Check out our brand new Local Poker Communities! Get updates and interact with poker players in your area.
Visit the United States Poker Community | Visit the California Poker Community | Read more about the Launch of P5s Local

How do you know you're a winner?

Reply to Thread
  1. BlueRenegade BlueRenegade is offline

    Vancouver, WA
    United States

    Posts: 58
    Joined: Jul 09

    Jan 1st, 2010 (1/1/2010 8:30am)

    I've got a statistics question...please somebody who still remembers how to do bell curves & standard deviations, etc., either answer this for me, remind me of the formula to calculate it, or direct me to a thread where this is answered (tried searching/Googling, no luck).

    Anyway, my question is this: How many trials does it take for your ROI to be statistically certain and not due to randomness? Obviously this answer will vary depending on the game and the number of players, so I'll give 3 different scenarios:

    1.) One-table SNG's

    2.) 180-man SNG's

    3.) 2k-3k player MTT's

    It's clear to me that one-table SNG's will require the least number of trials, that 180-man SNG's will require more trials, and that MTT's will require the most trials (in other words, variance increases as you add more players)...but if you have a 15% ROI after 100 one-table SNG's, is your ROI statistically reliable/valid (i.e. you're more than a couple standard deviations away from the mean)? How about MTT's? Do you need to play 5,000 of them before you know what your statistically true ROI is?

    Note I'm not asking for bankroll advice, moving up/down levels, etc. I'd just like to know at what point, for each of the above 3 types of games, I could consider myself a winning player and not just a lucky recipient of positive variance.

    Thanks in advance for any answers or pointing me in the right direction,

    Blue
  2. virtualthief virtualthief is offline

    NJ
    United States

    Posts: 325
    Joined: Nov 06

    Jan 1st, 2010 (1/1/2010 10:16am) in reply to BlueRenegade

    1) 1000
    2) 2500
    3) 5000
  3. aquavitix aquavitix is offline

    Solomon Islands

    Posts: 308
    Joined: Aug 07

    Jan 1st, 2010 (1/1/2010 2:23pm) in reply to BlueRenegade

    If you don`t know if you`re a winner you probably aren`t.
  4. rayfox111 rayfox111 is offline

    United Kingdom

    Posts: 833
    Joined: Mar 09

    Jan 1st, 2010 (1/1/2010 2:24pm) in reply to BlueRenegade

    Not an authority on the calculation of ROI, but you can never be statistacally certain about future results.
    I think the best you can do is make an educated hopeful guess.
    It's very difficult to play enough MTT's especially, to obtain meaningful figures.
    So sorry, can't personly help.
  5. XNASTYNATEX XNASTYNATEX is offline

    Bend, OR
    United States

    Posts: 940
    Joined: Jan 07

    Jan 1st, 2010 (1/1/2010 2:27pm) in reply to rayfox111

    i'd guess like 500x entrants
  6. BlueRenegade BlueRenegade is offline

    Vancouver, WA
    United States

    Posts: 58
    Joined: Jul 09

    Jan 1st, 2010 (1/1/2010 3:58pm) in reply to virtualthief

     
    Originally Posted by virtualthief View Post


    1) 1000
    2) 2500
    3) 5000

    Thief, those numbers "feel" about right to me...just curious how you came by them?

    Thanks,
    Blue
    Thread Starter
  7. BlueRenegade BlueRenegade is offline

    Vancouver, WA
    United States

    Posts: 58
    Joined: Jul 09

    Jan 1st, 2010 (1/1/2010 3:59pm) in reply to aquavitix

     
    Originally Posted by aquavitix View Post

    If you don`t know if you`re a winner you probably aren`t.

    Try reading the entire post instead of just the title next time, maybe you'll have something useful to contribute instead of being reduced to a lame one-liner.
    Thread Starter
  8. apestyles apestyles is offline Moderator

    Vancouver
    Canada

    Posts: 5,087
    Joined: Sep 05

    Jan 1st, 2010 (1/1/2010 4:22pm) in reply to virtualthief

     
    Originally Posted by virtualthief View Post


    1) 1000
    2) 2500
    3) 5000

    Unfortunately you are on the low end - especially for high stakes... I would say in general though, if you're losing over those given sample sizes you are probably a losing player and vice-versa. But it is definitely possible to be a winning player and lose over those sample sizes and a losing player and winning over those sample sizes.
     282 12 

    apestyles is an instructor at PocketFives Training . To get more of his advice and to watch his training videos, click here.

  9. lockdownpokah lockdownpokah is offline

    United States

    Posts: 2,289
    Joined: Sep 07

    Jan 1st, 2010 (1/1/2010 4:27pm) in reply to apestyles

     
    Originally Posted by apestyles View Post

    Unfortunately you are on the low end - especially for high stakes... I would say in general though, if you're losing over those given sample sizes you are probably a losing player and vice-versa. But it is definitely possible to be a winning player and lose over those sample sizes and a losing player and winning over those sample sizes.

    this
     
  10. m_hawk_1 m_hawk_1 is offline

    Australia

    Posts: 752
    Joined: Mar 07

    Jan 1st, 2010 (1/1/2010 6:07pm) in reply to BlueRenegade

    I ran some simulations re: variance in 1000 player MTTs a while back. It looked to me like a lot more than 5000 MTTs before any close to accurate ROI estimations could be made. But like apestyles mentioned, you should definitely know if you're a winning player by this point. Obviously if you're talking about 2-3k fields then you're looking at even more variance...

    Original thread here - http://www.pocketfives.com/poker-for...62?pageindex=1
    471 
  11. wes_mc wes_mc is offline

    Raleigh, NC
    United States

    Posts: 2,028
    Joined: Dec 09

    Jan 1st, 2010 (1/1/2010 8:25pm) in reply to m_hawk_1

     
    Originally Posted by m_hawk_1 View Post

    I ran some simulations re: variance in 1000 player MTTs a while back. It looked to me like a lot more than 5000 MTTs before any close to accurate ROI estimations could be made. But like apestyles mentioned, you should definitely know if you're a winning player by this point. Obviously if you're talking about 2-3k fields then you're looking at even more variance...

    Original thread here - http://www.pocketfives.com/poker-for...62?pageindex=1

    very enlightening post. thanks for the link.
  12. SigmaBetaRho SigmaBetaRho is offline

    Glen Oaks, NY
    United States

    Posts: 989
    Joined: Nov 06

    Jan 2nd, 2010 (1/2/2010 1:31am) in reply to wes_mc

    For 1 table SNGs after 3,000 games you can be 95% sure of your true ROI. For more info google jhub3000's blog.
  13. BlueRenegade BlueRenegade is offline

    Vancouver, WA
    United States

    Posts: 58
    Joined: Jul 09

    Jan 2nd, 2010 (1/2/2010 3:21am) in reply to SigmaBetaRho

    Cool, guys. Thanks much for some very useful info!

    OMG Apestyles wrote something in my thread...my brush with fame, lol.
    Thread Starter
  14. MDG0RD0N MDG0RD0N is offline

    Posts: 401
    Joined: Dec 08

    Jan 2nd, 2010 (1/2/2010 3:33am) in reply to BlueRenegade

    Nice link above. As for the OP, a better question is 'what sample size is needed to know my true ROI within a few points', as opposed to, 'am I a winning player'. Remember that the larger edge you have, the smaller the necessary sample size to determine if you are > 0% ROI. It won't take as many games for a 20% player compared to a 5% player to know if they are +EV, but it will take the same sample to determine their true ROI's to within x%.

    At the end of the other thread, Stevie posted a [now dead] link to RVG's ROI Simulator. I'm not sure where to find a working link for this, but I have a copy of it if anyone wants to check it out. I've used it to do a little bit of analysis on SNG variance.
  15. Autolobotomist Autolobotomist is offline playing air is my style? so what

    Tempe, AZ
    United States

    Posts: 10,066
    Joined: Jun 07

    Jan 2nd, 2010 (1/2/2010 5:13am) in reply to BlueRenegade

    Just uploaded it. RVG ROI simulator Scan it first, I might not be trustworthy :)
  16. DSKekaha DSKekaha is offline

    Puyallup, WA
    United States

    Posts: 583
    Joined: May 07

    Jan 2nd, 2010 (1/2/2010 11:21am) in reply to BlueRenegade

    If you don't know if you're a winner, you prob -- oh, wait

    nm
  17. aquavitix aquavitix is offline

    Solomon Islands

    Posts: 308
    Joined: Aug 07

    Jan 2nd, 2010 (1/2/2010 11:22am) in reply to BlueRenegade

    no sense of humour then. unlucky.

    ahhahhahhahahahah
  18. DEEZZZ_NUTS DEEZZZ_NUTS is offline

    Minneapolis, MN
    United States

    Posts: 723
    Joined: Mar 06

    Jan 2nd, 2010 (1/2/2010 3:20pm) in reply to BlueRenegade

    when you don't have to ask that question
     
  19. BlueRenegade BlueRenegade is offline

    Vancouver, WA
    United States

    Posts: 58
    Joined: Jul 09

    Jan 2nd, 2010 (1/2/2010 10:42pm) in reply to MDG0RD0N

     
    Originally Posted by MDG0RD0N View Post

    Nice link above. As for the OP, a better question is 'what sample size is needed to know my true ROI within a few points', as opposed to, 'am I a winning player'. Remember that the larger edge you have, the smaller the necessary sample size to determine if you are > 0% ROI. It won't take as many games for a 20% player compared to a 5% player to know if they are +EV, but it will take the same sample to determine their true ROI's to within x%.

    At the end of the other thread, Stevie posted a [now dead] link to RVG's ROI Simulator. I'm not sure where to find a working link for this, but I have a copy of it if anyone wants to check it out. I've used it to do a little bit of analysis on SNG variance.

    You're absolutely right, that's what I was really looking for the answer to. I will check out the ROI Simulator also, thanks again guys.

    Blue
    Thread Starter
  20. shootair shootair is offline

    Honolulu, HI
    United States

    Posts: 432
    Joined: Mar 08

    Jan 2nd, 2010 (1/2/2010 11:56pm) in reply to BlueRenegade

    if u play teh pokers, u put chip in pot. if u get back, u make teh win!
Reply to Thread

Similar Threads