Originally posted by lessthanthreee
The most important thing is to make sure you understand the difference between the statistical definition of 'variance' and the slang poker term 'variance'. They are essentially the same, but poker players have misused and misinterpreted the term 'variance' which causes a lot of confusion in the poker world to people who have no clue about stats.
This, definitely. I sometimes see people using 'variance' when they mean 'risk of ruin'.
Originally posted by lessthanthreee
i think the link between bigger roi and "variance" is not directly related, but indirectly related. variance is the square of the standard deviation. the standard deviation comprises of (is a measure of) a number of risk factors. in poker its things like number of players, quality of players, payout structures, your own skill. Often poker players dont understand that the term 'variance' is a statistical measure of risk. Whether one has a 'low variance' or 'high variance' style is something that a lot of people will misunderstand because there are a lot of factors that go into the variance of poker, most of which have never been studied or sampled over millions of hands.@tim, im talking about a 150BI downswing in the green line. The red line isn't really relevant to a statistical sample in poker (it would be a purely theoretical sample) and a risk of ruin study because the whole point of it is to include the 'swings'. The red line is by definition, a measure (well an attempt because its not 100% accurate) to remove a large factor of risk in poker (luck).
What you would need to do is run a risk of ruin on a red line sample, and a risk of ruin on a green line sample. you will discover the red line sample has a significantly reduced chance of downswings compared to the green line because the 'luck factor' is such a massive component of risk in poker which will comprise most of the 'variance'.
Hope that makes sense..... brings me back to 2nd year finance.
The most important thing is to make sure you understand the difference between the statistical definition of 'variance' and the slang poker term 'variance'. They are essentially the same, but poker players have misused and misinterpreted the term 'variance' which causes a lot of confusion in the poker world to people who have no clue about stats.
Thankyou thankyou thankyou. I just did a simple math quiz to help me understand some of these concepts ( http://www.mathsisfun.com/data/standard-deviation.html - hope np with this link, PS 
).
I remember hearing that the std deviation for 9 (or was it 10?) man sngs is 1.7 And that this std deviation is in buyins. So if ROI is 10% then you make 0.1 of a BI per Sng, with a std deviation of 1.7 BI. But what does that mean? Because, intuitively, you cant lose more than 1 BI per SnG?
Now, say we have the following ROI results for 5* 1000 * $10 game tranches of Sngs: 5%; -3%; -9%; 14%; 6%.
($500 -$300 -$900 +$1400 +$600 = $1300 up at the end) i.e. overall ROI of 2.6%.
My SD Calc ( http://www.mathsisfun.com/data/standard-deviation-calculator.html )
tells me that the SD is 7.9% and the Variance is 62.64
But, what does this mean in practice?
Originally posted by jbpatzer
Originally posted by lessthanthreee
The most important thing is to make sure you understand the difference between the statistical definition of 'variance' and the slang poker term 'variance'. They are essentially the same, but poker players have misused and misinterpreted the term 'variance' which causes a lot of confusion in the poker world to people who have no clue about stats.This, definitely. I sometimes see people using 'variance' when they mean 'risk of ruin'.
That's nothing. I hear people calling it variance when someone hits a two outer on them. 
Originally posted by dydukas
Hi Tim,Thanks for picking my quote! Glad to be the winner.
But I'm currently grinding micro limits on Ongame for BRB challenge so there's not much coaching or sweating needed there... would be just boring for you and not that necessary for me.If we could postpone that for a few weeks until I move at least to $11s that would be great. If not I kindly give the prize to Leito99 as he would benefit more from that atm.
Thanks, I appreciate it!
Good luck in the challenge!
lets assume the distribution of ROIs is normal
it will looks like this
the increments on the X axis are the standard deviation intervals. the % inside the graph is the % chance that an individual result is inside that portion (ie. all results lie under that graph and the middle point is the average).
using your example. you have a 68.2% chance of being between -1 and +1 SDs from the mean. 27.2% chance of being between -2-1 and +2-1 (the next portion on each side) etc. So the average roi is 2.6%, the SD is 7.9% so you have a 68.2% chance of being between (2.6%-7.9%) and (2.6%+7.9%) = between -5.3% and +10.5%. This shows you how volatile a 1,000 game sample is. there is still a pretty significant chance (27.2%) that your result will be in the next portion, etc etc..
more info here if u want to learn more ( http://en.wikipedia.org/wiki/Standard_deviation )
A lot of it is very technical and irrelevant, but the first few paragraphs should give you a pretty good understanding.
At the end of the day, no one has ever produced a statistically significant sample in any form of SnG poker to do a reliable test. even a 200k game sample of super turbos is going to show a very 'flat' graph , ie. the chances of being multiple SDs from the mean are relatively high. This is because of all the risk factors that poker involves.
[edit] in reality, the distribution of an entire poker population will not be normal but will be heavily skewed towards a very large proportion of big losing players and a very small proportion of big winning players with the mean less than zero (due to the rake).
Also, feel free to correct me (if any professors are reading). Its been a long time since I did this stuff at Uni.
also. I would HIGHLY recommend going to this website
scrolling down the front page (its in alphabetical order) you will find a whole section on statistics. This guy is amazingly good at explaining concepts simply and they are all free videos! I'm sure you will feel like a stats wiz after you finish the series.
Credit goes to GodlikeRoy for that link, found it on his blog and its the NUTS.
here is his video on SD - http://www.khanacademy.org/video/statistics--standard-deviation?playlist=Statistics
here is his video on variance - http://www.khanacademy.org/video/statistics--variance-of-a-population?playlist=Statistics
Originally posted by lessthanthreee
http://forumserver.twoplustwo.com/36/stt-strategy/stt-simulations-graphical-look-into-variance-73786/enjoy
kinda makes me want to stop playing SNGs
I've written a lot about this stuff for my book with TT.......but you'll just have to wait until it comes out and buy a copy. There are several good reasons why I'm affectionately called 'Professor Bastard' by the shits.....I mean the students.
Originally posted by lessthanthreee
also. I would HIGHLY recommend going to this websitescrolling down the front page (its in alphabetical order) you will find a whole section on statistics. This guy is amazingly good at explaining concepts simply and they are all free videos! I'm sure you will feel like a stats wiz after you finish the series.
Credit goes to GodlikeRoy for that link, found it on his blog and its the NUTS.
here is his video on SD - http://www.khanacademy.org/video/statistics--standard-deviation?playlist=Statistics
here is his video on variance - http://www.khanacademy.org/video/statistics--variance-of-a-population?playlist=Statistics
:heart: :heart:
OMG so useful
Main fear is that understanding variance better will make me instaquit poker ♦
At the end of the day, no one has ever produced a statistically significant sample in any form of SnG poker to do a reliable test. even a 200k game sample of super turbos is going to show a very 'flat' graph
the thing with big samples is inaccurate too because we have to take into consideration following facts:
1, we get better so we increase our ROI
2, the field gets better so it decreases our ROI
Gut instinct tells me that a bigger sample is more accurate notwithstanding the effects of our own and others' increasing skill.
Actually ran above EV today. Party's servers must be in some kind of meltdown. Obv can't continue. I shall greet tomorrow's return to normality (i.e running badly) almost with relief.
Deffo
I was running like shit and I run like god at Party ... Deffo smth wrong w their servers, we should contact support if it doesnt resolve itself ...
I am a new member to this community and accidentaly came up with this blog.Can you tell me please how many tables do you play,what stakes,how many hours per/days per week do you play in order to live your family?DO you think is possible to make a living from 9 man sngs?Do you think rakeback is a must for a sng grinder or bonus from ps/pp are good enough to hold variance?
Originally posted by ulairiiii
I am a new member to this community WELCOME! and accidentaly came up with this blog.Can you tell me please how many tables do you play ANYWHERE FROM 6 TO 10, what stakes ANYWHERE FROM $11 -- $33 ,how many hours per/days per week do you play in order to live your family? ROUGHLY 6 PER DAY/60 PER WEEK DO you think is possible to make a living from 9 man sngs? I AM NOT AN EXPERT IN 9 MAN SIT AND GO'S -- ASK <3. BUT HONESTLY, WITHOUT SUBSTANTIAL SAVINGS BEHIND YOU I WOULD NOT BE COMFORTABLE TRYING TO SUPPORTE A FAMILY THIS WAY. Do you think rakeback is a must for a sng grinder or bonus from ps/pp are good enough to hold variance? WELL BONUSES ARE A FORM OF RAKE BACK OF COURSE. HOWEVER WEEKLY RAKEBACK DOES HELP TO GET YOU THROUGH DIFFICULT PERIODS OF VARIANCE. THE MAIN THING IS TO HAVE SUFFICIENT BANKROLL IN THE FIRST PLACE SINCE WHETHER YOU RECEIVE YOUR RAKE BACK WEEKLY OR EVERY THREE MONTHS IS ONLY A MATTER OF CASH FLOW.
If you need more information send me a private message and I will try to help.
Sheer Brutality:
On the plus side I have managed to condense my thoughts and feelings about running so badly recently into one statement which I believe to be a true reflection of the reality of playing professionally:
If neither cash flow nor bankroll are issues (and they are not*), then all we need to do is to play a sufficient volume of games/hands each month so that, at our given level of profitability, average (EV) earnings are comfortably sufficient to live on.
I think Twice T (and I hope I am not misquoting him) said something similar -- and more succinctly: "I know my hourly, so I just sit the fu*k down and play."
With these thoughts in mind I decided to move up to the $55s on Party (or at least the few such games that run). In the back of my head the old voice said: don't move up until your runbad has ended, otherwise you will lose even more at the higher stakes. But the rational bit of me understands, now -- if not before, that this is simply nonsense. If my game is solid and profitable, I will make money. I've only managed 12 games so far but the ineptitude of some of the players with 55 bucks to splash around is pretty staggering.
__________
*well, not really 
I am wondering how you can survive this and still keep playing your A game.This is sick.Did you considered to start grinding in another platform?
Originally posted by ulairiiii
I am wondering how you can survive this and still keep playing your A game.This is sick.Did you considered to start grinding in another platform?
No. I have equity tied up at PP in the form of bonuses and I don't believe in rigged sites. The red line is what tells me I'm playing my A-game (or maybe B+)
Originally posted by Tim64
Originally posted by ulairiiii
I am wondering how you can survive this and still keep playing your A game.This is sick.Did you considered to start grinding in another platform?No. I have equity tied up at PP in the form of bonuses and I don't believe in rigged sites. The red line is what tells me I'm playing my A-game (or maybe B+)
No i dont believe that a site could be rigged.Ofc i agree with you.But maybe its an unlucky site for you.Maybe you should take a shot at stars
