn't possibly actually think that.

The argument on page 45 that since 75% of hands end without showdown that skill has to be involved is excellent and the government counter to it is ridiculous. They say because poker decisions revolve around random cards that it's all chance. Football game starts with a coin flip, doesn't negate the skill of rest of the game because decisions were made around that flip.

Good approximation but be careful with it. The % of winners is partially determined by the rake. Wherever you are getting that %-tile breakeven number, it has to be from the exact same game you are using the rake number for to back out that standard deviation. Someone looking at a different raked game could come up with a very different answer if they use the same % tile.