discuss

[Subfallen]

I don't think mike is saying that LHE cannot be +EV, just that the variance is ridiculously high.

[B1GF1SHY]

Maybe still +EV but it contains a whole ton of luck. It's so up and down it's crazy, silly crapshoot [img]/images/graemlins/smile.gif[/img]

[Bluffoon]

[ QUOTE ]
i posted privately to a friend but i might as well put it out here.

what online is teaching us, where seemingly good winning players can "run bad" for 100k hands, and not know anything with confidence until they reach a million hands, is that the long term is so very long term in limit hold em because luck is a much bigger factor than previously thought (or than "experts" have let on or even realised).

in fact, once you get beyond some very simple concepts like degrees of tightness and the logic of basic postflop play, the game is primarly luck. how you play 55 on the button after two limpers, what you do w/ KQ utg (or KJ for that matter), and so on and so forth are meaningless. all that matters is how you ran that night for a couple hours and whether youre smart enough to go spend your winnings on something worthwhile or chump enough to squander it away through another period of "running bad".

im calling bullsh*t on the whole thing. the only people ever sure to make money are the ones collecting the rake. and ive won every year ive played, more and more each year, for the past 5 years. and that's meaningless.

[/ QUOTE ]

If I were to graph my win rate I sense that I would get a line with up and downswings (little squiggles) lasting three to ten thousand hands or so with the downswings being shorter than the upswings in a ratio of about two to one in aggregate.

Going longer term I sense that I would see longer term variations (bigger squiggles) in my win rate lasting thirty to fifty thousand hands.

What Mike is saying is that if you play long enough you will see monster variations in your win rate lasting three to five hundred thousand hands. When you eventually hit this downswing you are going down and there is nothing you can do about it.

Ouch.

[Tommy Angelo]

Here’s a few other things to consider mike.

How much to do you think is raked/collected/tipped out of all poker games in the world each year. I'm talking every online game, every public B&M game, and every home game, in the world. Let's say it's one billion dollars per year. This means that the hugest +EV play in the history of earth is available to us right now, the poker players. If we thought of ourselves as a single organism, and we all quit playing for life, today, it would be a billion-dollar per-year swing.

Another thing. Add up how much in rake and tips that you, mike, have paid in the last five years. I’m guessing it’s about $100,000. Now add that number to your actual "winnings" as you calculate them. That sum is the true measure, for it is the only measure, of just how “lucky” you have been. If there’s one time to be results oriented, it’s when you’re talking about results. This idea you have in your head about needing to play millions of hands in order to determine what "should" happen in the next millions of hands is just goofy. The only way I know to keep score, in any game, is by looking at what happened, not by conjecturing about what might happen next.

Here’s another thing. Hold’em didn’t used to exist. And it used to have one blind. Now it has two blinds. Tomorrow it might have three. Right now hold’em is structured well for the house. Tomorrow it might be structured better for the house, meaning the affect of “skill” would be less, meaning the money would stay in motion on the table longer, meaning more of it would go down the hole, faster.

What if the house just raked everyone’s money before the first hand was even dealt and said okay, that’s it, everybody go home?

That’s what all public poker is mike. Just like that. The rate of rake is a balance between what the house can get, and what we’ll give. The game itself is just a product for sale.

The perfect arrangement for the house would be if nobody ever won, but everyone said they did. I’ve played in games like that before. But of course I’m the guy who really DID win!


Tommy

[gaming_mouse]

Mike,

You are overstating your case by quite a bit. This question has been answered many times in the Probability forum, but I'll do calcualtion here, since you mentioned something about the math.

Typical limit SD is between 14 and 17 BB/100. Let's call it 15. Then after 100K hands, the sample win rate of a player whose true win rate is 1 BB/100 has a normal distibution with:

mean = 1
SD = 15/sqrt(1000) = .47

This means that 95% of the time the player's sample win rate will be within .94 BB of his true win rate. In addition, there is only about a 1.6% chance that such a player will be even or losing after 100K hands. This kind of evidence is very far from meaningless, IMO.

gm

[TStoneMBD]

i really dont understand what the whole "hoohaa" is all about here. maybe youre right, poker really does suck. i dont know. all i know is all the money ive won during my short lived career. it may be miniscule to the people on here making 200k a year however, even though its only because they are all on the better side of variance than i am, according to you.

youre right though. you shouldnt base your future earn on what youre currently earning in poker right now. variance is fearce. when someone asks what it is that you do for a living, dont say poker player. say that youre retired, for thats what you are. you have money. you play poker on the side. until that money runs dry you are retired. look at it from that perspective.

[glorfindel]

Here's another way of looking at it mathematically, maybe. Correct me if I'm wrong, gaming mouse, but I'm working on understanding this standard deviation and variance stuff.

Assumptions:

100 hands = 1 hour of play.
SD for 1 hour of play = 15BB.
expectation for 1 hour of play = 1BB.
number of hands in 1000 hours = 100,000.
2SD = 95% confidence interval.

SD for 1000 hours = 15BB * sqrt(1000 hours) = 474 BB

There is a 95% chance that after 1000 hours of 100 hands per hour that a 1BB/hour player's results will end up between

1000BB - (474BB * 2) and 1000BB + (474 * 2) or
52BB and 1948BB.

If I did this correctly, it's clear that a 1BB/100 player should not be losing, but he could easily not be winning very much after 100,000 hands.

[JAA]

Well, are there any winning players on here who have logged a losing year (assuming they played a fair amount over that year)? I suppose that this would provide support for your argument.

I concede that how I play 77 after 2 limpers on the button with tight blinds is trivial, but I think the mistakes that your opponents make add up faster than you are giving it credit for. We make the vast majority of our money from the latter, not the former.

- Jags

[mikelow]

So is the standard deviation that huge?

I think you are a winning player and you know it. Skill has to take over at some point.

What might be happening, is that there are so many players with some skills, but not that good. So if there are fewer fish, it's harder to win, and the standard deviation might not go up that much but your edge is less and less.

[Turning Stone Pro]

I agree to some extent with Mike. However, I think that game selection and quality of competition is being lost in this discussion. This is the single biggest factor when trying to figure out standard deviation, what is the 'long run', etc., whether someone is a winning or losing player, etc.

I am not a mathmetician. Never will be, never have claimed to be. However, I know that if I play with players that are better than me, make better decisions than me, understand pot odds and things better than me, read me better than I read them, I will lose.

If I play with people, each one my exact speed, I will lose, because of the rake.

If I play with people who, on average, are slightly worse then me, I will make a little, not too much, because of the rake and luck and standard deviation.

However, if I sit down at a table with several inferior players, who do not have the experience, knowledge, and training that I do, and are concerned about things like grocery money and rent and the like, I will beat them. There may be some short-term downswings, but I will beat them. This math I understand.

It's all about game selection.

TSP