Announcement

Collapse
No announcement yet.

Applying MDF correctly?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Applying MDF correctly?

    I apologize in advance for the lengthy post and syntax, im hoping smarter minds then mine can help illuminate this for me.

    So for awhile now ive been a bit confused about exactly how to apply MDF from a theoretical perspective, assuming opponents bluff at the correct frequency with appropriate combos and using solid ranges. Ive gone through hand examples and through homework questions where we regularly make sure we are continuing with enough of our range on each street in one way or another depending on what strategy we are employing. This seemed to make sense to me intuitively and i was able to answer homework questions close enough to what Jonathan Little would qualify as correct. Recently however, when i find myself looking at situations where we are at a significant range disadvantage outright range vs range or based on board texture (or both) ive felt like this concept seems to break down. I'm now confused whether i understood this at all to begin with or whether im applying the concept incorrectly

    Example

    6max 100nl game, sitting in HJ with 100bb stack using the GTO Poker coaching charts i make a standard open and face a standard 3bet from the CO who for the purposes of this example ill assume is a thinking player also following the GTO CO vs HJ range chart also with a 100bb stack

    HJ open and CO vs HJ 3bet respectively


    Click image for larger version

Name:	HJ.PNG
Views:	71
Size:	350.4 KB
ID:	43100 Click image for larger version

Name:	CO vs HJ.PNG
Views:	57
Size:	325.4 KB
ID:	43101


    Based on the Theory learned from JLs cash game course, as well as the 6max course from JustGTO we are given information drawn from Micheal Acevedo book Modern Poker Theory which is used as a guide to explain important concepts, some of the information is shown as follows:


    Click image for larger version

Name:	HJ vs CO theory.PNG
Views:	54
Size:	102.2 KB
ID:	43102 Click image for larger version

Name:	HJ vs CO theory 2.PNG
Views:	49
Size:	94.8 KB
ID:	43103

    I used this and other content to then construct what i think is a reasonably simplified theoretical 4bet and 3bet flatting range for this given situation:

    Red corner = 4bet
    Blue corner = Flat


    Click image for larger version

Name:	HJ vs CO 4bet.PNG
Views:	58
Size:	331.6 KB
ID:	43104


    Ive tried to hold to the Ratios outlined by the theory as well as give as reasonable a board coverage as i can in an implementable style strategy without splitting hands into frequencies. For this particular situation im using JJ+ AK as 4bet for value (40) AQo, ATs and KJs as 4bet bluffs (20) and AQs AJs KQs 88-TT 76s-98s and JTs as flats (46)

    1. Is this Range reasonable Yes/No if No why?

    if we flat and the board comes Ah Qs 3h, we check and the CO fires for 2/3 pot

    If i understand MDF correctly, we need to defend at minimum 60% of our range facing our opponents 3bet cbet to not exploitably overfold.

    however using Equilab based on the range ive constructed and the board texture we have only approximately 23% equity with our flatting range vs CO entire range OOP.






    2. a) How do i go about seeing what hands i should continue with? as well as going into possible turns and rivers.

    b) Is there any point trying to apply MDF holding this amount of equity oop on this board with this range? Why/Why not

    c) What am i missing or incorrectly inferring? Does all this make sense?

    Any answer able to unravel this for me would be much appreciated

    Thanks!


  • #2
    Hey buddy.

    Here's the implementable range that I previously came up with for the exact same spot:

    Click image for larger version  Name:	HJ v CO 3-bet.png Views:	0 Size:	88.9 KB ID:	43107
    You'll see that my percentages line up very close to the GTO solution. 4! = 22.5 versus 22.2. Call = 14.1 versus 14.5. Fold = 63.4 versus 63.3.

    If you plug your range into the hand analyser you'll see you're perhaps calling a tad too wide. But it's not a major issue.


    Now I'm not sure what's happening with your numbers for the next part of your post, but for me it's easy to defend at MDF using your range.

    Here's flopzilla containing your range and suggested flop:

    Click image for larger version  Name:	Flopzilla 20102020.JPG Views:	0 Size:	125.0 KB ID:	43108

    So on this flop we have a total of 42 combinations of hands.

    Your MDF calculation was correct: if the opponent cbets 2/3 pot we need to defend with 60% of our range.

    42 * 60% = 25.2 - so therefore we need to defend with at least 26 hands to meet MDF.

    If we defend with any pair or better we're already at 26 hands before we factor in any bluffs we want to make.

    Does this make sense?

    I'm happy to try and answer any other questions you have but we need to make sure we're on the same wave length first.

    Comment


    • kkep
      kkep commented
      Editing a comment
      Do we need to to cover MDF in 3bet pots OOP tho?

      I'm probably only check calling 12 combos (29ish%) on that flop vs the Co 3! and flop cbet.

      I mean TT only has like 20% equity vs the Co range and range vs range we have like 25% equity with no nutted hands.
      How are you going to feel check calling a turn bet on any card really other than a T and the even tho the V has few FD the Th is still dicey?


      I think some spots are just bad and this is one of those with 2 over cards to our pairs on the flop.

    • LondonImp
      LondonImp commented
      Editing a comment
      Edit: Made a hash of this.
      Last edited by LondonImp; 10-21-2020, 08:05 AM.

  • #3
    Click image for larger version

Name:	image0.jpg
Views:	47
Size:	46.2 KB
ID:	43139

    Comment


    • LondonImp
      LondonImp commented
      Editing a comment
      Okay scratch what I said previously.

      Sorry.

      This is vitally important.

    • Wood
      Wood commented
      Editing a comment
      This is what i needed to read, thanks heaps

  • #4
    I think it's more likely because we will have 4! the top of our range along with a couple of *bluffs*, fold our worst hands and call mostly with hands that will play decently enough post flop.

    I think your defending range is way to tight. I think we should be defending roughly 135 combos pre flop. You're over folding roughly 13% which means on average your giving up like .33BB/100 with combos that have plenty of equity. We need to be defending somewhere between 48-50% maybe a little wider if the V is fishy or a little tighter if the V is nitty.

    Let's presume the hero opens 2.5x and the Co 3! 3.7x I don't know what the antes are but that only gives us a better price. So we need to call 6.75BB more to win 20BB.
    Given that we are OOP I would like about 35% equity and the range below has 42%


    YMMV 3bet pots are difficult so take it for what it's worth from rec.

    I recognize the x/r range isn't balanced but I think we should only be raising hands we're willing to go with on this exact hand. It doesn't make sense (to me) to be over bluffing into the V with the nut advantage and a range that can mostly continue. Again 3! pots are played differently than single raised pots and this is a learning exercise for me too.


    Click image for larger version  Name:	open.png Views:	0 Size:	74.8 KB ID:	43145
    Facing 3!
    I'm also still over folding but I think that is ok. Our check calling range is decent but not great. We either improve with most of it or we're done with the hand.

    @London I don't think we want to 4! bluff with are good suited hands. I believe are bluffs should come from the off suite hands that don't play as well post flop and we can happily give up when 5!
    Maybe 4! with AQo is better than AJo, I'm not sure. I still like calling with KQo because it blocks KK QQ AK AQ.

    I also think we can call with all of our suited Ax as long as we play them cautiously post flop.
    Click image for larger version  Name:	4bet.png Views:	0 Size:	80.0 KB ID:	43146

    Flop

    All of our PP pairs, well our top pair is TT and it only beats 99 and 88 in the V's range and I don't think it can withstand anymore pressure on basically any turn card.. Compound that with the fact that an A and K hit the flop which smashes his range. So I don't think we get to peel any of our PP.

    I'm not sure if we should x/r A3s or not. I decided to keep it in my check calling range on the flop mostly because I don't have a ton of draws I want to pile chips in the pot with yet.

    Edit: It might be wise to even fold some more of the junky AX s combos but I think we can call flop looking to ditch some more of those if he bets turn too.

    Click image for larger version  Name:	FlopAction.png Views:	0 Size:	81.1 KB ID:	43147

    We probably shouldn't fold any AXs on the flop actually and it looks like we should fold JJ.

    Click image for larger version

Name:	Mod.png
Views:	39
Size:	80.4 KB
ID:	43149



    Thoughts?


    Last edited by kkep; 10-21-2020, 08:36 AM.

    Comment


    • Wood
      Wood commented
      Editing a comment
      I feel like you might have too many hands in your 3bet calling range as i said below. Its gonna be very difficult to continue appropriately on turns and rivers with all your Ax and Fds. Ill defer to other Users on this one.

  • #5
    This might be fun and it surprised me a bit but it really shouldn't have. Using my cbet calling range which card or cards would you say are best for us?

    Comment


    • #6
      Hey guys thanks for the feedback, super helpful, i think im starting to see the issue here.

      The excerpt that kkep posted on cbet defence seems to drive at the core of the issue. It seemed that myself (and possibly Londonimp) were following the MDF rules too religiously even in situations that were clearly detrimental. It is amusing that the board AQ3 that made me realise this is also the example board that Modern poker theory used to point out why MDF shouldnt apply when equities and range distribution are strongly against us.

      Londonimp thanks for the initial post, it seems that we had very similar configurations for that spot although i suspect yours is slightly more correct and that i added slightly too many suited connectors. i was also aware when i did the math that i was slightly over calling more then was advised when i posted the ranges, but my main gripe wasnt the range itself but the concept of defending MDF in spots where common sense would give us pause on following the MDF rabbit hole to destruction. In my example we only have 23% equity on the flop vs Villians range, it makes no sense to me or i think anyone to defend approximately 60% of my range or thereabouts on each street when i perform so poorly against their range. I would have to defend something like TT for 2 or 3 streets im guessing which seems like a good way of setting money on fire. In reality i would defend probably the same as what you intially commented kkep, and for the exact same reason as you mentioned:

      https://forum.pokercoaching.com/foru...3110#post43110



      Click image for larger version

Name:	equilab output.PNG
Views:	42
Size:	45.0 KB
ID:	43152

      I think it's more likely because we will have 4! the top of our range along with a couple of *bluffs*, fold our worst hands and call mostly with hands that will play decently enough post flop.

      I think your defending range is way to tight. I think we should be defending roughly 135 combos pre flop. You're over folding roughly 13% which means on average your giving up like .33BB/100 with combos that have plenty of equity. We need to be defending somewhere between 48-50% maybe a little wider if the V is fishy or a little tighter if the V is nitty.


      im open to changing my defending range here, i was simply following the guidelines given from Modern Poker theory that i posted above which indicates a 22% 4bet range with a 15% flatting range, which works out to the 106 combos that make up my range. The reasoning for the larger 4bet range and smaller flatting range is outlined due to the difficulty in realizing our equity OOP in a 3bet pot. 135 combos would seem to be too many to me? and 48-50% seems like far too many hands. Can you direct me to the theory that suggests this?

      Let's presume the hero opens 2.5x and the Co 3! 3.7x I don't know what the antes are but that only gives us a better price. So we need to call 6.75BB more to win 20BB.
      Given that we are OOP I would like about 35% equity and the range below has 42%


      Sure but thats just raw equity, not realised equity. OOP that is gonna be a major factor to determining what hands we can profitably call with. I assume Modern Poker theory and the Solvers factored this in before it spat out the % breakdown of the 106 combos i posted above which for simplicity and clarity is what im trying to ground my thinking off of to understand the MDF issue. Im going off a 22% 4bet 15% flat and 63% fold. If you are flatting 55% of your range to a 3bet here OOP i think you are gonna find yourself in some tough spots quite often. At least versing an opponent following general GTO ranges and principles.

      I suppose to circle back and attempt going forward here, If using MDF as a guideline in spots where we have massive range disadvantage and range distribution problems is clearly detrimental, When exactly is it useful? When Equities run close?

      If we decide for example re my first example situation that following MDF here is ruinous, which i now think it may be. How then do we go about finding out how and what to continue here?

      Assuming that My preflop range is roughly correct in terms of how i arrived at this spot. GTO chart open (correct) vs GTO chart 3bet (correct) vs Crafted 3bet flatting range and 4bet range (almost certainly close to correct) as closely based on theory as possible which myself and Londenimp seem to have done. We end up in a spot where we have 23% Equity on the flop vs our opponents range.

      If i defend what intuition tells me to defend here im defending with only 29% or so of my range, raising my 2 combos of AQs and maybe 3 or so of my draws (flush draw or gutshot+Backdoor fd) and calling the rest, this means im seriously underdefending and folding out far too much of my range.

      However if i follow MDF principles i then have to call with hands that are almost dead vs my opponents range, I refer to TT here as a good example, MDF would have TT calling 2 streets of 2/3 size bets here.

      It would appear both options are losing. So whats the solution here? is there some sort of modified MDF formula that can be applied that takes into account your equity vs your opponents equity and allows you to lose the least amount of money? It seems in some spots we are naturally gonna be on the back foot whether due to our opponents range or the board texture or both. I feel like there is probably a simple solution or approach to this that is alluding me. What do you think?

      Sorry for any spelling or grammar mistakes and thanks for the feedback guys, super helpful.

      Comment


      • #7
        I certainly have issues over calling in 3! bet pots, my HUD shows me calling about 75%. I just came to realize this about a week ago. I think part of that leak is not 4! enough tho.

        I came up with the 48-50% both from a Matt Afflect Webinar and a Chapter by Ben Shuster in Excelling at Tough NLHE. That said I may be making a mistake by not factoring in my 4! as part of my defending/calling range but If I'm correctly recalling MA Webinar we can and should be defending at least many of the suited hands I choose. I'm going to watch that again today.

        MA used an analogy that OOP calling a 3! is sort of like defending the BB. We know a lot of hands we are calling with are in bad shape but they have plenty of equity when given the right price and we get to fold a lot on the flop.

        If you run my calling range (which is really only 12% of hands) vs the Co 3! you will see that it has way more than 33% break even equity. I like to add a 3% edge where as I think the Pros look for about 2%.
        There is one combo I think I should remove and that is 87s 76s which gets me down to 52% and those combos can't out flop any of the pairs in the V's 3! range so yeah I think they need to go.

        MDF is far more useful in single raised pots. It gets more complicated in 3! bet pots. (This is in a hand where both players play generally well) Worrying about MDF can be costly if we incorrectly identify how the V plays. If they are nitty we get to over fold and if they are laggy calling correctly is almost mandatory.

        IDK if you are a premium member or not but back in July Matt did 2 webinars on 3! pots.

        I don't play close to perfectly by any stretch and we should for sure be 4! some of worse suited hands like K8s Q8s when we are looking to be more polarized. My ranges are under development.
        BTW AKo is technically a *bluff I suppose so in this spot I think my range is decent if I also fold 87s and 76s

        Comment


        • #8
          There was another key concept. Most people play to passively OOP in 3! pots.The pre-flop aggressor typically auto Cbets for 1/3 pot. We get to x/r about 15-20% of the time. This wasn't an ideal spot for that but I still found 9%.

          Comment


          • #9
            Click image for larger version

Name:	pre.jpg
Views:	38
Size:	224.2 KB
ID:	43161

            Comment


            • #10
              So this is only for 40BB Co v BTN

              The Chart on the left was what the group fed Matt and the chart on the right is how we should play it.
              Blue is fold to 3! red is 4! all in but what I'm looking at is the calling range and compare that to how we as group thought should be mucked. Click image for larger version

Name:	group.jpg
Views:	36
Size:	254.2 KB
ID:	43163

              Comment


              • #11
                One more note, I don't study cash games but since this 100BB I wouldn't think the strategies for this spot wouldn't vary greatly?

                Comment


                • #12
                  For what it's worth here is the play-ability profile. BTW I'm doing this mostly for my own edification.




                  Click image for larger version

Name:	Untitled.png
Views:	34
Size:	209.7 KB
ID:	43197
                  Now look at your 4! Bluffing range.

                  I think using AJs and ATs and probably AKs is a big mistake. Ask yourself what part of the V's range is going to fold to a 4!

                  A4s A5s and probably A9s KTs How about AJo QJs KQo 88 99 I would think so?

                  AJs is doing really well against that range so we don't need to fold those out and everything else they call with has us in rough shape.

                  http://www.power-equilab.com
                  Equity Win Tie
                  MP3 58.76% 51.96% 13.60% { AJs }
                  CO 41.24% 34.44% 13.60% { 99-88, A9s, A5s-A4s, KTs, QJs, AJo, KQo }


                  If we get him to fold AQo which I think is reasonable that would be the only good result (9 combos, we block 3) but AJs still does well enough -

                  http://www.power-equilab.com
                  Equity Win Tie
                  MP3 54.11% 47.97% 12.28% { AJs }
                  CO 45.89% 39.75% 12.28% { 99-88, A9s, A5s-A4s, KTs, QJs, AQo-AJo, KQo }

                  Now suppose we used A7s as a bluff. I choose that over A8s and A9s because it unblocks 88 99, some of the pairs we could expect to fold.

                  TBH I'm not sure which pairs 5! jam, I think for sure QQ+ not so sure about TT JJ I would tend to fold those vs a 4! for 100BB but I think this concept stays in line.


                  This is a great result


                  http://www.power-equilab.com
                  Equity Win Tie
                  MP3 43.76% 41.01% 5.50% { A7s }
                  CO 56.24% 53.49% 5.50% { 99-88, A9s, A5s-A4s, KTs, QJs, AQo-AJo, KQo }

                  Lets look at ATs, that surely benefits more than AJs from some of the combos we get to fold but I still think it plays well enough post flop to just call with.

                  http://www.power-equilab.com
                  Equity Win Tie
                  MP3 50.44% 48.51% 3.86% { ATs }
                  CO 49.56% 47.63% 3.86% { 99-88, A9s, A5s-A4s, KTs, QJs, AQo-AJo, KQo }

                  I prefer using AJo over AQo both

                  http://www.power-equilab.com
                  Equity Win Tie
                  MP3 51.61% 44.34% 14.54% { AJo }
                  CO 48.39% 41.12% 14.54% { 99-88, A9s, A5s-A4s, KTs, QJs, AQo-AJo, KQo }




                  http://www.power-equilab.com
                  Equity Win Tie
                  MP3 61.91% 54.34% 15.13% { AQo }
                  CO 38.09% 30.53% 15.13% { 99-88, A9s, A5s-A4s, KTs, QJs, AQo-AJo, KQo }

                  Comment


                  • #13
                    I asked Matt tonight during his live stream - HJ open vs Co 3! What percentage of hands should the HJ defend other than the 4! range?

                    His answer was 55% about 20% 4! 35% call. I don't 4! enough and I'm still calling to wide.

                    Comment


                    • LondonImp
                      LondonImp commented
                      Editing a comment
                      Big total amount ain't it?

                      I'm pretty sure I don't flat or 4! enough.

                    • kkep
                      kkep commented
                      Editing a comment
                      It is and I have been defending way wider . I started making some new 4! defend charts. TBH I still think in my games I should have a little less 4! bluffs and absolutely 0 vs some players.

                      Matts answer was about right on with what he said in his webinar which was we should defend about oh it was either 50 or 52% and that pros will defend about 2% wider.
                  Working...
                  X