[5CD] Learning resources

    • BaggerSmurf
      Joined: 11.12.2010 Posts: 896
      I've recently given 5 Card Draw a try, and have come to like it a lot :f_love: .

      I have read through the articles available here on PS and watched the videos by EvGalois. But I would like to improve further! I have been playing poker for quite a while and am able to apply some of my previous knowledge, but maybe I'm developping a nitty habit for example. And those early habits are hard to get rid of :D .

      Could anyone recommend me a good place to learn more? Any more articles, videos, coachings or hand judgements available anywhere?

      Thank you in advance :) .
  • 26 replies
    • Hurrrrrrrrrrrrrrrrrrrricano
      Joined: 23.10.2012 Posts: 2
      In before jbpatzer and vorpalF2F

      // whoa sorry for weak english and being drunk:)

      EVG is a very fishy and nitty player, so you rather do not want to learn from him.
      Instead just search 2+2 for "Bigpooch"/"mangler241" posts and make sure you're head can take a math within it.

      Don't read any Wiesenberg or Mike Caro or other Hellmuths, because they are both nits and are ages ago behind our Hong Kong friend.

      You can find one or two articles on pokermedley.com or sth like that (guess who write them :f_confused: )
      or you can take a coaching there (guess who is the coach :f_confused: )
      or you can simply post a question on 2+2 (guess who will reply them :f_confused: )
      or simply read 2+2 DOP forum (guess who is the best poster there :f_confused: )
    • jbpatzer
      Joined: 22.11.2009 Posts: 6,944
      You are such a BigPooch fanboy. I watched EVGalois vids and made lots of $$$. I'm sure there must be a middle way.....
    • Hurrricano
      Joined: 23.10.2012 Posts: 60
      :evil: You are such a patzer. I watched Al Bundy's videos and also made $$$
      You were playing with Spanish players!
      If you were outside the Spain and two gamble-invalid nations: Australia and Canada, than you already was a winner :D
    • jbpatzer
      Joined: 22.11.2009 Posts: 6,944
      Originally posted by Hurrricano
      :evil: You are such a patzer. I watched Al Bundy's videos and also made $$$
      You were playing with Spanish players!
      If you were outside the Spain and two gamble-invalid nations: Australia and Canada, than you already was a winner :D
      Remind me how many r's there are in Hurrricano?

      I have no idea who Al Bundy is and cba to google it.

      And yes I am a pattttttttttzzzzzzzzer.

      And yes, R.I.P. the Spanish players. I will try to remember never again to think I had any idea what I was doing. Silly me.
    • BaggerSmurf
      Joined: 11.12.2010 Posts: 896
      So far I've been recommended articles/videos/posts of BigPooch and EvGalois.

      Just read through these two:

      Any more suggestions :) ?

      Could you guys only pick on each other in posts that also contain suggestions useful for this topic :D ?

      While I was looking through some of the hand evaluations here I've run into the term 'snow'. I have absolutely no idea what it means in a poker context :D . Could someone enlighten me?
    • Crox1710
      Joined: 11.10.2012 Posts: 22
      Yeah I'm also getting into 5CD and would like some more hints. I play at micro stakes right now and can win solidly but i know there are holes in my game, especially post draw.

      EVG is a very fishy and nitty player, so you rather do not want to learn from him.

      How so? 5CD seems to be quite a nitty game, theres not much room to maneuver especially on FL.

      I reckon a hand analysis thread would be great, I'll compile a few in the next day or so to get peoples opinions.
    • Hurrricano
      Joined: 23.10.2012 Posts: 60
      Let's make a one all-around 5 card draw thread...
      Theory, hand evaluation, reads, notes and boobs in one place :f_p:
    • jbpatzer
      Joined: 22.11.2009 Posts: 6,944
      Originally posted by Hurrricano
      Let's make a one all-around 5 card draw thread...
      Theory, hand evaluation, reads, notes and boobs in one place :f_p:
      I'll show my boobs if you show yours.

      Tell them your 'you can play really loose at the micros' theory. There may be something in it.
    • Hurrricano
      Joined: 23.10.2012 Posts: 60
      I've just found a good example of snowing with player comments!

      jbpatzer sits at table with EVGalois
      jbpatzer tries to chat and play at the same time
      jbpatzer is distracted by both EVGalois and number two son
      jbpatzer clicks the raise button not noticing UTG has raised
      jbpatzer cringes when UTG caps
      jbpatzer facepalms and decides the best plan is to snow
      jbpatzer fails to notice UTG only has 0.5BB left
      jbpatzer is an arse

      Poker Stars $1/$2 Limit 5 Card Draw - 6 players
      DeucesCracked Poker Videos Hand History Converter

      Pre Draw: (1.5 SB) Hero is BTN with K :diamond: K :heart: 6 :club: 5 :club: 3 :heart:
      Hero says "you're not really making a vid r u?", Candice0928 raises, sando8333 folds, EvGalois folds, Hero 3-bets, iivanoff folds, boymark folds, Candice0928 caps!, Hero calls

      First Draw: (9.5 SB) (2 players)
      Candice0928 draws 2, Hero stands pat
      Hand: K :diamond: K :heart: 6 :club: 5 :club: 3 :heart:
      Candice0928 bets all in, EvGalois says "yes right now.", Hero calls

      Final Pot: 5.75 BB
      Candice0928 shows A :diamond: Q :heart: 4 :spade: 4 :diamond: 4 :heart:( three of a kind, Fours)
      Hero mucks K :diamond: K :heart: 6 :club: 5 :club: 3 :heart:
      Candice0928 wins 5.5 BB
      (Rake: $0.50)
    • jbpatzer
      Joined: 22.11.2009 Posts: 6,944
      Nice example. I think I come out of the whole thing quite well.

      And lol @ EVG making another vid. :facepalm:
    • Hurrricano
      Joined: 23.10.2012 Posts: 60
      Quick guideline to stealing from Button:
      (do I have to post a author nick? :f_confused: )


      Here's a simple artificial "toy game" to look at: suppose
      there are only three players with the SB posted = exactly
      1/2 of the BB =$2, and there is no rake. The only option
      for the button is to raise to $4 or fold and the most anyone
      can make it is $4 to go (no other raises).

      If the button is raising, he is risking $4 to win $3, so the
      blinds need to defend with a frequency of at least 3/7;
      otherwise, the button will simply profit from bluffing. This
      would suggest that they defend 3/7 of the time, or fold
      about 4/7 of the time. Now, 1-(4/7)^(1/2) is about
      0.244071, so the BB needs to defend with at least 1/4 of
      the hands since the SB won't be defending as often simply
      because he has to put in more money and there is some
      chance the BB has a better hand.

      To be more precise, suppose the button has to give in
      advance his opening range, and say it is precisely [0,1/3].
      Obviously, since the BB is only risking $2 for a pot that will
      have $9, he only needs to win 2/9 of the time, so he simply
      needs a hand in [0, (1/3)(7/9)] = [0,7/27] if the SB has
      already folded.

      What does the SB do if the button raises? Here, things are
      a bit complicated now. Although he would be risking $3 to
      play in a pot that could have only $10 in it if the BB decides
      not to play, the exact hand (1/3)(1-(3/10))=7/30 will show
      a loss because there is some small chance the BB will play.
      Assume now that the SB has to give in advance his calling
      range when the button opens. Then, given that range, one
      can determine with which hands the BB can play (which
      will be even tighter). In any case, we know that for the
      SB playing the range [0, 7/30] is slightly too loose. For
      example, if the SB is playing [0, 0.22] it's still too loose
      since the BB can play 0.15 since his chance of winning with
      0.15 is then about 0.175 > 1/6. Now the SB playing the
      hand 0.22 only has at most a [1-0.22/(1/3)][1-0.15]=0.289
      chance of winning which is strictly less than 0.3, his "pot

      According to the above analysis, the SB only plays less
      than 0.22 of his hands and the BB only plays 7/27 of his
      hands once the SB folds, so these blinds "should" only be
      playing less than 0.22+(1-0.22)(7/27)~0.4222 of the time.
      If so, since the above number is strictly less than 3/7, this
      suggests the button's opening range is "too tight", i.e, he
      should be raising with more than 1/3 of his hands.


      At 6-max fixed limit, against tight blinds you can raise with
      as little as AK from the button (drawing three or standing
      pat if you don't have AKQJ to draw to), but you obviously
      can't do it all the time. You clearly don't want to raise
      with AQ because not having a king makes it more likely
      you'll get looked up by someone with KK (unless both of
      the blinds are oblivious).

      Do you really think that a raising strategy like 99+ plus some AK's OTB is good ?

      It works in practice, in fixed limit against specific blinds, if you
      must know. Theoretically, it could also be correct when you
      have something like AKQT, AKJT or AKQ9x since it's that much
      more likely the blinds won't have a pair of nines or better. The
      EV will be positive often enough at many of the fixed limit
      tables, but obviously won't work at some tables and even
      against "ideal blinds" it's not gaining much EV. If the big blind
      will defend with 99 (as he often should) and the legitimate
      range includes some pairs of eights, the button needs to bluff
      with 88 or weaker when the big blind has drawn three. If the
      button value bets AA (theoretically correct when heads up
      versus another three card draw), either the button would have
      to bluff with 99 (which has a tiny bit of showdown value
      especially against a tencacious big blind) or choose a range of
      weaker one pair hands to open raise with and draw three to
      (e.g., hands such as 77AKx or 22AKQ). The bad thing about
      pairs weaker than 88 is that they'll miss about 71.29% of the
      time and have "showdown value" (say make a pair of tens or
      better) less than 29% of the time, but when these hands
      improve, they are often worth betting when checked to.

      Drawing three to AK will give you a pair of tens or better
      between about 38% and 40% of the time (depending on the
      exact kickers you hold and whether the AK is "suited" or not)
      so it's not hopeless. A bad aspect concerning AK is that
      making a pair of aces will sometimes lose an extra bet
      postdraw when heads up against one of the blinds. Another
      bad feature is that when a blind reraises, you must surrender
      immediately; if you had a "normal semibluff", calling and
      drawing one would be standard.

      Do you remember that "toy game"? If you are playing more
      than 34% of the hands on the button, you probably want to
      add some hands to the range of 99+ and decent semibluffs.
      [ Work out how many hands you need to add; obviously,
      88AXy, and some lower pairs such as 55AKQ should be
      included. ] Assuming you won't be semibluffing with really
      weak draws from the button such as JT98x or ten-high flush
      draws, it's hard to get to 34% by adding to your "legitimate
      range" and semibluffing hands. The 34% is probably not
      even high enough theoretically because the solution to the
      "toy game" is about 0.3425 and from LHE practice (although
      it's an entirely different form of poker) the range is close to

      When the blinds are a bit too tight, you're right that having
      an ace and a king will make it less likely that the SB will call.
      With something like:

      A Spade K Spade Q Heart x Diamond y Club

      putting in raise here is good not only against tight blinds, but
      blinds that fold too often postdraw. Of course, the above
      hand is suited (adds about 1%) and having a queen makes it
      more likely nobody will play. It is important to value bet AA
      on the button when heads up against a player who has drawn
      three since you'll end up with a bust fairly often when opening
      with AK (suited) or AKQxy (not to mention when you don't
      improve when you open with 22AKQ). Against an opponent
      that draws three and never (or infrequently) checkraises, AA
      is a clear value betting hand if you work out the details from
      one of the MOP "toy games".

      In my earlier days, I would usually raise with AKQJ (usually
      draw one), AKQT or AKJT (drawing three), but it's better to
      open with any AKQ hands against players that fold too often
      predraw or postdraw. A caveat is that against light reraisers
      in the blinds, you probably want to restrict these type of
      semibluffs to when you have a queen and maybe when you
      are "onsuit" (even having a ten or a nine is of some help).

      A possible alternative is to include hands that look like a small
      pair with an ace and king side kicker (instead of so many AK
      hands), but it's not nearly enough to get the total open
      raising range on the button to above 34%. Of course, it may
      be the case that the "optimal" frequency from the button is
      less than 34%, but IMHO it's unlikely since from empirical data
      on LHE, the practical frequency is closer to 40%. I think it's
      not too far from optimal to open raise with about 35% of all
      hands from the button in fixed limit draw. Then, it's up to
      the individual to construct this range according to the type
      of opponents he will often be facing from the blinds.

      Should my range on the button include all pairs such as 88 or 77?

      It's not that "terrible" if somebody did include some pairs of
      eights for their range of hands, but would a pair of 77's be
      classified as a "semibluffing" hand? If somebody open raises
      with every pair of sevens from the button in limit draw against
      any decent opposition in the blinds, it wouldn't take much
      time or "rocket science" to figure out that it's going to be a
      "loss leader" in that a pair of sevens is going to "get caught"
      a bit too often. Now, if you held precisely the hand 77AKx,
      it's much better: you won't get caught bluffing as often
      postdraw, and you'll more likely have the best hand predraw
      as compared to something like 77J32.

      I don't have Zadeh's "Winning Poker Systems" with me any
      longer, but if you read the fine print in the "Ante and
      Straddle" section of his book, 77 is the minimum opening hand
      from the button, even though technically, this is for the limits
      2xBB - 4xBB (I clearly remembered the chart looked like
      JJ/hijack, 99/cutoff, 77/button, A/small blind). Zadeh likely
      worked out that "toy game" alluded to from your "Poker
      Theory" thread or at least had the computational means to
      find numeric solutions which he then "translated" to actual
      ranges of draw hands. Unfortunately, he didn't mention much
      about semibluffing hands such good flush draws (he did
      mention KQJTx somewhere though). In any case, for the
      three-handed "toy game", the button's optimal opening
      frequency is about 0.342527 or so and for limit draw, it could
      be a shade higher. If you take that number and "translate" it
      to a range of hands, it will be a pair of sevens AND about
      16 194 other combination of hands which can be "covered" by
      very good semibluffing hands (straight flush draws and A-high
      or better flush draws that don't have at least a pair of sevens).
      [Here, I'm ignoring card removal effects when three or four
      players have passed in front of the button, thereby skewing
      the frequency very slightly towards aces, kings and queens.]

      You did hit upon a key idea: having an ace, king or queen is
      nice because it will be less likely that the SB or BB will be
      able to call postdraw (it turns out a jack is important too
      but it's primarily for predraw).

      It's clear that some other hands should be used in the range
      of one card draws instead of most 77 hands. So, which ones?


      Well, suppose you open on the button, are heads up against
      the small blind in a pot that has 2.5 big bets (forget the rake
      for now) and so he's very likely drawing three if he is any sort
      of player (say his range is approximately JJ to KK and some AA
      hands he decides not to reraise with).

      You draw one and he checks. According to theory, if you bet,
      he should call you 5/7 of the time if he is not to be exploited.
      Well, if your range is rarely worse than AA, why should he even
      bother to pay you off with an unimproved pair of aces? (never)
      [ Some of those one card draws will be to trips and quads. ] If
      the small blind "always checks" to the button, he needs to pay
      off with almost 59.92% of unimproved one pair hands, so even
      if he never reraises with any pair of aces, his theoretical calling
      hands would include not just KK but also about 39.68% of QQ
      hands, so as not to be exploited (assuming his range is exactly
      JJ-AA) if the button's range of one card draws include enough
      "semibluffs" that "bust out".

      Thus, there are two reasonable "methods" of constructing a
      range of hands with one card draws on the button:

      1) drawing one to some range of one pair hands, such as
      one pair hands from 66 to TT.

      2) drawing one to some range of "semibluffing" hands that
      include not just straight flush draws, but also good flush
      draws, KQJT, AQJT9, and AKQJx (some debate the merits
      of this, but I think drawing one to this is fine, even though
      if the AK were "suited", drawing three is "reasonable").
      [It's also possible to "mix" 1) and 2) above and this is likely
      to be a decent approach if 1) is used sparingly. ]

      One of several "flaws" with 1) above is that chances of
      improvement are only 8/47 (identical to an openender), so
      why bother? [ The best hand you can make is only trips! ]
      Now, it's NOT out of the question with a hand such as 66AKx
      (probably draw to ace and x if the small blind has some
      chance of reraising with AA).

      Then, it is possible to "construct" a range of "bluffs" that
      will force an opponent to sometimes pay off with an
      unimproved pair as low as QQ. Of course, we will have to
      consider the other common scenario when heads up versus
      the big blind and the less common scenario when against
      both blinds, reraised pots, etc. Thus, one should be able
      to see that including some range of flush draws will be
      necessary to have flexibility in what "busts" to bluff with.

      Okay, how many "reasonable" flush draws are there? If
      you include all flush draws, the number I get is 110 940
      which includes any pair (obviously, some are legitimate
      hands) and some flush draws that probably should be
      tossed. [ Also, there are straight draws such as KQJTx
      and AQJT9 (only 7680 combinations). ] There can't be
      "too many" as this has to be considered in light of the
      scenarios primarily when heads up with a blind, but it's
      probably not too far off if the number of combinations is
      not far from 84 480 (the number of combinations for
      precisely a pair of sevens) if you work out the nitty


      Then, it IS possible to use some range of semibluffs and
      enough hands from 77 or less to make up another 16 000 or
      so combinations. Something like XXAKy where X = 3 to 7
      "works" because the blinds aren't going to be paying off
      as often postdraw with an unimproved one pair.

      Is including ALL pairs of eights going to "work", or will
      there be similar problems as those encountered with 77?

      Now, this is a bit more interesting; surely, these pairs of eights
      are "reasonable" to include in the "legitimate range":


      (X,Y,Z = any card higher than an eight; x,y = any).

      Since I've looked at the "problem of a pair of eights on the
      button" for awhile (since about 2005!), I'm not convinced any
      pair of eights unmentioned above will "work". It is important
      to realize that the BB (and to a lesser extent, the SB) should
      often reraise with AA against a button that opens with a wide
      range that includes 88. Then should you surrender most pairs
      of eights when faced with a reraise? When not holding an ace
      with a pair of eights, it's a bit more likely somebody with AA
      will reraise. Not holding many quality overcards is bad for
      several reasons with precisely 88. Contrast that with a hand
      like 77AKx: it's less likely somebody with AA or better will
      reraise and less likely that somebody with only a pair will pay
      off on the end versus a "bust".

      If we are "fine tuning" our range on the button, which other
      pairs of eights should we "throw out"? I think it's best to
      "replace" 88abc (where a,b,c <88), 88(9/T/J/Q)ab with some
      "other hands". Then, it's up to the individual preference to
      use some mix of hands including XXAYx (X=2 to 7; Y=K/Q)
      and very selective AK-type hands to "semibluff" with, such
      as AK(suited) with an offsuit queen or jack.

      Does 88 have showdown value? Against some opponents
      that will defend with a hand as weak as 22 it does! If that's
      the case, you need to "bluff" somewhat selectively with 88
      but I think you need to "bluff" with some pairs of eights to
      bluff close to the optimal game theoretic frequency unless
      there are many "other hands" that are three-card draws that
      don't end up as good as 88.

      The problem with not having a "spread" of hands (even if it's
      "thin") from 77 down to 22 or AK when one draws three is
      this: if your range of "bluffs" is including most unimproved 77,
      your opponent has information: if he sees a seven or two
      sevens, even with AA, why should he bother to pay off when
      you draw three? Or with something like QQ without having
      seen any seven or eight (if 88 is sometimes used as a "bluff"),
      he should pay off routinely. That's why it's nice to include
      the hands XXAKy (X<=7) for open raising from the button to
      prevent an extremely sophisticated opponent to use any
      information from his cards.

      In practice, most players in the BB won't defend with 44 or
      lower (dont' blame them!), but there's a few hands worse
      than 88 that the BB should strongly consider defending with
      and drawing three to. Of course, that's because many
      players on the button will be raising with a frequency much
      lower than 1/3 (which means they aren't getting "full value").

      All this is based on the frequency of 0.342527 which is likely
      "too low" theoretically! Can we get away from including AK
      hands worse than AK(suited)(Q/J)xy? A good question is
      what AK-type hands will work in practice.

      [Long question here...]

      This is very difficult to say for limit draw, but the button has
      position and that will give him hands to value bet when
      checked to (e.g., when the BB calls and draws three, the
      button draws three; I'm almost 100% sure that KKA is a
      "value bet" but for some reason that doesn't happen except
      at the highest limits). If I had to pick a frequency for the
      button's opening range in limit draw, I'd guess around 35%
      (maybe more!) since HE is the one who can often outdraw
      either the small blind or big blind. For most of this post,
      I'm going to consider that problem hand "AK", so bear with
      me! (As discussed before, I still prefer to have a queen or
      some "key cards" in the nine to queen range with an AK; e.g.,
      AK876 is a "fold predraw" from the button.)

      A useful tool is the twodimes.net poker calculator. For high
      draw, use "deuce to seven lowball" and invert the equities
      (this can only be used for heads up; for 3-handed, it's not
      possible to use directly, but I haven't bothered to try to use
      it indirectly). Of course, 5432A isn't a wheel but the
      calculator also has a "bug" in that flushes (for some reason)
      aren't counted (you'll see later). There may be better "tools"
      to use, but I haven't bothered to look for them (yet!).


      Sure, in pair versus pair, you're right that the underdog is
      going to win less than 25% quite often; e.g.,

      pokenum -mc 500000 -l27 ts th / qh 9d 2c - js jh / 8s 7h 2h
      5-card Draw 2-7 Lowball: 500000 sampled outcomes
      cards win %win lose %lose tie %tie EV
      Ts Th 379928 75.99 120071 24.01 1 0.00 0.760
      Js Jh 120071 24.01 379928 75.99 1 0.00 0.240

      Exceptions can occur if an "out" is held by the weaker hand
      that is live:

      pokenum -mc 500000 -l27 ts th / 4s 3h 2s - 9s 9h / tc 5s 2c
      5-card Draw 2-7 Lowball: 500000 sampled outcomes
      cards win %win lose %lose tie %tie EV
      Ts Th 132443 26.49 367557 73.51 0 0.00 0.265
      9s 9h 367557 73.51 132443 26.49 0 0.00 0.735

      pokenum -mc 500000 -l27 ks kh / 4s 3h 2s - 9s 9h / kc 5s 2c
      5-card Draw 2-7 Lowball: 500000 sampled outcomes
      cards win %win lose %lose tie %tie EV
      Ks Kh 128322 25.66 371678 74.34 0 0.00 0.257
      9s 9h 371678 74.34 128322 25.66 0 0.00 0.743

      AK vs PAIR

      On the other hand, look at that "ugly hand" again, AK versus a
      pair lower than KK:

      pokenum -mc 500000 -l27 as kd / qh 9d 2c - js jh / 8s 7h 2h
      5-card Draw 2-7 Lowball: 500000 sampled outcomes
      cards win %win lose %lose tie %tie EV
      As Kd 354324 70.86 145671 29.13 5 0.00 0.709
      Js Jh 145671 29.13 354324 70.86 5 0.00 0.291

      Here's why there's a "bug" in the calculator, so we can't
      necessarily believe these numbers are even close to "correct",
      but they may be close enough:

      pokenum -mc 500000 -l27 as ks / qh 9d 2c - js jh / 8s 7h 2h
      5-card Draw 2-7 Lowball: 500000 sampled outcomes
      cards win %win lose %lose tie %tie EV
      As Ks 354730 70.95 145267 29.05 3 0.00 0.709
      Js Jh 145267 29.05 354730 70.95 3 0.00 0.291

      Well, obviously As Ks will make a flush C(9,3)/C(42,3) ~
      0.007317073 of the time, but it's not reflected above (in
      fact, in this run As Ks is doing worse! Here, there's two
      spades out in the other hand, so I'll clean things up a bit:

      pokenum -mc 500000 -l27 as kd / qh 9d 2c - jd jh / 8d 7h 2h
      5-card Draw 2-7 Lowball: 500000 sampled outcomes
      cards win %win lose %lose tie %tie EV
      As Kd 354291 70.86 145704 29.14 5 0.00 0.709
      Jd Jh 145704 29.14 354291 70.86 5 0.00 0.291

      So, there is really a "bug".

      Things are now back under 25% when the pair has one of
      the "outs" for an AK offsuit (suited, the AK will do almost a
      percent better, so AK-suited will still be above 25%):

      pokenum -mc 500000 -l27 as kd / qh 9d 2c - jd jh / ad 7h 2h
      5-card Draw 2-7 Lowball: 500000 sampled outcomes
      cards win %win lose %lose tie %tie EV
      As Kd 375943 75.19 124054 24.81 3 0.00 0.752
      Jd Jh 124054 24.81 375943 75.19 3 0.00 0.248

      But if the AK holds an "out" as well, it's up to 27%:

      pokenum -mc 500000 -l27 as kd / jc 9d 2c - jd jh / ad 7h 2h
      5-card Draw 2-7 Lowball: 500000 sampled outcomes
      cards win %win lose %lose tie %tie EV
      As Kd 364942 72.99 135058 27.01 0 0.00 0.730
      Jd Jh 135058 27.01 364942 72.99 0 0.00 0.270

      So it looks like AK is one of the rare exceptions for the 3-1
      dog "rule" most of the time (but "pathetic" against KK) !


      If the opening range of the button is with about 35% of all
      hands, using a "1/3 rule", a blind will reraise with most (if
      not all) pairs of Aces if he is decent (and aggressive) so
      the button with an AK hand will usually know where he
      stands predraw when heads up with a big blind that draws
      three that just called (clearly, the AK mucks versus a
      reraise) or an aggressive small blind that just called and
      drew three. If we took that 35% and replaced it with 36%,
      now the "1/3" rule dictates that KKA is enough for a reraise,
      so it looks like 35% is likely an "upper bound", but it's stll
      not impossible for the frequency to be 36% or higher.

      Suppose the blinds will reraise with AA+ (and say the BB will
      also reraise with KKA, not entirely unreasonable). How often
      does an AK-type hand or something like 33AKx have to muck?
      (I'm assuming you'd fold this to a reraise, right?) In other
      words, how often will one of the blinds have AA+ or the BB
      has KKA+?

      Useful numbers to calculate this and other situations

      Here's some approximate numbers that will clearly help in
      the future for any non-pair type and one pair type hands:

      There are C(47,5)= 1 533 939 ~ 1 534 thousand combinations
      that an opponent can have if you are dealt a hand.

      There are very roughly about 120 thousand combinations that
      will be two pairs or better (but this obviously depends on the
      exact ranks and suit distribution of the hand).

      There are roughly 60 thousand combinations for any pair of a
      rank that you don't hold and about 32 thousand for a pair of
      a rank that you do hold: e.g., with AKQ32, there are about
      60 thousand combinations of JJ or TT out there, but only
      about 32 thousand of each of AA, KK and QQ. [ To be exact,
      there are 60 420 for an unseen rank when holding no pair,
      but only 60 198 when holding a pair; when holding the rank
      of the pair as a side kicker, there are 32 388 when holding no
      pair, but only 32 274 when holding a pair, but you should
      check this in case my calculations are incorrect. ]


      For somebody to have AA+, there are very roughly about
      152 thousand combinations out of the 1 534 thousand
      combinations, so for none of the blinds to have AA+, it's
      about 81.2%, but it's probably unrealistic as some players
      might semibluff reraise, so add about 10 thousand
      combinations for each of the blinds, so one gets about
      80% or a blind wakes up with AA+ or a semibluff reraise
      about 20% of the time. For the big blind to have KKA and
      the small blind has <AA, it's another (1- (162/1534)) x
      (32x(1/4)(3/4)/1534) 0.39% so altogether, maybe 20.4% of
      the time you'll have to surrender after trying to steal with
      AK. ( Note: for a hand like KK you'll get an ace kicker exactly
      1/4 of the time, but here there's a few complications, so it
      won't be exactly that)

      What about the small blind getting KK/QQ or JJ? That
      depends on whether the AK holds a Q or J and say we're
      looking at specifically AKQ92. The small blind will cold
      call with JJ-KK about 124/1534 ~ 0.080834 of the time or
      about 8.1% of the time.

      What about the big blind? Let's say he plays 77-KKxyz
      (don't forget that KKA was a reraising hand!). How often
      does he get that? [ (32-6)+4(60)+2(32) ]/1534 ~ 0.215124
      but just consider when the SB folds; that's about 19.8% of
      the time.

      Altogether, you'll be in "trouble" or "surrender" 20.4%+8.1%
      +19.8% of the time or only 48.3% of the time, and it will
      depend on how often the big blind will defend with a pair
      less than 88. It now seems that AKQ98 is an easier hand
      to raise with and AKQ62 might be a muck unless maybe if
      it's "suited".


      Postdraw, the button can value bet KKA against the BB, and
      the chances that AK improves to KK or better is anywhere
      from about 34.7% to 35.9% depending on whether it's suited
      and how many "honor" cards are discarded if we forget about
      the distribution of the BB's hand (he could "cripple" the AK
      with say KKA!). The other aspect is that AK complements
      the entire range of one pair hands the button open raises
      with that may improve to two pairs or better, so the button
      should clearly bluff with an AK that doesn't materialize to
      88 or better, although "bluffing" with 88 could be correct
      depending on the combinations of AK-hands and weaker
      pairs that the button will be opening with.

      On the other hand, when heads up against the SB with a
      range of approximately JJ-AA, the AK might only value bet
      with AA depending on how often the SB will reraise with
      AA: if the SB always does, then the button can again clearly
      value bet KKA.

      Also, the BB is theoretically supposed to bet most (but not
      all) of his improved hands when he draws three, in which
      case the button with AK makes a theoretical payoff with
      JJ or better (he needs to so as not to be exploited since
      most of his range when drawing three will be close to 88A to
      AA). When the SB bets out, he doesn't really have to pay
      off with JJ in practice even though it is a game-theoretic
      payoff since not too many players in the SB will defend
      with TT and bluff into a three-card draw (although it can
      happen). There is another problem with value betting KKA
      versus the SB though: if he "autochecks" (he's never going
      to have a hand without "showdown" value, right?) and his
      range includes AA, KKA is now not worth betting since he
      doesn't need to pay off with QQ very often depending on
      how often he defends the SB with JJ. I think in practice,
      against most opposition, I'd strongly consider mucking
      JJ-QQ versus a bet since not many players bluff "enough".

      Now, you may be concerned about value betting KKA on
      the button versus the big blind, but it's very likely the right
      play (except maybe a BB that never defends with less than
      QQ and "autochecks", but then you can "run over" such an
      opponent with more AK-type hands!), especially against a
      BB that will defend with 99+ (plus we know in theory, the
      BB must defend with even more one pair hands, and maybe
      even AK-suited).

      Now the big practical factor is how often the blinds will be
      "exploited" postdraw when they draw three. In theory, the
      SB should pay off with KK and some QQ hands and the BB
      should pay off with JJ+ and some TT hands (gulp!) but of
      course, we don't see that in practice! It looks like this is
      a common scenario where the button can pick up some
      small pots against exploitable opponents in the big blind.

      Now I'm wondering if any AK(suited)(Q/J/T/9)xy hand is
      worth stealing with on the button. If so, maybe
      that 35% number is a little low and maybe I'd even steal
      with A Spade Q Spade J Heart T Diamond 8 Club from the button versus some very
      tight blinds.


      1) From "toy game" analysis, it is highly suggestive that the
      opening range from the button in limit draw is about 35%
      of all combinations (solution to the toy game is just above
      0.3425). Zadeh likely computed this as evidenced from his
      section on "Ante and Straddle" in his book "Winning Poker
      Systems" but didn't mention semibluffing hands: his table is
      "extremely rough" and suggests 77+ as the opening range,
      even though technically it's not the "same limit".

      2) Without ANY semibluffing hands outside of straight flush
      draws (there's very few), one would need a range that is not
      just 77+, but also includes 66 as well from a simple
      computation based on a 35% raising frequency.

      3) In practice, raising with all pairs of sevens will run into
      problems against quality opposition: they'll figure out you
      raise with 77, so you get 3-bet by not only AA and KKA, but
      also by semibluffs such as good A-high flush draws or a
      "semilegitimate" pair or any semibluffing hand that doesn't
      specifically have any seven or eight if they looked carefully
      at the ranks of the cards. This suggests we only selectively
      play some pairs of sevens such as 77AKx or 77AQx so you
      won't "get caught" so often.

      4) Good flush draws are hands that "fit the bill" instead of
      playing specifically a pair of sevens. There's probably
      almost enough of the decent ones that we can be very
      selective of which pairs of sevens to play. Even with all
      pairs of eights, we still need a few more weak one pair hands:
      say 77AKx, 66AKx, ...33AKx should be enough; maybe even
      throw in 22AKx for good measure.

      The "SCENARIO" in the very long post is just to show that
      semibluffing with many flush draws is important strategically,
      especially when heads up against an opponent with a high
      pair so that he will at least need to consider paying off with
      an unimproved QQ or KK. There's also some upper limit to
      which ones can be played, so we prefer to have flush draws
      with at least two cards in the 9-A range (since the big
      blind must defend with 99) or straight draws such as KQJT,
      AQJT9 and likely AKQJx, AJT98 or KJT98. Then, there should
      be "enough" semibluffs to replace all the pairs of sevens.

      5) Some problems with raising with 77 indiscriminately will
      also occur with some 88 hands, especially those that don't
      have many high kickers. This suggests we should find some
      "other hands", but we can't choose decent semibluffs such
      as flush draws and quality straight draws, since there's
      already "enough" of those. Obviously, hands such as XXAQy
      with X = 2 to 7 come to mind, but it's still not enough to
      get to the 35% figure.

      6) What's left? AK-type hands seem to be the solution, but
      we want to be as selective as necessary since these are
      problematic for several reasons. Suggested are having the
      following: AK(suited)(Q/J)xy, AKQ(J/T/9)x and AKJT9 since
      there are enough "key cards" that even a BB playing a range
      of about 88+ and a SB playing about JJ+, you'll win the blinds
      often enough and get a chance to make a hand with some
      legitimate showdown value (and even if you don't, there is
      some chance a bluff will work postdraw). There might be a
      few other candidate AK-type hands to consider to get to
      the 35% mark, but now at least with the AK-type hands,
      we're almost there.

      7) Now there's a good range of hands to bluff with postdraw
      when drawing three: a sophisticated opponent can't use
      the little information from looking at his side cards that help
      him to decide to call with with an unimproved pair after a
      check-bet sequence.

      Of course, almost all of our opponents haven't considered
      anything close to the above, but having the above helps
      considerably: we adjust according to how the blinds will
      play. If they call too often predraw and postdraw, we cut
      back on the bluffs and semibluffs. The ideal opponent in
      the big blind is a player that will call with a wide range of
      pairs and will check-fold with an unimproved one pair too
      often: simply raise with more AK-type hands and value bet
      hands such as KK for "cover". Sure they can catch on, but
      they don't know exactly how selective we are about which
      AK-type hands we're playing, how many pairs weaker than
      88 we raise with. Eventually, they might even check call
      with JJ or QQ routinely, but then we can be cruel: tighten
      up on semibluffs and bluffs (maybe even only a little bit!),
      so that their play is theoretically even more unsound.
    • Borisian
      Joined: 28.01.2012 Posts: 947
      What would your ideal 5CD HUD have on it? Also what would you have in the pop-up?
    • jbpatzer
      Joined: 22.11.2009 Posts: 6,944
      Originally posted by Borisian
      What would your ideal 5CD HUD have on it? Also what would you have in the pop-up?
      Predraw: VPIP/PDR/coldcall/3bet/cap
      Predraw: OPR by position/Fold BB to button steal/Fold BB to SB steal
      Draw: 0/1/2/3
      Postdraw: Donk/FtDonk/Cbet/FtCB/RaiseCbet

      Something like that. Why, are you going to write something for us? :f_love:
    • Borisian
      Joined: 28.01.2012 Posts: 947
      Yeah PokerHands is playing up again!! (Even though I haven't buggered with it since I last used it) and FPDB (though working) just gets on my nerves.

      Mucked cards too? I'll assume yes ;)
    • sundancekid3
      Joined: 09.01.2011 Posts: 265
      It would be great if it have 3bet by position .
    • Crox1710
      Joined: 11.10.2012 Posts: 22
      Heres one from today

      Poker Stars $0.10/$0.20 Limit 5 Card Draw - 5 players
      DeucesCracked Poker Videos Hand History Converter

      CO posts a big blind (1 SB)

      Pre Draw: (2.5 SB) Hero is SB with A:spade: 5:spade: 4:club: 3:spade: 2:heart:
      UTG calls, CO raises, BTN folds, Hero 3-bets, BB calls, UTG folds, CO calls

      First Draw: (10 SB) (3 players)
      Hero stands pat, BB draws 1, CO draws 1
      Hero bets, BB folds, CO raises, Hero calls

      Villains reraise after drawing one tells me he has made a boat from his trips or two pair, so my call is clearly bad in this isolated example.
      But if I am seen to be folding pat hands after a reraise then am I exploitable?
      It is micro stakes so people probably aren't paying that much attention, but tehn again it is a fairly noticeable fold.
      Any thoughts? I'm siding with 1 BB lost every time someone hits a boat at 9% is a good price to pay for not making my pat hands exploitable.
    • Borisian
      Joined: 28.01.2012 Posts: 947
    • jbpatzer
      Joined: 22.11.2009 Posts: 6,944
      Originally posted by Crox1710
      Heres one from today

      Poker Stars $0.10/$0.20 Limit 5 Card Draw - 5 players
      DeucesCracked Poker Videos Hand History Converter

      CO posts a big blind (1 SB)

      Pre Draw: (2.5 SB) Hero is SB with A:spade: 5:spade: 4:club: 3:spade: 2:heart:
      UTG calls, CO raises, BTN folds, Hero 3-bets, BB calls, UTG folds, CO calls

      First Draw: (10 SB) (3 players)
      Hero stands pat, BB draws 1, CO draws 1
      Hero bets, BB folds, CO raises, Hero calls

      Villains reraise after drawing one tells me he has made a boat from his trips or two pair, so my call is clearly bad in this isolated example.
      But if I am seen to be folding pat hands after a reraise then am I exploitable?
      It is micro stakes so people probably aren't paying that much attention, but tehn again it is a fairly noticeable fold.
      Any thoughts? I'm siding with 1 BB lost every time someone hits a boat at 9% is a good price to pay for not making my pat hands exploitable.
      I haven't done the sums, but I expect that, to be unexploitable here you need a value raising range (probably a bluff raising range too, but I wouldn't be too bothered with that), and a calling range that is slightly wider than his value raising range so that you have some bluffcatchers in it. Personally, just off the top of my head, I'd probably raise a FH about 777XX or better, and call down to about a J high straight. The median pat hand is an A high straight so that way you're folding about the bottom 25% of your hands (assuming you're never snowing here, which you probably should be sometimes). Maybe you even need to call a little wider. As I say, I haven't done the maths, but I expect the idea's right, even if I've got the position of the boundaries slightly wrong. Of course, if your read on CO is that he's never bluffing (I expect he'll show you a flush a lot here), you can fold even up to small flushes. Anyway, whatever the right answer is, A2345 is the rock bottom of your value range, and a clear fold. I expect Hurrrrricano will tell me I'm a fish.
    • Crox1710
      Joined: 11.10.2012 Posts: 22
      Thanks. I had assumed that CO's raise pre draw meant that he had two pair or a set rather than a 4flush or straight draw making even AKQJ9 flush a bluff catcher.

      Damn assumptions! :f_cry:

      With Pat Hands in general I usually just lead out, unless someone else has stood pat as well. I get called most of the time by donks who can't fold a set or AAKK so the value bet does pay off at microstakes. Any thoughts on a check call, check raise? Clearly pat hands is something I need to improve on as up to know its just like a free pot.
    • 1
    • 2