Thought I would update my 2011 results and give some thoughts on SSNL live games and my evolving opinions on win-rates, bankrolls, and finances...

March, April, First half of May 2011: 230 Hours, 35 sessions - $5700, $25ish/hr - $200ish in comps, $600 in promotions.

YTD 2011: 350 hours, 50 sessions - $6700, $20ish/hr - $300ish comps, $800ish promotions.

I continue to put in 20-30 hours a week of live 2/5NL sometimes playing 2/5 PLO or 5/10NL. And it looks as if I am on pace for another abysmal yet consistent year. As I approach 2500 hours played at the tables over the past 3 years, It is becoming clear I am a 3bb/hr winner at 2/5NL. I must have significant leaks in my game, obviously.

What shocks me the most if my lack of variance. I would assume being a small winner in the game and considering my aggressive play style. I would experience tremendous variance - yet over the course of three years I win close to 75% of my sessions.

In another results thread, I asked about confidence % of a win rate and standard deviation. Basically, I was wondering if someone could correct my logic or my thought process -

I would assume that since I have 2500 hours I would have a reasonably high confidence % in my win rate and its accuracy. However, considering I am a small winner shouldn't my standard deviation be very high? So if that is the case, why don't I experience much session by session variance?

Essentially what I am getting at, is I have logged a somewhat significant amount of hours and as an average by obviously leaky player - If i am experiencing so little variance -

I think I could propose that people should be playing 2/5NL or shot taking more aggressively? I now believe that 10-15 buyins is more then sufficient to get started playing live poker as a bankroll.

How much does a stop loss play into things? I personally don't use one which should again lead to more variance? I try and play a set # of hours per session vs. having monetary goals or restrictions.

Basically where is my variance at? Why do I suck so bad?

-- Quoted Post from 03/02/2011 BELOW

very insightful boldman. Thanks for sharing you're results.

I'll probably share mines aswell as soon as i hit a significant number of hours (~500)

Note that i am deficient at detecting sarcasm, so im assuming this post is mostly serious.

First of all, you count promotions in your hourly. Because you pay for them out of the rake/promotional drop.

second your Q and ytd seem to be 5bb/hr and 4bb/hr respectively even without counting promotions. where are you getting 3?

third: what do you think you should be making? Im not saying that what youre reporting should be considered a ceiling, but "im bad at poker because i 'only' make 5bb/hr" seems a bit glass is half empty.

fourth: where are you playing? How are the games? is there any place that offers or might offer better ones? 3/5 games are not universally v. soft. I mean theyre soft, but youre not entitled to walk into a random 2/5 game and expect to be a 10bb/hr winner even if youre quite good. If youre in Sheboigan montana or something and you 2/3 of your hours playing with 7 WT fossils you just have to grind out what you can. Could you time/game select better?
How many hours are you playing in this game
and how many hours are you playing in this one

3bb/hr was kind of just a number I did a quick glance and composed. 2009 (20/hr), 2010 (15/hr), 2011 (20/hr). I assumed that because of my lack of variance and general lack of losing streaks - that I am running well, running very well, in fact over a 2.5k hours sample. So I assumed 3bb/hr is a more accurate estimate then 4bb-5bb/hr.

However, as I said when I posted my results in both post I have made in this thread, I round everything - so the one year that is 15/hr is actually closer to 17, the year that is 20/hr is actually closer to 19/hr, this year is actually closer to 22/hr. Just to make it easy for the sake of posting on a forum, I rounded everything basically to a 0 or a 5 last number. I'm aware of the impact this makes on estimating WR accuracy. I simply, don't care that much, I find it easier for people on the forum to process things like 20 vs 19.7, etc.

I try and separate promotions and comps from calculating my hourly. I understand it is essentially "rakeback" - despite that I try and keep that tabulated separate from my table winnings. Just the way I do things.

I wasn't trying to complain about my WR, I was mainly asking how come a player with a small win-rate such a mine - hasn't experienced many large swings. My standard deviation should be high, since my skill level is low - However, it seems to be low as I experience little variance. I am drawing that conclusion as my results seem to conform to my previous years results so far.

I guess I am curious what type of variance a 5-7bb/hr player is experiencing vs. myself? Considering I feel I am consistently running well at the tables, what do my current results say about my potential earnings?

I am mainly just shocked at how little variance I seem to have encountered over 2.5k hours.

To put into perspective, I also from 2009 til black friday I had played over 2million hands of online poker. Almost all SSPLO and some SSNL. I was essentially a break even poker player at the tables. All of my profits came from rakeback. But I was still significantly profitable. So I am well aware of what is within the scope of variance and I am capable of objectively analyzing my own skill level.

I understand I have only played 60kish hands of live poker over 3 years and that in it-self is a very small sample size. Maybe I have answered my own question 2.5k hours is still a small sampling of actual hands played. So maybe the variance will come?

I don't know - I guess I am just a little miffed at my lack of quality results. I feel I should be crushing the games, but I know that I am probably exceeding my expectation and it is very frustrating....

So I ramble and second guess everything haha.

IMO with a 75% win rate (unusually high IMO) and your hourly being what it is, I think it's just your style of play.

My guess on how you play: (I'm not judging your style vs good/bad/whatever). Very textbook (rarely slowplay or get tricky/take unusual lines), don't engage in coin flips (rarely call preflop allin w/AK/mid PP's or 3bet shove w/AK), rarely bluff (check raise flop heads up w/air or dbl-trpl barrel after missed cbet or bet when 3rd flush card hits), don't shove with a draw, don't play SC's regularly, your philosophy on winning at poker is to make (passively) fewer mistakes than opponents (vs trying to pressure your opponent into mistakes).

(Once again, I'm not judging you on weather your style is right or wrong, but if someone played this way their results would mimic yours.)

You sound like you might play TAG preflop, but don't take many risks postflop. Playing this style would also keep your "huge session wins" rare and "huge session losses" rare.

I decided to reevaluate life, took a month vacation to europe alone, grinded some casino poker out there. Crushed it, paid my trip, decided to come back and give poker a more serious go.

I can't post an actual graph of my sessions bc I don't know how, but:

Since april 16, I've played:

28 sessions (100 hours) of only 1/2nl
My profit is 3153
Win rate per hour is 31 bucks

At casino only (no home games):

20 sessions (78 hours)
3231 profit
Win rate is 41 per hour

Is this a large enough sample size to determine if I'm a luck box, decent, or who cares...

P.S. No wife, no kids, 28, life is good ...

you could be a great player, but this isnt nearly a large enough sample.

I'm in the same boat except I'm two years older. I'm running very well right now and probably the same rate you are. I know it will inevitably turn, but I'm going to enjoy the ride while it lasts and just focus on getting better. Read, study, play, profit.

Pretty sure your the next Phil Ivey

Although no wife or kids is clear evidence your running good

Wow [deleted post] was completely uncalled for. If your answer is that the sample size is too small then just say so. We shouldn't be here to put poeple down who don't clearly understand the concepts they are asking questions about. That's the purpose of this forum. To discuss concepts, and gain clarity into our own game and how others look at the situations we pose.

OP I would have to agree the sample size is still too small. I would say a good trend at this point that should give you some optimism, but don't let the optimism make you get complacent and stop learning and continuing to improve.

Pretty Much this. i was going to say the same thing, pretty uncalled for.

What is an adequate sample size? 200 hours? 500? 1000?

Bout tree-fiddy.

I think about 1,000 hrs will give you a good idea of win rate. After about 300-500 hrs you should have a good idea if you are a winning or loosing player. As you go from 1/2 to 2/5 live, you should see a big difference in talent and impact on your win rate.
Best of luck to you. Although 1/2 is filled with bad players, it is not necessarily easy to beat. Study the game and strive for continuous improvement.

Here's a question to ask yourself: Are you bluffing? If not, and you're relying on making hands, then I'd say you're simply running good. Because if you're not bluffing, then you are getting EXTREMELY lucky to be making so many strong hands to make that much profit in a short amount of time.

Try playing an 8 hour session w/out making any sets or better, going 5 orbits at a time w/out playing a hand. If that hasn't happened to you for at least 20 hours worth of play, I'd say you are running good.

Thanks for responses....

I'm definitely making lots of mistakes still...

Ya, sample too small... Def bluffing, def making hands though...

yea i agree with yoda. i mean we can all honestly look at our sessions and rate them to see if we played good or bad and if we got lucky or not. I would also throw in if we are folding hannds like TP or better and we usually right then its probably more skill then luck, if we are never folding those hands and it just happened to hold then it might be closer to 50/50.

i use this, but i use the "lite" version b/c its free... don't recall what the paid version offers that lite doesn't, but lite offers different types of graphs, winrates, hours logged etc

i also feel way too comfortable at 1/2 to move up. still working on my game, trying to log more hours, better sample size, etc... but eventually gotta baller up. i definitely don't have the roll right now... if i can net 10k in 1/2, then regardless of hours logged, i will assume that means i can beat the game and my skills are good enough to move up. without a good roll, i fear that i will get run over b/c the money is too important.

Sup guys. Another grinder here hoping to learn and contribute. Good luck.

Does anyone actually measure their own variance? I'm enough of a geek that i'm starting to try to. Each hour i count my stack and get an hourly result. Nerdy, but all in the name of science.

What i can't yet get my mind around is: If you double the period of measurement, you clearly don't double the observed std dev. So grouping the same data in fewer periods of longer duration would makes the std dev per actual unit of time go down, right?

To put it another way it's not clear to me how you go from std dev by the hour to std dev for each, say, 12 hour session. Just multiplying by 12 is not appropriate.

Any stats wizards who can help?

i have about 2100 right now for poker, is this around enough of a bankroll for 1/2 NL? i could make the 20 buy in rule work if i came in for 100, but than im shortstacked. so would it be better to come in for 200?

I don't know your winrate. I could be obnoxious and just assume it's -$0.01/hour, and so no, $2100 isn't enough. But how about you tell us your estimated winrate instead?

Well, think about it. What is standard deviation? Its the average distance from the mean.

We know that, as our sample size grows, our results tend to more accurately reflect our actual winrate. That is to say, its extremely common to have an especially good or bad hour. Less common to have an especially good or bad week, much less common to have an especially good or bad month, etc. So in general standard deviation /x unit of time is lower as x is longer.'

To figure out what it is for the specific period you want, you just re-plug the results into the sd equation... theres no way to add them up.