Two Plus Two Publishing LLC Two Plus Two Publishing LLC
 

Go Back   Two Plus Two Poker Forums > >

Live No-Limit Hold’em Cash Discussion of no-limit hold’em live cash games of all stakes.

Reply
 
Thread Tools Display Modes
Old 12-27-2016, 02:49 PM   #17326
Tbirdx24
old hand
 
Tbirdx24's Avatar
 
Join Date: May 2009
Location: NY
Posts: 1,630
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by Garick View Post
Or you could just look at the post above yours. ��
Lol we must have posted at the same time. Thank you very much.
Tbirdx24 is offline   Reply With Quote
Old 12-27-2016, 03:21 PM   #17327
homerdash
banned
 
homerdash's Avatar
 
Join Date: Sep 2006
Location: desert
Posts: 6,062
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by Garick View Post
Ok, I found it and it's probably worth posting again. To find the 95% confidence level on your bounds related to observed winrate:
Multiply your hourly standard deviation by 2 and then divide the result by the square root of your number of hours. For short: Hourly SDev*2 / sqrt of hours

As you can see, increased hours will tighten the heck out of these bounds, especially over the first couple hundred hours.

To use, combine with observed winrate. So if observed winrate is 7bbs/hr and your bounds are +/- 6.3bbs/hr, you can be 95% confident that you are a winning player, even though your true winrate could be as low as .7bbs/hr.


conversely, can my true winrate be as high as 13-14bb/hr?
homerdash is offline   Reply With Quote
Old 12-27-2016, 03:25 PM   #17328
johnny_on_the_spot
Carpal \'Tunnel
 
johnny_on_the_spot's Avatar
 
Join Date: Jun 2013
Location: S-Mart
Posts: 10,922
*** Official Winrates, bankrolls, and finances ***

Quote:
Originally Posted by Garick View Post
Ok, I found it and it's probably worth posting again. To find the 95% confidence level on your bounds related to observed winrate:
Multiply your hourly standard deviation by 2 and then divide the result by the square root of your number of hours. For short: Hourly SDev*2 / sqrt of hours

As you can see, increased hours will tighten the heck out of these bounds, especially over the first couple hundred hours.

To use, combine with observed winrate. So if observed winrate is 7bbs/hr and your bounds are +/- 6.3bbs/hr, you can be 95% confident that you are a winning player, even though your true winrate could be as low as .7bbs/hr.


Is there a reason why we're using 2 and not 1.96 as our z value? I get that they're practically the same and that 2 is "easier" than 1.96, but we're also calcing the square root of a random number, so it's not like anyone is doing this without a calculator...
johnny_on_the_spot is offline   Reply With Quote
Old 12-27-2016, 04:35 PM   #17329
Garick
Oberbiergenießer
 
Garick's Avatar
 
Join Date: Dec 2007
Location: Do you even math, bruh?
Posts: 24,603
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by homerdash View Post
conversely, can my true winrate be as high as 13-14bb/hr?
Exactly.
Quote:
Originally Posted by johnny_on_the_spot View Post
Is there a reason why we're using 2 and not 1.96 as our z value? I get that they're practically the same and that 2 is "easier" than 1.96, but we're also calcing the square root of a random number, so it's not like anyone is doing this without a calculator...
IDK. I just stole the formula from bip! Bip!?
Garick is offline   Reply With Quote
Old 12-27-2016, 04:45 PM   #17330
bip!
Slow Pony
 
bip!'s Avatar
 
Join Date: Oct 2012
Location: not on urban dictionary...
Posts: 13,669
re: Winrates, bankrolls, and finances

2 was just an approximation. 1.96 if you want to be exact.
bip! is offline   Reply With Quote
Old 12-27-2016, 04:56 PM   #17331
MIB211
veteran
 
Join Date: Jan 2013
Posts: 3,011
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by homerdash View Post
conversely, can my true winrate be as high as 13-14bb/hr?
I don't know the statistical answer here, but my strong suspicion is that if you have someone who's winning 7 bbs/hour with a 95% confidence interval of +/- 7bb/hour, it's significantly more likely that they're a break even player than they're winning 14bb/hour. This is an application of Bayes theorem and reversion to the mean. Basically, you have a ton of observed break even players, and very, very few 14bb/hour crushers. So, much more likely that any given person who's a 7bb/hour winner is running good than bad.
MIB211 is offline   Reply With Quote
Old 12-27-2016, 05:31 PM   #17332
OvertlySexual
old hand
 
OvertlySexual's Avatar
 
Join Date: Jan 2014
Posts: 1,518
re: Winrates, bankrolls, and finances

FWIW, I want to note once again that some of the poker tracking apps give wrong standard deviation numbers.

For example, Poker Income gives me standard deviation in the low fourties whereas RunGood gives me a standard deviation between 90 and 100 BB/hr.

This in turn changes drastically the results of the confidence interval from 2.9BB to 7BB.

I did try to find my standard deviation on excel and I think Poker Income is the one doing the right calculation. I could be wrong though.
OvertlySexual is offline   Reply With Quote
Old 12-27-2016, 05:48 PM   #17333
Garick
Oberbiergenießer
 
Garick's Avatar
 
Join Date: Dec 2007
Location: Do you even math, bruh?
Posts: 24,603
re: Winrates, bankrolls, and finances

90-100 seems much more standard. Very few of the folks ITT who have converted sdev/session into sdev/hr via crunching the numbers (vs just taking what the app says as gospel) have reported sdev <60bbs/hr. Like three of us, IIRC, and all three were upper 50s.

100bbs/hr, OTOH, seems pretty standard.

Last edited by Garick; 12-29-2016 at 09:13 AM. Reason: typo
Garick is offline   Reply With Quote
Old 12-27-2016, 05:59 PM   #17334
homerdash
banned
 
homerdash's Avatar
 
Join Date: Sep 2006
Location: desert
Posts: 6,062
re: Winrates, bankrolls, and finances

my Poker Income has it at 112
homerdash is offline   Reply With Quote
Old 12-27-2016, 06:46 PM   #17335
OvertlySexual
old hand
 
OvertlySexual's Avatar
 
Join Date: Jan 2014
Posts: 1,518
re: Winrates, bankrolls, and finances

Let me tell you what I did on excel.

1. I took the session profit from 48 sessions.

2. I then took the duration of each session and converted it into a decimal form. So for example a session of 11 hours 56 minutes was converted to 11.933.

3. I then divided each session's profit by the duration so I found the hourly profit of each session.

4. I then calculated the sum of the individual profit and came up with a mean profit for all sessions.

5. I then subtracted the mean from each hourly session profit.

6. I squared the resulting number.

7. Then I divided the resulting number by the number of sessions.

8. I then got the square root of that number.

That final number is supposed to be the standard hourly deviation. Using that method, I found a standard deviation of 43 BB, whereas poker income gives me 42.85 for the same sample. Runggood gives me a standard deviation of 95.21.

What did I do wrong?

Regardless. Can we determine which poker tracking app does it right? Can someone do the proper calculation on excel and compare with the number he gets on his poker app?
OvertlySexual is offline   Reply With Quote
Old 12-27-2016, 06:54 PM   #17336
feel wrath
The Situation
 
feel wrath's Avatar
 
Join Date: Nov 2010
Location: lost on the turn
Posts: 23,866
re: Winrates, bankrolls, and finances

Difference between bb and BB?
feel wrath is offline   Reply With Quote
Old 12-27-2016, 07:13 PM   #17337
Sneaky Pete
banned
 
Join Date: Aug 2015
Posts: 353
re: Winrates, bankrolls, and finances

Big Blind as position vs big blinds as quantifier.
Sneaky Pete is offline   Reply With Quote
Old 12-27-2016, 08:43 PM   #17338
bip!
Slow Pony
 
bip!'s Avatar
 
Join Date: Oct 2012
Location: not on urban dictionary...
Posts: 13,669
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by feel wrath View Post
Difference between bb and BB?


The main difference is that "BB" tilts me every time I see it in a NL forum.

"bb" on the other hand has no such tilting effect.

.. probably because it actually means "big blind" and exists in NL.
bip! is offline   Reply With Quote
Old 12-27-2016, 08:54 PM   #17339
bip!
Slow Pony
 
bip!'s Avatar
 
Join Date: Oct 2012
Location: not on urban dictionary...
Posts: 13,669
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by MIB211 View Post
I don't know the statistical answer here, but my strong suspicion is that if you have someone who's winning 7 bbs/hour with a 95% confidence interval of +/- 7bb/hour, it's significantly more likely that they're a break even player than they're winning 14bb/hour. This is an application of Bayes theorem and reversion to the mean. Basically, you have a ton of observed break even players, and very, very few 14bb/hour crushers. So, much more likely that any given person who's a 7bb/hour winner is running good than bad.


^ to address this we took a survey of 2+2ers and they all assured us they are running below EV... often accompanied by a story about "that time I got 2 outed in a big pot".

bip! is offline   Reply With Quote
Old 12-27-2016, 08:57 PM   #17340
bip!
Slow Pony
 
bip!'s Avatar
 
Join Date: Oct 2012
Location: not on urban dictionary...
Posts: 13,669
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by OvertlySexual View Post
Let me tell you what I did on excel.

1. I took the session profit from 48 sessions.

2. I then took the duration of each session and converted it into a decimal form. So for example a session of 11 hours 56 minutes was converted to 11.933.

3. I then divided each session's profit by the duration so I found the hourly profit of each session.

4. I then calculated the sum of the individual profit and came up with a mean profit for all sessions.

5. I then subtracted the mean from each hourly session profit.

6. I squared the resulting number.

7. Then I divided the resulting number by the number of sessions.

8. I then got the square root of that number.

That final number is supposed to be the standard hourly deviation. Using that method, I found a standard deviation of 43 BB, whereas poker income gives me 42.85 for the same sample. Runggood gives me a standard deviation of 95.21.

What did I do wrong?

Regardless. Can we determine which poker tracking app does it right? Can someone do the proper calculation on excel and compare with the number he gets on his poker app?


You can't convert the individual results to per hour and treat that as normalized.

Later I will post how to do this in excel. It is simpler, but maybe not intuitive.
bip! is offline   Reply With Quote
Old 12-27-2016, 09:04 PM   #17341
YGOchamp
old hand
 
Join Date: Mar 2013
Posts: 1,517
re: Winrates, bankrolls, and finances

Theres no such thing as losing (or break even) poker players. Simply those who are running good and those who are running bad

Also, so can anybody else confirm that the poker income app has correct stddev?
YGOchamp is offline   Reply With Quote
Old 12-27-2016, 11:35 PM   #17342
MikeStarr
Carpal \'Tunnel
 
Join Date: Jan 2016
Posts: 7,978
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by OvertlySexual View Post
FWIW, I want to note once again that some of the poker tracking apps give wrong standard deviation numbers.

For example, Poker Income gives me standard deviation in the low fourties whereas RunGood gives me a standard deviation between 90 and 100 BB/hr.

This in turn changes drastically the results of the confidence interval from 2.9BB to 7BB.

I did try to find my standard deviation on excel and I think Poker Income is the one doing the right calculation. I could be wrong though.
About 2 months ago I sent all my numbers to Bip! and he confirmed that my StnDev number was correct. I use Run Good.
MikeStarr is offline   Reply With Quote
Old 12-27-2016, 11:50 PM   #17343
meale
Carpal \'Tunnel
 
meale's Avatar
 
Join Date: May 2013
Location: Pattaya, Thailand
Posts: 9,840
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by YGOchamp View Post
Theres no such thing as losing (or break even) poker players. Simply those who are running good and those who are running bad
:')
meale is offline   Reply With Quote
Old 12-28-2016, 05:27 AM   #17344
WereBeer
Carpal \'Tunnel
 
Join Date: Oct 2013
Posts: 9,650
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by bip! View Post
^ to address this we took a survey of 2+2ers and they all assured us they are running below EV
I think runbad is a broader problem than 2+2.

Casino regs all run super unlucky as well, like every one of them I ask or in fact don't ask, is keen to relay their sob story about how bad they run.
WereBeer is offline   Reply With Quote
Old 12-28-2016, 05:40 AM   #17345
feel wrath
The Situation
 
feel wrath's Avatar
 
Join Date: Nov 2010
Location: lost on the turn
Posts: 23,866
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by bip! View Post
^ to address this we took a survey of 2+2ers and they all assured us they are running below EV... often accompanied by a story about "that time I got 2 outed in a big pot".

unless of course, they posted higher win rates than anybody else, in which case 2+2ers assured them that they were either lying or were running unsustainably well and that a fall was right around the corner
feel wrath is offline   Reply With Quote
Old 12-28-2016, 05:41 AM   #17346
VolumeKing
adept
 
Join Date: Jan 2016
Posts: 729
re: Winrates, bankrolls, and finances

161 hrs since the 9th
VolumeKing is offline   Reply With Quote
Old 12-28-2016, 01:52 PM   #17347
iraisetoomuch
banned
 
iraisetoomuch's Avatar
 
Join Date: Aug 2013
Location: New Jersey
Posts: 34,453
re: Winrates, bankrolls, and finances

Quote:
Originally Posted by cAmmAndo View Post
That's a $5.40/hr bump yes compared to not playing those pots or running neutral but I think it's good to realize it's actually a $10.80 swing in short term wr if you lose those pots.
Not really.
I didn't profit those amounts ($2,700) I won pots of that amount.
Which means that I only profited $1,350 (really closer to $1,700 since some of them were not heads up pots, but whatever). So if I don't play them, I'm down $1,700 and if I lose them I'm down $2,700.

Quote:
Originally Posted by MikeStarr View Post
I wouldnt feel bad if you're not there. No matter what you see on 2+2 with people bragging about win rates, I would estimate that only 1-2% 2/5 players in the world are making that much. Being in the top 1-2% of anything is pretty impressive.
This is likely true as a general idea (but obv lol at people making up statistics on the spot with little to go on really).
But since the average poster on 2+2 is likely better than the average non-poster it is hard to really judge where we may lie in the poker spectrum and how many of us should be making XXbb/hour.

Quote:
Originally Posted by MIB211 View Post
I don't know the statistical answer here, but my strong suspicion is that if you have someone who's winning 7 bbs/hour with a 95% confidence interval of +/- 7bb/hour, it's significantly more likely that they're a break even player than they're winning 14bb/hour. This is an application of Bayes theorem and reversion to the mean. Basically, you have a ton of observed break even players, and very, very few 14bb/hour crushers. So, much more likely that any given person who's a 7bb/hour winner is running good than bad.
This is only true if we expect that the population is not normally distributed.
Otherwise it is by definition that the two results (+14bb/hour and +0bb/hour) are equally likely.
iraisetoomuch is offline   Reply With Quote
Old 12-28-2016, 04:15 PM   #17348
cAmmAndo
The Situation
 
cAmmAndo's Avatar
 
Join Date: Nov 2008
Location: Philly
Posts: 4,712
re: Winrates, bankrolls, and finances

^^^
Re: pot vs profit. Agreed.
cAmmAndo is offline   Reply With Quote
Old 12-28-2016, 04:24 PM   #17349
Maskk
veteran
 
Join Date: Feb 2009
Location: DC Area
Posts: 2,083
re: Winrates, bankrolls, and finances

I am curious how run bad effects play styles. I ran terrible for premium hands around the time I was really moving into a profitable expected WR (probably 500ish key hours)--this forced me to work on storytelling, bluffing, long-shot+small-risk hands, and other ways of gaining EV which is now an integral part of how I play.

I am curious if people think that how you ran while learning has affected your style of play at all? I wonder if it is something that solid tags and rocks even consider?


Sent from my iPhone using Tapatalk
Maskk is offline   Reply With Quote
Old 12-28-2016, 04:42 PM   #17350
Sneaky Pete
banned
 
Join Date: Aug 2015
Posts: 353
re: Winrates, bankrolls, and finances

All the time, but that's more psychology than WR related.
Sneaky Pete is offline   Reply With Quote

Reply
      

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


Forum Jump


All times are GMT -4. The time now is 11:30 AM.


Powered by vBulletin®
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.
Copyright © 2008-2020, Two Plus Two Interactive