Edge% in Sng wizard

    • Leito99
      Leito99
      Bronze
      Joined: 27.07.2009 Posts: 754
      What should you set the edge% in sitngo wizard

      I know it depends on how many BB so lets say at 10bb/5bb/20bb

      Is their default usually best?

      Does it depend on your position as well?

      Thanks!! :)
  • 9 replies
    • pzhon
      pzhon
      Bronze
      Joined: 17.06.2010 Posts: 1,151
      The edge should almost always be set to 0%.

      The edge % is vestigial. It is a remnant of the time when push/fold calculators used a short cut. They assumed that if you chose to fold, the hand ended instead of allowing your opponents to collide. This undervalued folding, and so an edge % was added to counter this error. SNG Wizard no longer takes this short-cut, but it still allows you to "correct" for an error it doesn't make, which now creates an error instead.

      The patterns used by SNG Wizard to make the errors vary by position, stack size, etc. do not reflect how you should deviate from the ICM in theory based on your position relative to the blinds and a skill advantage over your opponents.

      As I have discussed, I think in my video "Independent Chip Model," you can try to account for the cost of hitting the blinds. When you are UTG, folding may be worth less than the ICM says. You can adjust for this by using a small negative edge whose magnitude I showed how to compute. The value does not agree with the values suggested by the edge models within SNG Wizard.

      Using a positive edge % at the wrong time won't increase your edge over your opponents. It will damage your game.
    • Leito99
      Leito99
      Bronze
      Joined: 27.07.2009 Posts: 754
      Originally posted by pzhon
      The edge should almost always be set to 0%.

      The edge % is vestigial. It is a remnant of the time when push/fold calculators used a short cut. They assumed that if you chose to fold, the hand ended instead of allowing your opponents to collide. This undervalued folding, and so an edge % was added to counter this error. SNG Wizard no longer takes this short-cut, but it still allows you to "correct" for an error it doesn't make, which now creates an error instead.

      The patterns used by SNG Wizard to make the errors vary by position, stack size, etc. do not reflect how you should deviate from the ICM in theory based on your position relative to the blinds and a skill advantage over your opponents.

      As I have discussed, I think in my video "Independent Chip Model," you can try to account for the cost of hitting the blinds. When you are UTG, folding may be worth less than the ICM says. You can adjust for this by using a small negative edge whose magnitude I showed how to compute. The value does not agree with the values suggested by the edge models within SNG Wizard.

      Using a positive edge % at the wrong time won't increase your edge over your opponents. It will damage your game.
      Thanks for the detailed answer!

      I play the superturbos and I was using ranges with 0 edge %. However, after watching a faarcyde video, he mentioned in it that his standard for edge % is .1 when he is higher than 10BB and then almost 0 for when he gets shorter than that. I adjusted my ranges and in the last 300 games my $EV has been rising noticably faster.

      This could be just variance or more likely makes up for the odd calls some fish make with hands like QJs, 22 etc
    • pzhon
      pzhon
      Bronze
      Joined: 17.06.2010 Posts: 1,151
      I'm glad your recent adjusted results have been good, but I wouldn't extrapolate from a sample of 300 for something like this. The all-in adjustment works well for super turbos, but after 300 tournaments your confidence interval should still be about your observed ROI +- 10%.

      If you think your opponents are calling wider on average, then you should adjust the ranges instead of using a minimum edge. Never blindly trust the SNG Wizard default ranges.

      Although you get to see bad or ugly calls, you don't get to see bad folds of hands which are easily good enough to call. I reviewed aggregate calling ranges in my video "Super Turbo Ranges" and while in some situations players call wider than the Nash equilibrium, in others they call more tightly.
    • petersveter
      petersveter
      Bronze
      Joined: 18.10.2011 Posts: 339
      wow this is pretty shocking for me...I have been working on some numbers for how wide I should be pushinng from each possitions in each stack sizes. I now realized that i was using the edge that SnG Wizard gave me.

      When I get back to my numbers, lets say 10 BB and am on the BUTTON, with the avarege tight blinds, I could steal with about 50 % hands. Now that I changed the edge to 0 % I can push 82,5 %.

      Do you think it's fine leaving the edge on 0 from 10 bb's lower? this will really change my pushing ranges thats why I'm asking :s_confused: :)
    • kurrkabin
      kurrkabin
      Bronze
      Joined: 12.10.2010 Posts: 5,976
      Hi again, petersveter!

      The edge has primary to 2 purposes-to reduce variance when you are deep stacked and to help account for future hands when you are very short stacked.

      When we are big stack, we have a lot of opportunities to exploit our opponents. We are not in a hurry to make a move. Therefore, we should not be willing to stack off with marginal hands that small + EV winners as that will also increase the variance a lot.
      Usually I will give myself some negative edge when I don't have a healthy stack and FE decreases or is about to decrease significantly in the next hand- f.e. the famous minus EV push with 5bb from UTG.

      Please keep in mind that nash ranges always operate with edge set to zero. With 10bb and fishy opponents I think I will still give myself some edge and not be willing to make some very close + EV pushes as we are still deep enough to pick our spots. However, it is really player dependent. Fishes usually have more static calling range- that means that they don't adjust accordingly when effective stack sizes go down in numbers. Therefore, we can usually push a super wide range bvb vs a fish, because they won't be willing to call as wide as a reg will. On the other hand, they also won't be willing to push as wide as a reg will, so our call ranges have to tighten up.


      You better post the situation as a lot of factors matter, but honestly, I don't see myself pushing 82-83% on the BTN with 10bb eff.
    • Kamikaze001
      Kamikaze001
      Bronze
      Joined: 20.11.2007 Posts: 5,832
      Ok, now I'm a little confused about the edge topic.

      Am I correct to think that for the reasons Phzon explained the edge should always be set to 0, unless
      1) hero either has a very healthy stack -> small positive edge to decrease ur own be/slightly+ev-pushing range or
      2) a very small stack -> small negative edge to widen our shoving range for a future game +ev perspective
      ?

      If I'm correct so far, what numbers would u usually put in for "small" positive/negative edges?
      And do u sometimes increase the edge numbers and if so, for what reasons and to what numbers?

      FYI I usually play mtsngs on Stars.
      Probably stupid question: But is there any correlation between payout structures and the edge I should put in?

      :)
    • kurrkabin
      kurrkabin
      Bronze
      Joined: 12.10.2010 Posts: 5,976
      Hi, Kamikaze001!

      Sorry a bit for the delay.

      I think you are correct and I am also almost certain that Phzon explained the same things. I will, however, also give myself an edge if I assume people will do many mistakes pre and postflop, which we benefit from. Say we are not that deep(as first point suggests), SB and BB and are the two midstakes, CO is the big stack and we are sitting in the BTN as the shortstack. CO is very aggressive and abuses the bubble, while SB and BB are very loose aggro fishes that are unfamiliar with risk aversion and will simply call a hand that they think is ahead of CO's range or maybe even spite him. If CO folds, then it's very likely that the blinds will play with eachother and do mistakes, which we benefit from!

      That will be a reason for me to give myself a small edge, considering we can get a free roll ITM.

      Small stacks-yes, as loosing FE is crucial, so taking a negative edge is reasonable and it will also advocate for looser ranges.

      As for you last question- I asked a fellow coach for a bit of a help :) We will be back shortly.
    • lessthanthreee
      lessthanthreee
      Bronze
      Joined: 30.06.2009 Posts: 16,300
      there is no direct correlation between edge and payout structure.
    • Kamikaze001
      Kamikaze001
      Bronze
      Joined: 20.11.2007 Posts: 5,832
      Thanx a lot guys :)