Select Page
Poker Forum
Over 1,291,000 Posts!
Poker ForumShort-Handed NL Hold'em

Hand vs. range equity and preflop strategy

Results 1 to 6 of 6
  1. #1
    Renton's Avatar
    Join Date
    Jan 2006
    Posts
    8,863
    Location
    a little town called none of your goddamn business

    Default Hand vs. range equity and preflop strategy

    OK, allow me to first warn readers that this isn't really a definitive theory post or anything, just some stuff I've been thinking about, most of which lacks any grand conclusions. I just have nowhere better to put this stuff and I figure this forum needs a breath of new life back into it.

    So I was playing with the propokertools simulator. (http://www.propokertools.com/simulations, sorry for the spam but it's necessary for understanding the rest of the post) Most people use this simulator to find PLO equities, but there's an interesting tool that's of interest to Hold'em players, its called "Graph HvR."

    This tool takes a given range and hand, and plots how much equity the hand flops vs this range how much percentage of the time.

    For example:




    This depicts the spectrum of flop possibilities when you hold 97s vs a 15% UTG opening range.





    And this is AJ offsuit vs the same range.

    After looking at these graphs for a number of different types of starting hands, the following becomes apparent:

    - Low card hands like 97 have very reliable feedback as to their equity on the flop. 50% of the time they flop 20 percent equity which is basically garbage. This is intuitively obvious, when you hold 97 on a AJ2 flop you're drawing nearly dead to 2nd pair, you can never make a large mistake folding that hand vs a cbet. Conversely, when you flop AhTs2d with 9s8s, you only have 10% equity vs top pair but you know exactly what turn cards will allow you to improve to beat that hand. This hand actually has value in floating or raising the flop despite the fact that the equity is so low.

    - When low cards get boards where they have 30 or more percent equity vs a range, this means that they have a draw of some kind. Draws are valuable beyond the actual equity they have in the pot, as we all know.

    - High card hands have very poor feedback on the flop, and have very poor equity distribution. For example, in the depicted AJ hand, 70% of the time it flops a measly 30-40% equity vs the UTG opening range. And where, 30-40% was valuable with 97s, AJ high isn't nearly as floatable vs this range as a flushdraw or oesd would be. This is why we tend to fold AJo to utg opens, despite it being a slight favorite vs a utg opening range.


    So as I continued looking at these graphs it struck me that the area of these curves has to have some meaning as relates to EV/playability of a hand. So I started doodling on these charts with mspaint. For example:






    I've VERY roughly defined the categories as air, x, y, and z where:

    air - flops where we are very likely to have our equity share in the pot stolen or otherwise having to pay too much to realize it
    x - flops where we don't have much but can make some profitable bluffs
    y - flops where we make a medium strength hand thats ahead of his range and we attempt to play a small to medium size pot with it and get to showdown
    z - flops where we attempt to extract value, above and beyond just trying to realize our equity and win a share of the pot.

    The most striking things comparing 97 suited to AJ offsuit that I noticed:

    1) AJo distributes its equity in such a way that we are very often forced to fold a significant equity share. We are regularly folding with 30-40% equity which sucks. 97's air has 20% equity at best so we don't lose as much equity when we miss and have to fold.

    2) That said, AJo has significantly more equity to distribute to begin with. Theoretically there should be a point where raw equity wins over. AQo vs a utg range would be an example of this (nevermind the fact that AQ high is often strong enough to call the flop anyway, even if it weren't, pair value gets us past the point of +ev in most cases).

    3) AJo has fewer opportunities to effectively semibluff when it misses. This should be obvious to most, but it is seen very clearly in the fact that the "X" slice of its range is tiny when compared with 97s.


    Anyway, that's where I am now. I'm thinking that there should be a way to directly valuate the area under the curve. Obviously the area under the part where we have over 80% equity should be worth significantly more than the area under 60%. I'm wondering if we can just count the squares under the line and attach a value to each square, add up the values and result in an accurate estimate of the value of a hand vs a range, compare this value with the amount to call preflop (whether this be coldcalling an open, flatting a 3-bet, or flatting a 4-bet), and decide whether to play the hand or not.

    What do you guys think?
    Last edited by Renton; 03-02-2014 at 08:08 AM.
  2. #2
    Good post man. I'll have to take a look at this some more and ponder. I'm assuming the AQo curve is similar to the AJo? How does adding suitedness improve AJo?
    Quote Originally Posted by Jay-Z
    I'm a couple hands down and I'm tryin' to get back
    I gave the other grip, I lost a flip for five stacks
  3. #3
    Join Date
    Aug 2007
    Posts
    8,697
    Location
    soaking up ethanol, moving on up
    Quote Originally Posted by Renton View Post
    - When low cards get boards where they have 30 or more percent equity vs a range, this means that they have a draw of some kind. Draws are valuable beyond the actual equity they have in the pot, as we all know.
    unless they have over 70% and then they're a made hand already right? This reminds me of something i read on here once -= i'm going to have to go and read the plo post someone wrote about nutty vs crunchy (or was it sticky) hands, and the no such thing as a made hand posts.

    Quote Originally Posted by Renton View Post
    - Low card hands like 97 have very reliable feedback as to their equity on the flop. 50% of the time they flop 20 percent equity which is basically garbage.
    looks like 30% equity, i could be failing at graph reading though.. actually, it looks as though you have chosen different equity values for the x section for 79s vs for AJo , which would extend the AJ x-area out a long way further right and make it a far larger share.

    Quote Originally Posted by Renton View Post
    1) AJo distributes its equity in such a way that we are very often forced to fold a significant equity share. We are regularly folding with 30-40% equity which sucks. 97's air has 20% equity at best so we don't lose as much equity when we miss and have to fold.
    is this more related mostly to the fact that AJ flops made but dominated hands more often whereas 97s is either made and able to improve (similar to 98 vs TT on 962rb in this post http://www.flopturnriver.com/pokerfo...ot-197158.html) or drawing to a hand that dominates? Or maybe it's just because it's harder to be confident with AJhi postflop, even when we are ahead or have 6 clean outs so we have to fold this equity..


    Quote Originally Posted by Renton View Post
    Anyway, that's where I am now. I'm thinking that there should be a way to directly valuate the area under the curve. Obviously the area under the part where we have over 80% equity should be worth significantly more than the area under 60%. I'm wondering if we can just count the squares under the line and attach a value to each square, add up the values and result in an accurate estimate of the value of a hand vs a range, compare this value with the amount to call preflop (whether this be coldcalling an open, flatting a 3-bet, or flatting a 4-bet), and decide whether to play the hand or not.

    What do you guys think?
    i think my head is broken
    but i don't like the idea that greater area under the curve = greater calling strength, again because of intuitive issues with domination in the example where AJ has a far bigger area under curve than 79s. First thing that jumps to mind is that there is a lot of information in the x-intercepts when y = say, 75% min equity and maybe a couple of other chosen points (65%, 50%) but i can't yet figure out how to use them.

    In the example you gave:
    for 97s: x(75)= 6, x(65) = 9, x(50) = 30, x(30) = 55
    for AJo: x(75)= 12, x(65) = 28, x(50) = 30, x(30) = 95
  4. #4
    Renton's Avatar
    Join Date
    Jan 2006
    Posts
    8,863
    Location
    a little town called none of your goddamn business
    Quote Originally Posted by griffey24 View Post
    Good post man. I'll have to take a look at this some more and ponder. I'm assuming the AQo curve is similar to the AJo? How does adding suitedness improve AJo?



    This is AQo/AJo overlay, surprisingly dominating KQ/AJ doesn't make as large of a differences as one might think. After all, it's only distributing 3 more percent of equity than AJo. So this brings us to the easy conclusion that not all 30-40% equity holdings on the flop are created equal. AQ overs are far more floatable than AJ overs, as they have somewhat (3%) more showdown value, but the outs are also of greater quality, and have more implied odds.

    Heres AJs on top of AJo, again vs a 15% opening range.




    Again similar, but the suitedness adds an equity bump where it counts, in the X part of the line.
  5. #5
    Renton's Avatar
    Join Date
    Jan 2006
    Posts
    8,863
    Location
    a little town called none of your goddamn business
    There may be a slight problem with all these charts, sigh. Apparently PPT doesn't use the same suitedness syntax as pokerstove.

    Anyway that shouldn't drastically affect the shape of the graphs, you have the tools to do your own investigation at least.
    Last edited by Renton; 03-02-2014 at 10:11 PM.
  6. #6
    Cool thanks for the overlays!
    Quote Originally Posted by Jay-Z
    I'm a couple hands down and I'm tryin' to get back
    I gave the other grip, I lost a flip for five stacks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •