Some 'long term' questions for the big boys.

So... been a fair bit of interesting talk on the forums lately and even though i've posted a lot lately, the questions keep coming faster than the answers.

This post here sparked my imagination a lot, i'm sure you've all read it (the one with the variance calculator):

http://www.flopturnriver.com/phpBB2/...99.html#871501

First reaction was like: wow!! incredible, i'm going to win at poker even though my graph is a bit flat, 23k hands REALLY is insignificant. I finally understand now. Just keep getting the hands in etc.

Then, through that post span off other links and articles about SD or 'standard deviation' and what a downswing really is and how common they are AND how EVIL they can be.... even though you're a top winning player... how frequent they are, how loooong they last etc. etc. etc.

In short... question is.... so why are we playing poker again?

In detail:

1. If variance is SO long term, we could be winning losing players for months or more (i.e. poor players with good variance) or vice versa, great players with shit variance over 200k hands or so. Just how can we every know if we're on par or doing well or how can we know anything!?

2. I'd love to see some REAL long term stats. Would any mid stake hard core grinders care to show what a real graph can look like over a long sample?

3. For future reference... those of you who earn a living from poker.... specifically what do you do to pay the bills if you have one or two break even (or worse) months?

4. SD and variance in mind... what is the ratio like for long term, successful mid stakers between those shit months and the good months?

Thank you kindly. Following the white rabbit through Wonderland.

oh dat FTR search function

kinda like the slightly chubby kinda cute girl in high school: she's always there, we take her for granted, yet she satisfies your needs.

.... so yeah, as i was saying...

ok. For graphs you should use the search function here on FTR. Also, variance happens, but if you are beating the stake you're at by a big enough margin - even the 1/100 -ve variance streak could be breakeven or better. Go to that site (i can't be bothered) and note that a 3ptBB/100 winner always does pretty well over 100k+ samples.

as for months - again, it depends on your true win-rate (including bonuses/rakeback/etc) and your hand volumes. I haven't had a losing month yet, but i have had plenty of losing 40k hand stretches. And this is from someone who is only a marginally above break-even player at the stakes played.

Re playing pro and paying bills etc. Most pro players play off a very deep roll for the stakes they play and have savings set aside to cover these sorts of variance situations. I finish work in two weeks. I have enough money in savings to travel for a few years if I want (thanks poker!), and like to run a poker bankroll of 60-100 buyins. If I somehow have a breakeven or losing month I'll still be able to withdraw 20buyins and that's fine to live on for that month - without even thinking about dipping into savings...

I'm sure many of the other grinders have graphs that look like this, it can be done:

http://img23.imageshack.us/img23/3296/100khands.png

If you look there's a few break even spots over 5K and even 10K hands. And this is still from having a win rate within a range of 4ptbb-9ptbb/100 (only 2ptbb/100 during my short stack experiment). Not saying that these little stretches were purely variance, as I've found that I played pretty bad during them, but there of course were a few coolers in there which stimulated some poor play. There's also multiple sharp declines in the graph from getting stacked deep more than once in a row, but it's hard to tell exactly because the graph smoothes out a bit over a large sample.

If you look at a shorter term graph though, it's more swongy.
This is April so far, which is actually a mix of $25nl and $10nl, so I actually have a positive win rate here depsite the fact that I'm just under break-even because my $25nl results (and some $10nl) have put me below positive:

http://img23.imageshack.us/img23/8300/aprils.png

I wouldn't call this purely variance though. I went back and discovered a few stack offs that I wasn't too pleased with.
Also I wouldn't worry about what your standard deviation and variance figures are. Just learn the game and practice good bankroll management and you'll be fine. Also, don't be to quick to blame your losses on variance.

Cool. Thanks very much for showing this. It's that significant leap from theoretical statistics to actual living examples that make all the difference. And i appreciate the fact that you put the two graphs up which really gives perspective.

I too had to look at my graph after all this graph gazing i've been doing lately and admit... actually, my graph aint that bad.

It's all about the zoom factor if you get me. The widest point we can possibly look at (100k hands in this case) ought to remind us where we're at. It's an awesome achievement in itself to HAVE 100k hands to look at. Otherwise one is really groping blindly through poker.

Good point about not blaming variance too much. Oh, as soon as i can i'll post some hands and PM you. I got sweated tonight for the first time, was awesome, realised how much i had to learn just by the general feel of how lacking i felt in comparison to the sweatER if you will. It's all fine tuning, but man the fine tuning is significant. Like you're selling a few thousand loaves of bread a week and you save a few pennies on each loaf's production cost. Then look at overall profit for the year, but there are dozens of tiny factors to improve upon so it really gains momentum.