BTN opening

    • Kyyberi
      Kyyberi
      Coach
      Coach
      Joined: 09.07.2010 Posts: 10,511
      There are two different ways to play in position preflop. One is to open with a wide range from BTN/CO and one is to open with stronger range. I made a little math to find out how it really is.

      The Myth:

      In PLO25 Zoom, folding equity preflop (from BTN against the blinds) is so low that one shouldn't open trashy hands.

      Lets assume that the Myth is right. Of course we open any4 against nits, but against unknown players we open only for hand strength value. We are not facing a lot of 3bets from the blinds, so we don't have to adjust our betsizing against that. So we should open full pot to get the maximum value. Here is why:

      When we open with strong range, we will often hit stronger than our opponent. And with good hands, we will hit more than we miss. When we compare 2bb and 3,5bb raises preflop to the amount we will win postflop when we cbet 3/4 pot otf, ott and otr, it looks like this:

      Pot Pre - Flop - Turn - River (our value)
      4 - 10 - 25 - 62,5 (31)
      7 - 17,5 - 43,75 - 109,4 (55)

      So that small difference preflop is quite big postflop.

      BUT!!! Is the myth right?

      I looked at my database and found out that when it comes to BTN unopened, and hero opens, blinds actually fold 50% of the times on average. When hero gets called, half a pot cbet is succesfull 40% of the times. Lets analyze a situation where hero doesn't even look his cards, but plays stab and fold poker (raises pre, SB folds and BB calls, makes a half a pot cbet and gives up if called/raised).

      Raise 3,5bb = preflop we risk 3,5bb to win 1,5bb. To be breakeven we need 70% FE. We have only 50% so we lose -1bb in EV. OTF we cbet 3,75bb to a pot of 7,5bb. We need 33% FE to breakeven, so our EV is +0,75bb. So our total EV is -0,13bb.

      Raise 2,5bb = preflop we risk 2,5bb to win 1,5bb. To be breakeven we need 63% FE. We have only 50% so we lose -0,5bb in EV. OTF we cbet 2,75bb to a pot of 5,5bb. We need 33% FE to breakeven, so our EV is +0,75bb. So our total EV is +0,13bb.

      So if hero doesn't even look his cards and makes cbet always against one opponent, the other is -EV and the other is +EV.

      Results: if you open with a wide range from the BTN, it's better to open 2,5bb than 3,5bb. If you open with valuerange, you should open 3,5bb.

      Note: these are just the vacuum lines against unknown opponents.

      edited post with precise calculations 4.6.2014
  • 23 replies
    • PerusJamppa
      PerusJamppa
      Bronze
      Joined: 18.01.2008 Posts: 2,714
      Originally posted by Kyyberi
      Results: if you open with a wide range from the BTN, it's better to open 2,5bb than 3,5bb. If you open with valuerange, you should open 3,5bb.
      If FE is exactly the same regardless your bet size and you actually play stab and fold poker, i guess this is the truth.

      However, i'm not sure is it really that simple..
      Like many other theories, this one have that common problem called "real world." :D
      What it means is you're not folding when you have quads otf and villain raises, etc, etc.

      Based on your chart....
      Pot Pre - Flop - Turn - River (our value)
      4 - 10 - 25 - 62,5 (31)
      7 - 17,5 - 43,75 - 109,4 (55)

      .... when you actually have a good hand and are getting value, on average you get it more when you've raised full pot pre.
      So i think that you should somehow add this section to these calculations. I'm too lazy (read: stupid) to do the math, but i'm quite sure that with 100% range it's not enough to gain even nearly that whole gap up, but at least it should even it out a little bit.
      Outcome would be probably the same (-EV open pot with 100% range), but maybe not exactly that -0,7bb.
    • Skodljivec
      Skodljivec
      Bronze
      Joined: 17.12.2011 Posts: 5,709
      Imho "trashy" (but playable ofc) hand IP = Value hand OOP.

      In your example, you are suggesting we don't even want our opponents to call us preflop? But isn't it better for them to call preflop, but fold on future streets? More dead money? :D
    • Kyyberi
      Kyyberi
      Coach
      Coach
      Joined: 09.07.2010 Posts: 10,511
      Of course it's not that simple :)

      But if you have trashy hand, which is primary for stealing, you are happy if blinds fold preflop. Sometimes they call, and often you cbet and are happy if they fold. On the other hand, when you have a strong hand you want to get as much value as possible. So if most of your range is value hands, raise bigger. If most of your hands are for stealing, raise smaller.

      The calculations are something to base our gameplan on. It answers some questions. The best would be to raise accordingly to our hand strength, but that would be too transparent for most players. So if we want to open the same with steals and valuehands, we need to figure out the best sizing based on our range (value vs steals).
    • Kyyberi
      Kyyberi
      Coach
      Coach
      Joined: 09.07.2010 Posts: 10,511
      Originally posted by Skodljivec
      Imho "trashy" (but playable ofc) hand IP = Value hand OOP.
      Are you sure?
    • PerusJamppa
      PerusJamppa
      Bronze
      Joined: 18.01.2008 Posts: 2,714
      Imho "trashy" (but playable ofc) hand IP = Value hand OOP

      I guess this should be other way around. :D

      Before i write an essay, are you questioning your current bet size, or are you planning to widen your opening range?
      Yes, i've been following you guys dicussing on PT -forum about this matter. ;)
    • Kyyberi
      Kyyberi
      Coach
      Coach
      Joined: 09.07.2010 Posts: 10,511
      If I open a lot from the BTN, I adjust my betsizing.

      I wasn't questioning my betsizing, when I open for hand value from BTN. I was questioning if my assumption about the EV for opening a ton from the BTN was right. It wasn't. And by "a ton" I mean +85%.

      And of course, if we have a reason to adjust, we adjust. As in all things.
    • Arzapy
      Arzapy
      Bronze
      Joined: 27.10.2009 Posts: 28
      I dont understand, if your 3,5 preflop raise has a -EV of 1BB and your flop CBET has an ev of 0,75 then why your total ev is -0,13BB? Shouldnt it be -0,25BB (-1+0,75)?
    • Kyyberi
      Kyyberi
      Coach
      Coach
      Joined: 09.07.2010 Posts: 10,511
      Cos they both have 50% chance of happening. You can't just deduct one from another, you have to use percentage. So it goes:

      (-1bb*0.50)+(+0,75bb*0.50)=total EV

      The reason why just deducting one from another doesn't work can be seen if you think the other instance has 60% chance of happening and the other has 40%.
    • Arzapy
      Arzapy
      Bronze
      Joined: 27.10.2009 Posts: 28
      From already thank you for your answer, my problem now is that I can not get the results when we open 2,5bbs. If you could look my calculations, I would appreciate if could you please correct me. greetings.

      OPENING TO 3,5 BBs

      BEFORE THE FLOP (betting 3,5 to steal 1,5)

      - 3,5*50% + 1,5*50% = -1,75 + 0,75 = -1bb

      AFTER THE FLOP (betting 3,75 o steal 7,5)

      -3,75*60% + 7,5*40% = -2,25 + 3 = 0,75bb

      Then our total ev is -1bb*0,5 + 0,75bb*0,5 = -0,125 bb

      ----------------------------------------------------------
      OPENING TO 2,5 BBs

      BEFORE THE FLOP (betting 2,5 to steal 1,5)

      - 2,5*50% + 1,5*50% = -1,25 + 0,75 = -0,5bb

      AFTER THE FLOP (betting

      -2,75*60% + 5,5*40% = -1,65 + 2,2 = 0,55 bb

      Then our total ev is -0,5bb*0,5 + 0,55bb*0,5 = 0,05 bb
    • Kyyberi
      Kyyberi
      Coach
      Coach
      Joined: 09.07.2010 Posts: 10,511
      It's good that someone checks my numbers as I haven't included the calculations to the original post. I don't know how I did it, but in 2,5bb calculations there is a clear mistake.

      After the flop EV is indeed 0,55bb. Total EV is 0,025bb (not the 0,05 you had unless you rounded it up).

      I think this clearly shows that when posting results of calculations, it's good to post those calculations too. So everyone can check if there are some unnoticed errors. :)
    • Arzapy
      Arzapy
      Bronze
      Joined: 27.10.2009 Posts: 28
      thank you for uot answer, I was really cracking my head trying to find my posible mistakes, lol. best regards, (sorry for my bad english)
    • Krach-Bumm-Ente
      Krach-Bumm-Ente
      Black
      Joined: 01.05.2006 Posts: 10,405
      The sheer weakness of a wide BTN range makes us not want to open with a large raise. Some top players even openlimp the BTN.
    • AlvisR
      AlvisR
      Bronze
      Joined: 27.02.2009 Posts: 656
      what i implemented in my game - if my plan is to steal i think of

      2 factors:
      1st are there fishes who are willing to play from blinds
      2nd are there somewhat aggressive villains who like to 3bet

      If 1st: i like to steal 3bb big because vs smaller stealing they like to play more often (here comes fold equity)
      if 2nd: i like to steal smaller 2,1bb 2,2bb 2,3bb 2,4bb 2,5bb 2,6bb etc because we will face 3bets often enough and it will be easier to play smaller pots (pot control?)

      and this is in micros of holdem. can you implement this in omaha? :p
    • Kyyberi
      Kyyberi
      Coach
      Coach
      Joined: 09.07.2010 Posts: 10,511
      Originally posted by AlvisR
      If 1st: i like to steal 3bb big because vs smaller stealing they like to play more often (here comes fold equity)
      Is this based on any statistics or is it your belief?
    • AlvisR
      AlvisR
      Bronze
      Joined: 27.02.2009 Posts: 656
      stats on 200k hands from 6max cash games

      do you think its skewed up and i shoulnd believe in such a fact?
    • Kyyberi
      Kyyberi
      Coach
      Coach
      Joined: 09.07.2010 Posts: 10,511
      Do you have sample size on both 2bb and bigger btn open on same stakes? If yes, what is the folding % of villains in both cases? I am very curious, cos I havent seen that on decent sample size ever.
    • Krach-Bumm-Ente
      Krach-Bumm-Ente
      Black
      Joined: 01.05.2006 Posts: 10,405
      are we talking 200k btn steals or 200k hands total?
    • Kyyberi
      Kyyberi
      Coach
      Coach
      Joined: 09.07.2010 Posts: 10,511
      200K BTN steals would be pretty big sample size. :D

      I mean what are the sample sizes for different BTN openings in the same stakes? As you can't compare if you open bigger on PLO25 and smaller on PLO50.

      So have you experimented with different BTN opening sizes in same environment (stakes)?
    • AlvisR
      AlvisR
      Bronze
      Joined: 27.02.2009 Posts: 656


      250k hands nl5+nl10 zoom

      we are interested in BU sizing. Here is some sample.

      we care only on non sd line do you see how big difference might be when hand count would be equal? Steal range about 46-48% (holdem) no im not stealing tighter with bigger opensizes.
    • 1
    • 2