Decent Sample Size?

    • mccol001
      mccol001
      Bronze
      Joined: 31.07.2010 Posts: 170
      Hey, I recently started playing the turbo 45/180 SNGs and seem to have been running incredibly hot but today I had a quick session this morning and it was a complete disaster... Just wondering what a good sample size would be to know if I was on a upswing, or a downswing now or what is happening..

      I'm playing up to 30 tables at a time, here are most of my results so far:

      45 Mans:

      $0.55 - 70 played - ROI% = 14.3 - ITM% = 15.7 = Up $5.50 in total
      $1.10 - 206 played - ROI% = 12.1 - ITM% = 20.9 = Up $27.40 in total
      $3.25 - 588 played - ROI% = 44.1 - ITM% = 18.5 = Up $843.00 in total
      $6.50 - 259 played - ROI% = 19.0 - ITM% = 18.5 = Up $320.50 in total
      And played 3 $12 and didn't cash in any :(

      180 Mans:

      $2.20 - 262 played - ROI% = 15.5 - ITM% = 12.6 = Up $89.60 in total
      $7.70 - 28 played - ROI% = -86.0 - ITM% = 7.1 = Down $185.36 in total
      And played 2 $12 and didn't cash :/

      But then this morning I had a quick session, played 6 $2.20's, 26 $3.25's, 15 $6.50's, 1 $7.70 and 5 $12.00 and got a ROI of -92%, ITM 3.8% and down $241.78 for the day

      Are things looking okay for me? Is today just a standard downsing or should I step back and play some more $3.25's before playing more $6.50's and starting $12's?

      I gotta go to work in a bit and will have a long session after, and will post results..
  • 6 replies
    • Heffron89
      Heffron89
      Bronze
      Joined: 23.01.2009 Posts: 813
      Hi Im playing 45 and 90 mans myself now and just today i lost 40 buy ins... I keep getting sucked out hehe but I looked em all up and most of the hadns were played fine so its just variance

      And in these mtt sngs u can expect huge swings.
    • TilTmuch
      TilTmuch
      Global
      Joined: 06.10.2010 Posts: 1,185
      play more 3.25's till u get ATLEAST 1k sample, obv more would be better. and if u crush it like u do now, then move up to 6's and so on
    • mccol001
      mccol001
      Bronze
      Joined: 31.07.2010 Posts: 170
      Hmm after this post I got a new Macbook and I haven't got Holdem Manager or any tracking software yet but I think I've ended 2 or 3 sessions with a profit out of around 10 sessions... You can see my results and the downswing on SharkScope... But I literally now have no idea if I'm a losing SNG player, running really bad or the cashout curse or what's happening lol...
    • pzhon
      pzhon
      Bronze
      Joined: 17.06.2010 Posts: 1,151
      I had overlooked this thread earlier.

      I think what you want is a confidence interval for your ROI. A rough 95% confidence interval after n tournaments is your observed ROI +- 2 SD/sqrt(n), where SD is the standard deviation for one tournament. For 9-player SNGs, the standard deviation is about 1.5 buy-ins. For 45 player tournaments, the standard deviation is about 2.8 buy-ins. For 180 player tournaments, the standard deviation depends more on your playing style, but it may be 5-8 buy-ins, more for a big winner.

      So, if you play 100 45-player tournaments and have a 20% ROI, you can say your 95% confidence interval is 20% +- 2(280%)/sqrt(100), or 20% +- 56%. This means you have strong evidence that your ROI is between -36% and +76%.

      It takes a very long time to figure out how often you win MTTs. You have to play long enough that you win a few times. Much of your ROI comes from the times you win the tournament, so until you play enough to win the tournament a few times, you can't expect an accurate estimate for your ROI.
    • amplifyd
      amplifyd
      Bronze
      Joined: 03.08.2009 Posts: 1,769
      Originally posted by pzhon
      I had overlooked this thread earlier.

      I think what you want is a confidence interval for your ROI. A rough 95% confidence interval after n tournaments is your observed ROI +- 2 SD/sqrt(n), where SD is the standard deviation for one tournament. For 9-player SNGs, the standard deviation is about 1.5 buy-ins. For 45 player tournaments, the standard deviation is about 2.8 buy-ins. For 180 player tournaments, the standard deviation depends more on your playing style, but it may be 5-8 buy-ins, more for a big winner.

      So, if you play 100 45-player tournaments and have a 20% ROI, you can say your 95% confidence interval is 20% +- 2(280%)/sqrt(100), or 20% +- 56%. This means you have strong evidence that your ROI is between -36% and +76%.

      It takes a very long time to figure out how often you win MTTs. You have to play long enough that you win a few times. Much of your ROI comes from the times you win the tournament, so until you play enough to win the tournament a few times, you can't expect an accurate estimate for your ROI.
      Do you know the figure for a 1000 man scheduled - in the ballpark?
    • pzhon
      pzhon
      Bronze
      Joined: 17.06.2010 Posts: 1,151
      A rough approximation is that the standard deviation grows as the number of participants to the power 0.4. So, with 1000 participants, you expect to see a standard deviation of about (1000/180)^0.4 times the standard deviation for 180 player tournaments, or about 9.5 buy-ins.