Two Plus Two Newer Archives (http://archives1.twoplustwo.com/index.php)
-   Poker Theory (http://archives1.twoplustwo.com/forumdisplay.php?f=17)
-   -   Determining the probability that you're a winning player (http://archives1.twoplustwo.com/showthread.php?t=555120)

 jesse8888 11-26-2007 09:01 PM

Determining the probability that you\'re a winning player

I play live limit texas holdem and would like to determine the probability that I am a winning player. What information do I need to do this? I have heard that for solid, TAG play the standard deviation for an hour of play is around 15 big bets. Is that true? That seems massive to me....If I assume this to be true, how do I apply it to the fact that I have won X big bets in Y hours to generate a probability distribution for my true win rate?

In case I haven't been clear so far, I'm looking to come up with statements such as the following:

There is an X percent chance that I am a losing player.
There is a Y percent chance that my true win rate is above .5 big bets per hour.
There is a Z percent chance that my true win rate is above 1 big bet per hour.

Mods, please move this if you feel I've posted in the wrong forum.

 JulioYalil 11-26-2007 09:37 PM

Re: Determining the probability that you\'re a winning player

there is no such thing as the probability of u being a winnin player. u either r or r not. and it's not that difficult to determine.

cash outs - money invested in poker = money made playin poker

if it's a positive real number then u r a winnin player.

i must add: i just reread ur post and now i'm even more confused. wtf u talkin about?

 Jimbo 11-26-2007 09:52 PM

Re: Determining the probability that you\'re a winning player

[ QUOTE ]
there is no such thing as the probability of u being a winnin player. u either r or r not.

[/ QUOTE ]

well, this tells us that you don't really know if you are a winning player or not then.

To the OP: You will get a more concise reply in the probability forum. You wil need your SD and a reasonably large sample of results to determine the likelihood of your being a "long term" winning player and also to compute your ROR using the size of you bankroll, the limits you play and your style of play and type of game you play.

Jimbo

Re: Determining the probability that you\'re a winning player

play 10k hands. if you win, post the stats in a forum and ask people to estimate your chances of being a winning player.

if you lose, make some trivial changes to how you approach the game, and....

play 10k more hands. if you win, post the stats in a forum and ask people to estimate your chances of being a winning player.

if you lose, make some trivial changes to how you approach the game, and....

play 10k more hands. if you win, post the stats in a forum and ask people to estimate your chances of being a winning player.

if you lose, make some trivial changes to how you approach the game, and....

play 10k more hands. if you win, post the stats in a forum and ask people to estimate your chances of being a winning player.

if you lose, make some trivial changes to how you approach the game, and....

play 10k more hands. if you win, post the stats in a forum and ask people to estimate your chances of being a winning player.

if you lose, make some trivial changes to how you approach the game, and....

 pzhon 11-27-2007 12:20 AM

Re: Determining the probability that you\'re a winning player

[ QUOTE ]
I play live limit texas holdem and would like to determine the probability that I am a winning player. What information do I need to do this?

[/ QUOTE ]
You can't determine this without a prior distribution. If your prior distribution doesn't make sense, your results may not make sense. You can estimate how likely your results are for each win rate, and update this to get your new distribution for your win rate.

What you might be looking for instead is a confidence interval. A rough 95% confidence interval is your observed rate +- 2 standard deviations. The standard deviation of your observed rate after n independent periods is sqrt(n) times the standard deviation for a single period.

[ QUOTE ]
I have heard that for solid, TAG play the standard deviation for an hour of play is around 15 big bets. Is that true?

[/ QUOTE ]
No, not in most games. It's usually 8-10 big bets. The figure of 15 BB is the standard deviation per 100 hands in an online full ring game (or a very tight full ring live game). It's about 10% higher shorthanded.

 GeeBeeQED 11-27-2007 08:25 AM

Re: Determining the probability that you\'re a winning player

Purchase a PDA or a phone with a spreadsheet program. Track game type, location, \$ in, \$ out, \$ net profit/loss, hours played.

If your not keeping track of what your doing then I'd say there is about a 99% chance your not a winning player.

What is the point in running some long drawn out calculation to deturmine how tall you are or what color your eyes are when you can measure these things directly. Your either winning more than your losing or not. If your late to work too often, did your boss go through some long drawn out calculation to arrive at this conclusion?

If you muck about a keyboard long enough, you can figure out all kinds of ways to feel good about a result that you havn't measured.

It might be an interesting question mathmatically to try to extrapolate how an adjustment in play might effect long term ROI......in the future. However, trying to figure this out after the fact shows your also not playing out of a bankroll. Oh, it's so easy to chalk it up to a bad night, get a dozen Benjamins out of the hiding place and go after those boys again. Some ups, some downs and a month or two later after may winning sessions we are back in our hidding place again retrieving more Benjamins. Sound firmiliar to anybody?

Hey, I know the feeling, your surrounded by a bunch of players playing terrible and you want to feel like a winner. I want to say "I'm a winning player". Yet it's a most complicated game and the patience to work out a profit is hard to come by. Make a real measurment. When a years gone by and you've got 12 months to look at, you can begin to decide if your a long term winner or not. It's an illusive goal and probably only 5% get there. Most of us, in fact probably most of the players lurking on this forum are in the other 95%. There is no shame in being there. Only shame in lingering there.

Dave

 jesse8888 11-27-2007 01:41 PM

Re: Determining the probability that you\'re a winning player

Wow. Clearly like two of you had a freaking clue what I was talking about. Upon reading my post, I can see why. Let's try an example:

Suppose you have three coins, two of which are loaded. Specifically, coin A lands heads up 25% of the time, coin B is fair, and coin C lands heads up 75% of the time. Now suppose you don't know which coin is which, and you pick one up at random.

Now suppose you flipped the coin 8 times, and it landed heads up on only 2 trials. What is the probability that you have coin A on your hands? Pretty high, right? Yeah, pretty high, but not 100%. If you think it's 100%, well....please don't click the reply button. If you can teach me how to find the probability, please do.

So now let's suppose I'm Joe Bloggs poker player, and I have a win rate. It's either 0 big bets per hour, .5 big bets per hour, or 1 big bet per hour; only problem is, I don't know which one it is. So I go out and I play 170 hours and end up winning 142 big bets. If the world were so simple and discreet, it would be the exact same problem as the coins. The world's not discrete, so this version of the problem will involve an integral or two, but the concept is the same.

Have I been a little more clear?

 jesse8888 11-27-2007 01:46 PM

Re: Determining the probability that you\'re a winning player

Oh, and thanks for the 8-10 BB standard dev estimate. That sounds a lot more likely to me.

 RustyBrooks 11-27-2007 02:03 PM

Re: Determining the probability that you\'re a winning player

Right, OP is looking for a way to find out how close his CURRENT win rate is to his TRUE win rate.

I don't remember how to do this. I do remember there being a program that helps you with this calculation though. It's also designed to help you track live sessions (and it can use the data in the live sessions to do the calculation). I uh can't think of that program's name at the moment but maybe someone else remembers.

 pzhon 11-27-2007 02:04 PM

Re: Determining the probability that you\'re a winning player

[ QUOTE ]

What is the point in running some long drawn out calculation ... Your either winning more than your losing or not.

[/ QUOTE ]
[img]/images/graemlins/spade.gif[/img] It is very reasonable to collect evidence of being a winning player, and then ask whether that evidence is strong, or weak.

[img]/images/graemlins/spade.gif[/img] "Results +- 2 standard deviations" is not long and drawn out.

[img]/images/graemlins/spade.gif[/img] The contraction of "you are" is "you're," while "your" is a possessive pronoun.

[ QUOTE ]
When a years gone by and you've got 12 months to look at, you can begin to decide if your a long term winner or not.

[/ QUOTE ]
This is terrible. It generally takes much less than 12 months to determine whether you are a winning player. It could be a huge waste of time and money to spend 12 months at each level before worrying about whether you are a winning player or not. Further, do you really think "years" is a contraction of "year has?" Why did you write that sloppy, rambling, incorrect post in response to my simple, correct answer, which also gives you a range of win rates supported by the evidence?

All times are GMT -4. The time now is 07:19 PM.